EcoProduction.

Environmental Issues in Logistics and Manufacturing

Paulina Golinska-Dawson
Kune-Muh Tsai
Karolina Werner-Lewandowska   Editors

Smart and
Sustainable Supply
Chain and Logistics—
Challenges,
Methods and Best
Practices
Volume 2
EcoProduction

Environmental Issues in Logistics and Manufacturing

Series Editor
Paulina Golinska-Dawson, Poznań, Poland
The EcoProduction Series is a forum for presenting emerging environmental issues
in Logistics and Manufacturing. Its main objective is a multidisciplinary approach
to link the scientific activities in various manufacturing and logistics fields with the
sustainability research. It encompasses topical monographs and selected conference
proceedings, authored or edited by leading experts as well as by promising young
scientists. The Series aims to provide the impulse for new ideas by reporting on the
state-of-the-art and motivating for the future development of sustainable manufac-
turing systems, environmentally conscious operations management and reverse or
closed loop logistics.
It aims to bring together academic, industry and government personnel from
various countries to present and discuss the challenges for implementation of
sustainable policy in the field of production and logistics.
Paulina Golinska-Dawson · Kune-Muh Tsai ·
Karolina Werner-Lewandowska
Editors

Smart and Sustainable

Supply Chain
and Logistics—Challenges,
Methods and Best Practices
Volume 2
Editors
Paulina Golinska-Dawson Kune-Muh Tsai
Faculty of Engineering Management Department of Logistics Management
Poznań University of Technology National Kaohsiung University of Science
Poznań, Poland and Technology
Kaohsiung, Taiwan
Karolina Werner-Lewandowska
Faculty of Engineering Management
Poznań University of Technology
Poznań, Poland

ISSN 2193-4614 ISSN 2193-4622 (electronic)

EcoProduction
ISBN 978-3-031-15411-9 ISBN 978-3-031-15412-6 (eBook)
https://doi.org/10.1007/978-3-031-15412-6

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature
Switzerland AG 2023
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse
of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and
transmission or information storage and retrieval, electronic adaptation, computer software, or by similar
or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, expressed or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface

The COVID-19 pandemic has caused numerous disruptions in supply chains and
has significantly changed customer’s behaviours. Moreover, these new challenges
had been merged with the existing pledge for sustainability and greenhouse gas
emissions reduction. As a result, companies and public entities need to adapt their
processes to meet new challenges in the macro environment. New smart technologies
are becoming increasingly important and support business operations amid staff
shortages and restrictions on the movement of people and goods. Smart technologies
are also increasingly being used in crisis management to reduce risk and make supply
chain processes more resilient to disruption caused by natural disasters or pandemics.
This book entitled Smart and Sustainable Supply Chain and Logistics—Chal-
lenges, Methods and Best Practices: Volume 2 presents the original methods, tools
and case studies on topics, as follows:
. Smart solutions for supply chain management,
. Modelling, simulation and optimization of supply chain, production and logistics
operations,
. Sustainable, social and legal challenges in supply chain management and logistics.
This book includes the selected papers which have been submitted to the 15th
International Congress on Logistics and SCM Systems (ICLS 2021) organized by
the Faculty of Engineering Management, Poznań University of Technology and the
International Federation of Logistics and SCM Systems (IFLS). We would like to
express our gratitude to the Board of the International Federation of Logistics and
SCM Systems (IFLS) for the invaluable contribution to the volume:
. Honorary Chairman—Prof. Karasawa, Yutaka, Kanagawa University, Japan.
. Advisors—Prof. Kachitvichyanukul, Voratas, Asian Institute of Technology,
Thailand; Prof. Katayama, Hiroshi, Waseda University, Japan.
. Chairman—Prof. Tsai, Kune-Muh, National Kaohsiung University of Science
and Technology, Taiwan.

v
vi Preface

. Vice Chairmen: Prof. Lai, Kin Keung, City University of Hong Kong; Prof. Liu,
Xiaohong, Central University of Finance and Economics, China; Prof. Rim,
Suk-Chul, Ajou University, Korea; Sethanan, Kanchana, Khon Kaen University,
Thailand; Prof. Wakabayashi, Keizo, Nihon University, Japan and Wu, Yenchun
Jim, National Taiwan Normal University, Taiwan.
. The Board Members.
This scientific monograph has been doubled blind reviewed. We would like to
thank all Reviewers whose names are not listed in the volume due to the confiden-
tiality of the process. Their voluntary service and comments helped the authors to
improve the quality of the manuscripts. Although not all of the received manuscripts
appear in this book, the efforts spent and the work done for this book by the Authors
and Reviewers are very much appreciated.

Poznań, Poland Paulina Golinska-Dawson

Kaohsiung, Taiwan Kune-Muh Tsai
Poznań, Poland Karolina Werner-Lewandowska
Contents

Smart Solutions for Supply Chain Management

Impact of Internet of Things on Food Supply Chains . . . . . . . . . . . . . . . . . . 3
Elkafi Hassini, Mohamed Ben-Daya, and Zied Bahroun
An Agent-Based Approach to Evaluate Freight Consolidation
Center Strategy for Last Mile Deliveries in Jaipur City, India . . . . . . . . . . 13
Pankaj Kant and Sanjay Gupta
A Proposed Framework for Designing Blockchain Solutions
for Logistics in post-Covid Scenario and Future Pandemics . . . . . . . . . . . . 29
Javed Aslam, Aqeela Saleem, Nokhaiz Tariq Khan, and Yun Bae Kim
Application of Big Data Analytics in Modern Logistics Solutions . . . . . . . 37
Jozef Fras and Waldemar Osmólski
Case Study Validation of a Predictive Maintenance Implementation
Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Marcel André Hoffmann

Modeling, Simulation and Optimization of Supply Chain,

Production and Logistics Operations
The Simulation for the Ideal Optimum Site Model in the Actual
Distance Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
Angela Yu Yen Chen and Yutaka Karasawa
Solving a Location-Allocation-Sizing Problem Using Differential
Evolution Algorithm: A Case Study of Agricultural Water
Resources in North-Eastern Thailand . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
Rerkchai Srivoramas, Ponglert Sangkaphet,
Chutchai Kaewta, Rapeepan Pitakaso, Kanchana Sethanan,
and Natthapong Nanthasamroeng

vii
viii Contents

Simulation-Based Application for Improving Carton Production

Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
Nara Samattapapong and Thiti Mhoraksa
Solving Multi-Echelon Location Allocation Problem Using
Modified Differential Evolution Algorithm: Case Study
of Agricultural Products Warehouse in Greater Mekong Subregion . . . . 145
Kiatisak Pranet, Ponglert Sangkaphet, Rapeepan Pitakaso,
Natthapong Nanthasamroeng, Thanatkij Srichok, Kanchana Sethanan,
and Peema Pornprasert
Mathematical Modelling for Optimizing Tourist Trip Design
with Considering Scoring on Arc Visits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
Ryo Geoffrey Widjaja, A. A. N. Pewira Redi, Parida Jewpanya,
Muhammad Asrol, and Nur Layli Rachmawati
Two-Echelon Vehicle Routing Problem for Agricultural Area
Mapping Operation with Drone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
A. A. N. Pewira Redi, Muhammad Reza Chandra Kusuma,
Bertha Maya Sopha, Anna Maria Sri Asih, and Rahmad Inca Liperda

Sustainable, Social and Legal Challenges in Supply Chain

Management and Logistics
Environmental and Economic Value Prediction of Waste Electrical
and Electronic Equipment Recycle Using Reverse Logistics: The
Case of China’s Waste Mobile Phones . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
Sung Woo Kang and Yu Quan
A Financial Feasibility Study of Zero Waste Manufacturing:
Opportunity and Value Added of Agricultural Products on the R9
Route . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
Supattraporn Saisomboon, Arunrat Sawettham, Sumalee Ngeoywijit,
and Monika Kosacka-Olejnik
The Impact of Tighter Regulations of Export Controls in Global
Logistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
Kuninori Suzuki, Yoji Murayama, and Yi Xing
Two Echelon Vehicle Routing Problem for Disaster Mapping
Process Using Cooperated Vehicles Assignment . . . . . . . . . . . . . . . . . . . . . . 233
A. A. N. Perwira Redi, Nandini N. Sekaringtyas, Handina B. Astiana,
Rahmat Inca Liperda, Anna Maria Sri Asih, and Bertha Maya Sopha
Contents ix

A Relief Aids Distribution Model for the Pre-cooked Meals: The

Case of the Probable Disasater During Fasting Periods in Padang
City, Indonesia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
Rahmad Inca Liperda, Nikorn Sirivongpaisal, Reinny Patrisina,
and Sirirat Suwatcharachaitiwong
The Impact of the Use of Intelligent Supply Chain Tools
on the Transport, Forwarding and Logistics Industry . . . . . . . . . . . . . . . . . 265
Sylwia Konecka and Anna Łupicka
Smart Solutions for Supply Chain
Management
Impact of Internet of Things on Food
Supply Chains

Elkafi Hassini , Mohamed Ben-Daya, and Zied Bahroun

1 Introduction

The food supply chain (FSC) is a complex network of farmers and food producers,
including makers of farm equipment and chemicals, agribusinesses and processors,
distributors and support industries, such as transportation and financial services,
retailers and consumers. It is a critical component of the global economy. For
example, one billion people, or one in three of all world employees, are employed
in agriculture, and employees in FSC businesses account for more than half of the
total employment in sub-Saharan Africa and South Asia (FAO 2012).
The practice of supply chain management in FSCs has its distinctive challenges.
While in a typical supply chain, such as in retail, we largely focus on reducing costs,
in a FSC other important dimensions must be considered such as minimizing food
waste and loss as well as meeting certain health and safety regulations at all phases
of the supply chain. IoT application in FSCs reduces the time between data capture
and decision making. This enables supply chains to react to changes in real-time and
reduces food waste. In addition, it allows for quick action in case of recalls or safety
concerns. Most of the IoT applications in the FSC focus on technological solutions
or deal with conceptual models. The reader is referred to Ben-Daya et al. (2021) for
a recent review of IoT applications in the FSC. Using IoT data for better decision
making and for dealing with various operations management aspects are issues that

E. Hassini (B)
DeGroote School of Business, McMaster University, Hamilton, Canada
e-mail: [email protected]
M. Ben-Daya · Z. Bahroun
Industrial Engineering, American University of Sharjah, Sharjah, United Arab Emirates
e-mail: [email protected]
Z. Bahroun
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 3

P. Golinska-Dawson et al. (eds.), Smart and Sustainable Supply Chain
and Logistics—Challenges, Methods and Best Practices, EcoProduction,
https://doi.org/10.1007/978-3-031-15412-6_1
4 E. Hassini et al.

are not addressed adequately in the literature (e.g., see Ben-Daya et al. 2019 and
Ivanov et al. 2021). Another important issue that needs further investigation is the
justification of IoT investment through adequate cost–benefit analysis. It is thus our
goal in this paper to develop models for assessing the efficiency and impact of IoT
technologies applications in FSCs.
We consider a fresh produce product with a quality that deteriorates over time.
The introduction of IoT sensors allows for the continuous monitoring of the product
quality parameters through the supply chain. Consumer demand is a function of
price and quality. We develop Stackelberg game models that allow us to quantify the
impact of the use of IoT at the retailer or distributor levels. Such models can be used
to justify the investment in IoT infrastructure in the FSC and if it is viable, it can help
decide in which echelon of the FSC the investment will have more impact on FSC
performance improvement. To the best of our knowledge, models that quantify the
impact of IoT on FSC management and that can be used to make the business case
of IoT investment are lacking in the literature.
The remainder of this paper is organized as follows. A literature review of relevant
research is discussed in the next section. In Sect. 3, we present the problem statement,
notation and assumptions. We then present two Stackelberg pricing models in Sect. 4
where one includes IoT and the other does not. The two models will be used in the
following two sections to study the impact of IoT. In Sect. 5, we provide a numerical
study to investigate the impact of IoT as well as the impact of IoT investment on profits
and supply chain coordination. In Sect. 6, we provide some concluding remarks and
directions for future research.

2 Literature Review

Ben-Daya et al. (2019) have reviewed the literature on IoT and supply chain manage-
ment and pointed out the need for the development of analytical models to assess
the impact of IoT on supply chain operations. They have also found that a promising
area for IoT applications is in FSCs. IoT can enhance product freshness and aid
in quality management. A more detailed review of the role of IoT in supply chain
quality management can be found in Ben-Daya et al. (2020). Here we would like to
focus on the following two related literature areas: (i) food waste inventory control
models and (ii) the role of pricing and quality in supply chain coordination. In each
area we focus on the literature that is most relevant to our research problem and
methodology.
(i) Food waste inventory control models: There is evidence that a retailer’s choices
for how to manage inventory and consumer choice behaviour are key drivers of food
waste at the retail level (Cicatiello et al. 2017). Thus, inventory models for produce
retail must account for their impact on food waste. Unfortunately, this has not been the
prevailing practice in the operations management literature, where food cost is often
lumped up with inventory costs (Ozbilge et al., 2021). Yavari and Geraeli (2019)
Impact of Internet of Things on Food Supply Chains 5

proposed a multi-objective model for designing a sustainable food supply chain

where one of the objectives is to minimize the environmental footprint. Janssen et al.
(2018) consider waste minimizing in a grocery retail store by explicitly incorporating
non-sale periods, such as when the store closes. Mallidis et al. (2020a, b) and Ozbilge
et al. (2021) have considered the use of donations to minimize food waste.
(ii) Role of pricing and quality in supply chain coordination: The literature on supply
chain coordination is vast. In this paper we investigate the impact of IoT investment,
using the leverage of pricing and quality, on reducing the double marginalization
effect between the retailer and the distributor. Ma et al. (2013) have considered a
two-stage supply chain where the retailer tries to sell, and the supplier invests in
quality. They find that the two-part tariff as well as the two-part tariff combined
with the supplier quality effort do not coordinate the supply chain. They offer a new
contract that can achieve coordination. In a similar way we show, and quantitatively
determine, the impact of coordination through two-part tariff and IoT quality invest-
ment. Taleizadeh et al. (2018) consider closed-loop supply chains of a manufacturer,
a retailer and a third party. They use Stackelberg game models to find optimal prices,
quality levels and sales and recycling collection efforts and consider channel structure
impacts on coordination.
There are several opportunities for studying the emerging field of IoT quality-
controlled supply chain operations. This study is one of such efforts and makes the
following contributions: (1) A quantitative assessment framework for IoT impact
of FSC operations; (2) We investigate how pricing can play a role in food waste
reduction with the aid of IoT technology; (3) Analysis of IoT investment thresholds
that allow for deciding on who should invest in IoT and how much they should invest;
and (4) An assessment of the role of IoT investment in enhancing coordination in
the supply chain.

3 Problem Statement and Assumptions

We use the following notation:

t Time
i Supply chain entity, i ∈ {1(producer), 2(distributor), 3(retailer)}
qi Quality level at supply chain echelon i
ti Product arrival time at supply chain echelon i
q(t) Quality at time t
μi Instantaneous quality deterioration rate at supply chain echelon i
μ31 Instantaneous quality deterioration rate at the Retailer without IoT
μ32 Instantaneous quality deterioration rate at the Retailer with IoT
qh Highest quality level beyond which a customer is indifferent to quality
th Time at which customer starts recognizing different quality levels
qc Critical quality level below which a customer considers the product
unsalable
6 E. Hassini et al.

tc Time beyond which a customer finds the product unsalable due to bad
quality
D( p, t) Demand rate at time t for price p
D0 Market size
α Demand sensitivity to price
β Demand sensitivity to quality
T Replenishment period, T = tc − t3
c Distributor’s cost
w1 (w2 ) Price charged by the distributor to the retailer without IoT (with IoT)
p1 ( p2 ) Unit selling price at the retailer without IoT (with IoT)
D1 , D2 Total demand with IoT
We assume that the producer starts with perfect item quality. The product quality
then deteriorates linearly at rate μi that change at times ti . To analyze the impact of
different quality levels, we assume that the product is at a premium quality when it
reaches the retailers’ premises at the time t3 and remains at that state until the time th .
During the time interval [t 1 , th ], the product quality cannot be visibly differentiated,
and it is referred to as apparent stability. After th , the product quality becomes regular
and the customer can visibly tell the product quality level, such as due to discoloration.
The product quality remains acceptable until time tc at which point all remaining
inventory is not sellable and becomes food waste. Thus, we experience two types
of quality: a visible quality that remains constant at the level qh and then linearly
decreases with the rate μ3 and an effective quality level that reflects the actual quality
level that would be measurable if we use IoT technology.

4 Models

We use a Stackelberg game to model the retailer pricing problem in the absence and
presence of IoT.

4.1 Without IoT

Similar to Bahroun et al. (2020), we use the following quality dependent demand
functions in the absence of IoT:
(
D0 − αp1 + βqh t ≤ (q3 − qh )/μ31
D( p1 , t) = (1)
D0 − αp1 + β(q3 − μ31 t) t > (q3 − qh )/μ31
( (
1 1 [ ]
D1 = − β (q3 − qh ) + (q3 − qc ) + (D0 − αp1 + βq3 )(q3 − qc ) (2)
2 2
μ31 2
Impact of Internet of Things on Food Supply Chains 7

Using a Stackelberg game theory model, the distributor acts as a leader and the
retailer as a follower. The retailer optimizes the retail price p1 to maximize his profit,
which is subsequently used by the supplier to optimize the wholesale price w1 to
maximize her profit. We first determine the retailer and distributor profit as
( (
( p1 − w1 ) 1 ) ]
π R1 = − β (q3 − qh )2 + (q3 − qc )2 + (D0 − αp1 + βq3 )(q3 − qc )
μ31 2
(3)

and
( (
(w1 − c) 1 [ ]
π D1 = − β (q3 − qh ) + (q3 − qc ) + (D0 − αp1 + βq3 )(q3 − qc )
2 2
μ31 2
(4)

Finally, the overall decentralized supply chain profit π S1 is found by combining

the individual profits of the retailer and the distributor given in (3) and (4), i.e.,
π S1 = π R1 + π D1 , thus
( (
( p1 − c) 1 [ ]
π S1 = − β (q3 − qh )2 + (q3 − qc )2 + (D0 − αp1 + βq3 )(q3 − qc )
μ31 2
(5)

Proposition 1 (Optimal Retailer and Distributor Prices without IoT) The distributor
optimal wholesale price and the retailer’s optimal price are given by
[ ]
−β (q3 − qh )2 + (q3 − qc )2 + 2(D0 + βq3 )(q3 − qc ) c
w1∗ = + (6)
4α(q3 − qc ) 2
[ ]
−3β (q3 − qh )2 + (q3 − qc )2 + 6(D0 + βq3 )(q3 − qc ) c
p1∗ = + . (7)
8α(q3 − qc ) 4

Proof Applying the first and second-order optimality conditions to the retailer’s
profit function given by (3) we find that
[ ]
−β (q3 − qh )2 + (q3 − qc )2 + 2(D0 + βq3 )(q3 − qc ) w1
p1 = + (8)
4α(q3 − qc ) 2

Substituting this value into the distributor’s profit given by (4), the distributor
finds the optimal value w1 that maximizes its profit function, which is given by (6).
Now Substituting (6) into (8) yields the optimal retailer’s price given by (7). □

The optimal profit expressions for the retailer and distributor are given by the
following proposition.
8 E. Hassini et al.

Proposition 2 (Optimal retailer and distributor profits without IoT) The optimal
profit expressions for the retailer, distributor and supply chain are given by:
[ [ ] ]2
∗ −β (q3 − qh )2 + (q3 − qc )2 + 2(D0 + βq3 − αc)(q3 − qc )
π R1 = (9)
64αμ31 (q3 − qc )
[ [ ] ]2
∗ −β (q3 − qh )2 + (q3 − qc )2 + 2(D0 + βq3 − αc)(q3 − qc )
π D1 = (10)
32αμ31 (q3 − qc )
[ [ ] ]2
∗ 3 −β (q3 − qh )2 + (q3 − qc )2 + 2(D0 + βq3 − αc)(q3 − qc )
π S1 = (11)
64αμ31 (q3 − qc )

Proof The optimal profit expressions are obtained by substituting p1∗ and w1∗ in the
profit functions given by (3) and (4), respectively. The relationship between them is
due to the linear nature of the demand function. □

4.2 With IoT

When the supply chain players invest in IoT infrastructure to monitor product quality,
the quality level can be always detected even when visible changes do not occur. As a
result, the quality is assumed to be a linearly decreasing function of time throughout
the demand period at a constant unit retail price p2 as mentioned previously.
With IoT data, the quality level is known at all points of time and the demand
function at the retailer is given by

D( p1 , t) = D0 − αp1 + β(q3 − μ31 t)

and the total demand is

[ ]
1 1
D2 = − β(q3 − qc )2 + (D0 − αp2 + βq3 )(q3 − qc ) .
μ32 2

The derivation of the model in this case is like that of the previous section, except
that the demand function is given by (2) and the retailer incurs an additional fixed
cost per cycle, I r , that corresponds to the investment in IoT infrastructure.
The profit functions of the retailer, distributor, and supply chain are given by
( (
( p2 − w2 ) 1
π R2 = − β(q3 − qc ) + (D0 − αp2 + βq3 )(q3 − qc ) − Ir
2
(12)
μ32 2
( (
(w2 − c) 1
π D2 = − β(q3 − qc )2 + (D0 − αp2 + βq3 )(q3 − qc ) (13)
μ32 2
Impact of Internet of Things on Food Supply Chains 9

( (
( p1 − c) 1
π S2 = − β(q3 − qc ) + (D0 − αp2 + βq3 )(q3 − qc ) − Ir
2
(14)
μ32 2

Following the same approach described in the previous section, we present the
following propositions for optimal prices and profits.

Proposition 3 (Optimal Retailer and Distributor Prices with IoT) The optimal prices
that maximize the retailer’s and distributor profits are

6D0 + 3β(q3 + qc ) c
p2∗ = + (15)
8α 4
2D0 + β(q3 + qc ) c
w2∗ = + . (16)
4α 2
Proposition 4 (Optimal retailer and distributor profits with IoT) The optimal profit
expressions for the retailer, distributor and supply chain are given by:

∗ (q3 − qc )[−β(q3 − qc ) + 2(D0 + βq3 − αc)]2

π R2 = − IR
64αμ32

∗ (q3 − qc )[−β(q3 − qc ) + 2(D0 + βq3 − αc)]2

π D2 =
32αμ32

∗ 3(q3 − qc )[−β(q3 − qc ) + 2(D0 + βq3 − αc)]2

π S2 = − IR.
64αμ32

The proofs of the above propositions are like those of Proposition 1–2.

5 IoT impact on Reducing Food Waste

We numerically study the impact of IoT investment on food waste reduction. Inspired
from the parameters used in (Chen et al. 2019), in Table 1, we show the values for the
initial deterioration rate without IoT and with IoT, the initial quality at the retailer,
the potential market size, the price sensitivity, the quality sensitivity, and the unit
product cost.
In Fig. 1, we show how demand changes depending on where IoT has been
deployed. IoT impact on food waste is reflected through changes in demand after

Table 1 Initial parameter assignment

μ31 (/hr) μ22 (/hr) q3 qh qc D0 (units/hr) α β c ($/unit)
0.007 0.004 0.95 0.60 0.30 15 1.8 1.8 5
10 E. Hassini et al.

an improvement in the deterioration rate μ32 when it is implemented at the retailer

level only (Fig. 1a), an improvement in retailer initial quality q3 when implemented
at the distributor level only (Fig. 1b), or an improvement in both deterioration rate
and initial retailer quality if IoT is implemented at both the distributor and retailer.
We note that in all cases the demand increases and therefore the food loss
decreases. The behaviour of how demand changes are similar to those of the changes
in profits. The benefit of IoT deployment has diminishing marginal benefits. i.e.,
there is a point beyond which additional investments in IoT do not lead to as much
food waste reduction as the prior investment. Besides, based on the parameters used,
and if only one party will invest in IoT, it should be the retailer as deploying the
technology at the retailers can save more food waste as seen in Fig. 1a, b. We can
see from Fig. 1a that the demand follows an iso-elastic pattern with respect to the
deterioration rate when only the retailer invests in IoT. When only the distributor
invests in IoT, we see in Fig. 1b that the demand is an increasing function of the
initial quality. However, the demand does not vary significantly as it would if the
retailer is the only investor in IoT. The demand increases the most, and so will the

Demand Demand
1800 450
1600 400
1400 350
1200 300
Units

1000
Units

250
800 200
600 150
400 100
200 50
0
0
0.001
0.004
0.006
0.009
0.011
0.014
0.016
0.019
0.021
0.024
0.026

0.86
0.80
0.81
0.82
0.84
0.85

0.87
0.88
0.90
0.91
0.92
0.93
0.94
0.96

µ32 q3
(a) Retailer only. (b) Distributor only.

Demand
1600
1400
1200
1000
Units

800
600
400
200
0
1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55
(q3,µ32)
(c) Retailer and distributor.

Fig. 1 Impact of IoT deployment on food loss reduction

Impact of Internet of Things on Food Supply Chains 11

reduction in food waste, when both the retailer and distributor invest in IoT as shown
in Fig. 1c.

6 Conclusions

We have presented models to assess the impact of IoT on coordination and waste
reduction in food supply chains. We provided analytical and numerical results. Our
analysis reveals that IoT can coordinate the supply chain. We also show how IoT can
lead to a reduction in food waste.
We believe that the models we presented in this paper will serve as a good starting
point for investigating the impact of IoT on supply chains. As such it is natural that it
will have some limitations and there are several ideas for extending our work. First, an
extension of the discounting model may include competition where two retailers may
invest in IoT and compete on prices and quality. Second, while we have modelled the
impact of IoT investment implicitly through the deterioration rate, it will be useful
to consider a direct approach where one would develop an investment function and
find optimal investment levels. Third, in our model we used a deterministic demand
function and it would be worthwhile investigating the impacts of demand uncertainty
on the value of investing in IoT technology.

References

Bahroun Z, Ben-Daya M, Hassini E (2020) IoT Quality-controlled demand pricing model in food
supply chains. In: Golinska-Dawson P, Tsai KM, Kosacka-Olejnik M (eds) Smart and sustain-
able supply chain and logistics—trends, challenges, methods and best practices. EcoProduction
(Environmental issues in logistics and manufacturing). Springer, Cham. https://doi.org/10.1007/
978-3-030-61947-3_26
Ben-Daya M, Hassini E, Bahroun Z (2019) Internet of things and supply chain management: a
literature review. Int J Prod Res 57(15–16):4719–4742
Ben-Daya M, Hassini E, Bahroun Z, Banimfreg B (2021) The role of internet of things in food
supply chain quality management: a review. Qual Manag J 28(1):17–40
Cicatiello C, Franco S, Pancino B, Blasi E, Falasconi L (2017) The dark side of retail food waste:
evidences from in-store data. Resour Conserv Recycl 125:273–281
Chen X, Wu S, Wang X, Li D (2019) Optimal pricing strategy for the perishable food supply chain.
Int J Prod Res 57(9):2755–2768
FAO (2012) Labour in FAO statistical yearbook 2012. https://bit.ly/3a1i5HZ
Ivanov D, Tang CS, Dolgui A, Battini D, Das A (2021) Researchers’ perspectives on industry
4.0: multi-disciplinary analysis and opportunities for operations management. Int J Prod Res
59(7):2055–2078
Janssen L, Sauer J, Claus T, Nehls U (2018) Development and simulation analysis of a new perishable
inventory model with a closing days constraint under non-stationary stochastic demand. Comput
Ind Eng 118:9–22
Ma P, Wang H, Shang J (2013) Contract design for two-stage supply chain coordination: integrating
manufacturer-quality and retailer-marketing efforts. Int J Prod Econ 146(2):745–755
12 E. Hassini et al.

Mallidis I, Sariannidis N, Vlachos D, Yakavenka V, Aifadopoulou G, Zopounidis K (2020a) Optimal

inventory control policies for avoiding food waste. Oper Res 22: 685–701
Mallidis I, Vlachos D, Yakavenka V, Eleni Z (2020b) Development of a single period inventory
planning model for perishable product redistribution. Ann Oper Res 294:697–713
Ozbilge A, Hassini E, Parlar M (2021) Donate more to earn more. Submitted for publication,
DeGroote School of Business, McMaster University
Taleizadeh AA, Moshtagh MS, Moon I (2018) Pricing, product quality, and collection optimization
in a decentralized closed-loop supply chain with different channel structures: game theoretical
approach. J Clean Prod 189:406–431
Yavari M, Geraeli M (2019) Heuristic method for robust optimization model for green closed-loop
supply chain network design of perishable goods. J Clean Prod 226:282–305
An Agent-Based Approach to Evaluate
Freight Consolidation Center Strategy
for Last Mile Deliveries in Jaipur City,
India

Pankaj Kant and Sanjay Gupta

1 Introduction

The negative impacts of urban goods distribution for social, economic and externali-
ties is an area of key concern in sustainable development of cities. The sustainability
of urban goods transport is the key policy objective of various countries (OECD
2003). In recent decades there is a thrust in improving passenger transport in Indian
cities but urban freight is still a neglected area in national and local policy discourse.
Even the master plan development (MPD) exercise and comprehensive mobility plan
(CMP) exercise by local development authorities of Indian cities do not address the
critical concerns of urban freight issues and freight stakeholders (MoUD 2014).
Urban freight strategies are often selected without comprehensive information of
their potential benefits in transport management plans (T.M.P) (DoT 2012). Sustain-
able transportation is essential for the overall sustainable development of cities due
to its contribution to externalities, economic and social issues (Gudmundsson et al.
2015). Urban freight strategies and key performance indicators are helpful to local
policy makers to reduce the negative impacts of urban goods distribution in context
of sustainable development objectives (Litman 2019; Patier and Routhier 2008).
Replication of freight stakeholders behaviours and attributes are essential and key to
sustainability in city logistics domain (Dablanc et al. 2013).
Awareness regarding freight issues and evaluation for freight strategies needs to
be addressed among all relevant urban freight stakeholders (Martin et al. 2012). Its
tedious and time-consuming task to predict the outcomes of urban freight policies

P. Kant (B) · S. Gupta

Department of Transport Planning, School of Planning and Architecture, 4-Block-B Indraprastha
Estate, New Delhi 110002, India
e-mail: [email protected]
S. Gupta
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 13

P. Golinska-Dawson et al. (eds.), Smart and Sustainable Supply Chain
and Logistics—Challenges, Methods and Best Practices, EcoProduction,
https://doi.org/10.1007/978-3-031-15412-6_2
14 P. Kant and S. Gupta

including financial implications (Browne and Allen 2011). The effect of urban freight
transport can be quantified under various categories, and each category requires a
different kind of freight strategy (Lindholm 2010). A research study concluded Seven
groups of sustainable policy measures to mitigate the negative impact of urban freight
and barriers in implementing freight strategies for the sustainability of urban freight
transport (BESTUFS II 2006).
International best practices in sustainable urban freight strategies are available for
local policy makers and other urban freight stakeholders (CIVITAS 2015). Divergent
objectives of shippers and transporters act as a barrier in the implementation of
urban freight strategies (Van Duin et al. 2018). A suitable framework is essential for
selecting the freight strategies based on the requirements, behaviour and attributes
of freight stakeholders for sustainability in urban freight sector (Sharma et al. 2017).
City logistics is a difficult domain for local policymakers due to of lack informa-
tion and knowledge about urban freight. Among all available city logistics solutions,
Urban Consolidation Centre as a logistics space strategy is an appreciated one (Faure
et al. 2016). Urban consolidation centres (U.C.C.s) have a major role in solutions to
reduce the impact of urban freight transport (Quak 2014). An urban consolidation
centre (U.C.C.) is a logistics space strategy used in the periphery of cities for trans-
shipment, load consolidation and selection of carriers for last-mile delivery (Browne
et al. 2005).
The urban freight planning related aspects in the developing country context are
not amply researched upon. There is almost negligible literature available on the
freight stakeholder behaviour analysis and its relation with urban goods distribution
strategies like logistics spaces and consolidation centre strategy in the Indian context.
Inputs from wholesalers, retailers, and transport operators in the modelling frame-
work are missing in policy discourse by local policymakers in Indian cities to arrive
at acceptable solutions as city logistics measures.
The next section of research presents a literature review related to consolida-
tion centre strategy in urban goods distribution. Section 3: research methodology.
Section 4: case city profile and commodity selection. Section 5: descriptive view of
data set. Section 6: model validation. Section 7: Evaluation of urban consolidation
centre strategy and Sect. 8 presents the conclusions and policy implications.

2 Literature Review

An urban consolidation centre (U.C.C.) can mitigate several urban freight issues like
emissions, congestion, parking, road safety of inner roads and business deliveries.
The success of the U.C.C. has several barriers like financial, social, cultural, and legal
barriers. The most important related to U.C.C. is unclear business terms and low
acceptability among private stakeholders. Acceptability for all freight stakeholders
is key to the success of U.C.C. (Elvsaas 2020). The location of U.C.C. has been
analyzed by the centre-of-gravity method to optimize freight deliveries in the seaside
tourist resorts of the West Pomeranian Region of Poland. The research has applied
An Agent-Based Approach to Evaluate Freight Consolidation Center … 15

the interview method for the analysis of the tourist traffic impact on deliveries in the
selected area. The research study was concentrated on the aspects of transport issues
and organizational assumptions. Two different optimum locations were proposed for
summertime and after summertime (Kijewska et al. 2012).
A research study assesses the potential role of clean vehicles in association with
U.C.C. to reduce the freight traffic and environmental impact in urban areas of
London. A pilot was conducted for stationery and office supplies company to use elec-
tric tricycle for deliveries to customers from a micro-consolidation centre located in
the delivery area. The results show the reduction in CO2 emissions per parcel delivery
by 14% and 55% respectively as a result of this delivery system. The pilot trial was
successful in the context of transport, environmental and financial terms (Leonardi
et al. 2012). Challenges and factors affecting the implementation of U.C.C. in Hague
have been analyzed for the local municipality. Two factors that affect the implemen-
tation of U.C.C. were the allocation of the costs and benefits and the willingness of
transport companies to cooperate. The municipality had a major role in bringing the
costs and benefits together. The scientific theory could provide more accurate advice
to local policymakers, but evaluations of U.C.C.s are often poorly documented (van
Duin and Quak 2010).
A UCC is viable under some given conditions like the number of delivery parcels,
the approach distance, and so the urban sprawl. The morphology of the city also
influences the performance of U.C.C.s. Productivity of U.C.C. shall address fixed
costs rather than variable costs. Location of U.C.C. and the number of U.C.C. required
is needs to be evaluated before implementation. Finally, it is certainly useful to study
the influence of demand and location on the viability of U.C.C (Faure et al. 2016).
Utilizing U.C.C. gives a chance to mitigate the negative effects of freight trans-
port in urban areas. Research analysis of urban consolidation centres development
in the Westpomeranian Region of Poland suggests that U.C.C. is a good solution
enabling more effective functioning of tourist, recreational, health resort services and
improvement in the quality of the services due to reduction in the noise and pollution
in the tourist zones. Implementation of a U.C.C. requires adjusting regulations of
entrepreneurs, wholesalers and distributors (Chwesiuk et al. 2010).
Results of Binnenstadservice.nl (B.S.S.) U.C.C. show positive results after its first
year of service in Nijmegen. Due to the B.S.S. consolidation centre, the number of
trucks and number of truck kilometres in the city centre decreased. The effects on
inconvenience for residents, traffic safety and shopping environment are promising.
The effects on local air quality are limited due to passenger and bus traffic in addition
to freight traffic in the city. The fiscal deficit will be there in the second year of oper-
ation without local municipality support. The positive results of B.S.S. in Nijmegen
give rise to B.S.S. franchise initiatives in other Dutch cities (van Rooijena and Quak
2010).
A framework for the classification of U.C.C. and initiatives was developed based
on a survey of already adopted alternatives. The main cause for the inefficiencies of
the U.C.C. is the lack of direct consultation between carriers and receivers in urban
areas. Urban consolidation was never treated as a stand-alone concept. Finally, there
16 P. Kant and S. Gupta

is a need for the classification of consolidation-oriented measures and initiatives like

physical and behavioural concepts (Verlindea et al. 2012).
The results of the U.C.C. of Belo Horizonte city show a significant improvement
to all actors involved in the process, such as society, carriers and retailers. There was a
reduction in the number of vehicles and a reduction of pollutant emission in the urban
environment. In addition, the adoption of green vehicles can further improve environ-
mental parameters (de Assis Correia et al. 2012). Research results from Sweden and
Scotland suggest that freight policies in conjunction with stakeholders collaboration
are key to support public-led urban consolidation centre (U.C.C.) developments.
The key finding reveals that urban freight policies such as time window restric-
tions can support successful U.C.C.s; they cannot be considered in isolation by the
local authority. A successful development of U.C.C. also requires a commitment to
financially support for the medium-term (Zehra Akgün et al. 2019).
An agent-based simulation (ABMS) framework with receivers and carriers was
applied to the city of Copenhagen for sustainable business models and the supporting
role of administrative policies for the feasibility of U.C.C. Research results show
that environmental improvement may be achieved through the use of a U.C.C. by
reduction of trucks in urban areas by up to 60% and reducing emissions by about
70%. However, it is challenging to find schemes that are also financially sustainable.
There is a need for committing carriers to use the U.C.C. for its success. The success
of the U.C.C. is not possible without supporting measures and temporary subsidies
to the carriers (van Heeswijk et al. 2019).
An agent-based modelling framework was developed for multiple stakeholders
for the city logistics measure (U.C.C.). The model integrates multiple stakeholders,
which can be used to analyze the city logistics solutions measures satisfying domain
stakeholders. The modelling framework used a semantic model (ontology) and
its validation by the participatory simulation game. The model uses city logistics
ontology as a fundamental building block with events and additional behavioural
attributes to capture real-world behaviour (Anand et al. 2019).
The literature review suggests that there is insignificant research on the appli-
cation of agent-based modelling and simulation in the decision-making process of
different stakeholders in a developing environment like India. Moreover, no attempt
has been reported in demonstrating the use of ABMS for evaluating urban consoli-
dation centre strategy in India. The present study provides the much-needed insight
into the freight stakeholder behaviour towards scientifically arriving at sustainable
urban goods distribution strategies like consolidation centre strategy using an agent-
based modelling system (ABMS) modelling approach in a developing environment
of India with the case city of Jaipur.
An Agent-Based Approach to Evaluate Freight Consolidation Center … 17

3 Methodology and Data Collection

3.1 Research Methodology

The basic fundamental block of ABMS is the state action charts of its agents (stake-
holders). The stakeholder’s interaction model uses agent-based modelling and simu-
lation (ABMS) framework to evaluate the consolidation centre strategy in the case
city with the help of KPIs. The proposed ABMS modelling framework for agents
interaction is shown in Fig. 1. The consolidation centre strategy is tested once the
ABMS model is calibrated and validated. The primary agents for this research study
are administrators, wholesalers, retailers, carriers and loader/unloaders.
The proposed framework has two types of movement involved, one is the flow of
information, and the other is the flow of actual goods movement. The information
flow and sequence are depicted by the solid line, and the sequence of goods movement
is shown by the dotted line in the proposed ABMS framework. The retailers place the
weekly orders to the wholesaler via the message/call function. Wholesaler process
the order received from retailer and call/message to transport carrier available at the
parking area in the near vicinity of wholesaler. A carrier moves to the godown of
the respective wholesaler after receiving the message from the wholesaler. Loaders
start loading the carrier once it reaches godown. After loading, the carrier moves
to the respective retailer in the city. Unloading of goods starts once the carrier is
reached to the retailer. The carrier again moves back to the initial parking lot after
the goods are delivered to the retailer and wait for his next consignment from the
wholesaler. Transport carriers use the shortest routes option for route selection in
delivering goods from wholesalers to retailers in existing time window scenarios

Fig. 1 Conceptual ABMS modelling framework for agents (Source Author)

18 P. Kant and S. Gupta

Table 1 State action transition and functions for agents

Agents State Transition type or Function
triggered by
Retailers Frequency of ordering Rate Uniform
goods (weekly)
Receiving goods Message/call
Working hours Time out Boolean
(10 a.m.–9 p.m.)
Monday to Saturday
Wholesalers Shipping goods as per Message/call Function of retailer
(distributor) received orders orders
Working hours Timeout Boolean
(10 a.m.–6 p.m.)
Monday to Saturday
Trucks/carriers Got order from distributer Msg. from the
wholesaler
Loading (Hours) Rate Triangular
Going to client (retailer) Agent arrival MSG/call
Unloading Rate Triangular
Going back to distributor Agent arrival M.S.G./call
Route restriction Time out
(6 p.m.–8 p.m.)
Source Author

and restrictions implemented in case-city by local administrators. The parking lot

for the transport carrier and warehouses are situated in the wholesale market areas
across commodities. The behaviour and interaction rules of identified agents for this
research study are shown in Table 1.
Modelling assumptions adopted in the proposed ABMS research framework are
the following.
. Capacitated vehicle routing problem with a time window (CVRPTW) is assumed
for this research framework.
. All freight vehicles are initially assumed to be located at wholesale markets.
. The location of retailers for both wholesale markets is assumed to be located at the
centroid of residential zones (total 92 zones) as per the comprehensive mobility
plan of the case city. Model results for (92) retailers are expanded for the full
population of building hardware and electronics goods retailers in Jaipur city.
. Tour routing and planning by transport carriers are not considered in the modelling.
. The model only captures the urban goods distribution from wholesalers to retailers
as last-mile delivery.
. The ABMS model only simulates the transport model, not the total logistics cost
model.
An Agent-Based Approach to Evaluate Freight Consolidation Center … 19

. The simulation time for the ABMS model is 90 days (12 weeks). Unit time setting
for simulation experiment is per day.
. The ABMS model was validated by G.E.H. statistics with model counts and counts
observed from cordon points traffic volume around the wholesale market.
AnyLogic 8.4 java-based student version software is used for ABMS modelling.

3.2 Data Collection

The primary establishment surveys were conducted in the case wholesale markets
and retailer’s premises. Face to face pen and pencil survey method was used for
the establishment surveys to collect the information from wholesalers, retailers and
origin-destinations surveys from transport operators. Manual traffic count of freight
vehicles was conducted at market cordon points of wholesale markets to capture
the total flow of incoming and outgoing goods vehicles. Transport carrier surveys
were conducted in the wholesale market during the loading of goods, and transport
operators survey at their offices in transport Nagar in Jaipur city. Table 2 shows the
sample size collected by various sampling methods for this research study.

4 Case City Profile and Consolidation Space Selection

Jaipur city which is the capital city of the Rajasthan state in India is famous for its
tourist attraction, stone work, textile, jewellery and its local architecture. Jaipur is
also known as pink city in India. Jaipur city is well connected with rest of Indian
Cities with road, rail and air transport. The total population of Jaipur city is 31
million (yr. 2011) and its total area is 2940 km2 . The primary land use of the Jaipur
city is residential (45%). The commercial land use constitutes 6.7% and circulation
is 16%. (Anand et al. 2016). There are around 12 wholesale markets in the Jaipur city
involved in intra-city and inter-city goods distribution. Some of these of wholesale

Table 2 Data collection

Establishment survey Establishment survey Truck driver survey
(wholesalers) (retailers)
Sampling method Stratified Systematic Stratified
Confidence level (%) 95 98 98
Population size 510 5100 Unknown
Sample size required 109 189 392
Sample size collected 110 200 550
Source Primary survey
20 P. Kant and S. Gupta

markets are situated in walled city of the city like textiles, furniture, iron market,
food grain, chemical markets and electronics markets. Dairy and meat products and
vegetable wholesale market on the periphery of Jaipur city. Building hardware &
timber market is situated centre of city near Gopalpura bypass and stone market at
Ajmer bypass. Fertilizer and chemical market is situated at Durgapura area in Jaipur
city (Anand et al. 2016).
Building Hardware market (BH) and Electronics market (EM) were selected for
the evaluation of freight consolidation centre strategy in Jaipur city. Goods distri-
bution in the building hardware market is weigh based, whereas it is number and
item-based in the Electronics market. Electronics market is situated in old city area
and there is no scope of expansion for wholesalers due to rising demand in case city.
Building hardware market is developed by local development authority to cater the
need of construction industries in fast developing of Jaipur city and satellite towns.
Five consolidation spaces are identified for the relocation of the building hardware
(Aatish) market (B.H.) and electronics (Jayanti) market (E.M.) as recommended
by the comprehensive mobility plan of Jaipur city (Jaipur Development Authority
2019). Figure 2 shows the location of the new consolidation centre for wholesale
market relocation. All these new locations are on the periphery of the city along with
incoming highways in Jaipur city. These new locations are tested for the relocation
of wholesale markets without changing the retailer’s location in ABMS modelling
framework for Jaipur city.
Five major highways connect Jaipur city for receiving goods by wholesalers from
respective manufacturers. Kukas (1) area is located on the periphery of Jaipur city
connected to Delhi city by national highway (NH) 48. Similarly, Kanuata (2) is
located on NH 21 connected to Agra city, Chaksu (3) on NH 52 connected to Kota city,
SEZ (4) on NH48 connected to Ahemdabad city and Harmada (5) on NH48/NH21
connected to Bikaner city at the periphery of Jaipur city.

Fig. 2 Location of
consolidation centres
An Agent-Based Approach to Evaluate Freight Consolidation Center … 21

5 Descriptive Statistics of Data Set

Descriptive statistics of wholesalers attributes is shown in Table 3. The mean value of

the shop area in the electronics market is higher than the building hardware market.
Employment per 100 m2 in the electronics market is higher than the building hardware
market. Total tonnage handled, including incoming and outgoing tonnage per 100
m2 in the building hardware market, is almost double of the electronics market. Total
tonnage trips frequency, including incoming and outgoing trips per 100 m2 in the
building hardware market, is marginally lesser than the electronics market. Incoming
freight trips in the building hardware market are almost double the electronics market.
Outgoing trips in the electronics market are higher than building hardware market.
Descriptive statistics for retailers of building hardware and electronics goods
retailers are presented in Table 4. The median value of employment is two people in
most of the retail shops in the building hardware market, similar to electronics market
retailers. The number of trips attracted (incoming trips frequency) to building hard-
ware retailers has almost half compared to electronics goods retailers. The weekly
tonnage attracted to building hardware retailers is much higher than the electronics
goods retailers.
Descriptive statistics of transport operator attributes are shown in Table 5. There
is a significant difference in haulage time in the case of L.C.V. and 4w commercial
vehicles mode in both markets, suggesting a significant difference in distribution
leads. Loading time is slightly higher than unloading for all modes in both markets.
Idle time is almost similar in both markets for 3 W and 4 W, whereas, in the case of
L.C.V., idle time is higher in the building hardware market compared to the electronics
market. There is no significant difference in loading unloading cost in both markets
for respective freight modes.

Table 3 Descriptive statistics of wholesalers

Indicators Unit Electronics market Building H. market
Mean Median SD Mean Median SD
Shop area m2 140.2 149.5 67.7 89.1 80 62.5
Employment 100/m2 5 4 2.1 3.1 3 1.5
Incoming Weekly 4.1 4 1.4 9 9 2.7
frequency
Incoming Weekly 14.1 15 3.04 27.6 23 15.5
tonnage
Outgoing trips Weekly 22.1 23 4.1 14.9 15 3.2
frequency
Outgoing Weekly 11.8 12 2.04 24.8 20 13.8
tonnage
Average nos. of Per wholesaler 5.5 8.8
retailers
Source Author
22 P. Kant and S. Gupta

Table 4 Descriptive statistics of retailers

Indicators Unit Electronics Building hardware
Mean Median SD Mean Median SD
Shop area m2 77.19 70 59.7 94.5 83.6 66
Employment 100/m2 2.85 2 1.64 3.07 2 1.9
Incoming trips Trips/week 4.16 4 1.51 2.48 2 1.3
Incoming tonnage Tons/week 5.22 5 3.42 8.99 7.5 2.8
Source Author

Table 5 Descriptive statistics of transport operators

Mode Commodity Haulage time Loading time Unloading time Idle time (Hr)
(Hr) (Hr) (Hr)
Mean SD Mean SD Mean SD Mean SD
3W EM 0.7 0.2 0.3 0.05 0.27 0.06 0.95 0.15
BH 0.7 0.4 0.6 0.20 0.51 0.07 0.9 0.17
4W EM 0.7 0.2 0.4 0.08 0.34 0.09 1.15 0.16
BH 0.8 0.3 0.6 0.17 0.49 0.15 1.13 0.20
LCV EM 0.9 0.4 0.7 0.07 0.66 0.07 1.4 0.14
BH 1.4 0.6 0.7 0.12 0.69 0.09 1.85 0.22
Source Authorb

5.1 Cordon Point Traffic Volume Counts (T.V.C.)

Traffic volume counts of motorized freight vehicles were observed at the cordon point
of the building hardware market for nine hours. A total of 1256 freight vehicles were
observed, out of which there were 313 3w commercial vehicles, 388 4w commercial
vehicles, and 555 LCV, respectively observed in the T.V.C. survey. In the case of
the electronics market, a total of 1030 freight vehicles are observed in the case of
the building hardware goods distribution. There were 345 (3w) commercial vehicles,
362 (4w) commercial vehicles, and 323 LCV observed in the T.V.C. survey.

6 ABMS Model Validation Results

Table 6 shows the ABMS modelling and validation results for electronics and building
hardware goods distribution in Jaipur city. A total of 1318 freight trips for all retailers
were predicted by the ABMS model in Jaipur city. Total 1256 freight trips counts
are observed from the cordon survey of building hardware wholesale market by
commercial vehicles (3w, 4w and L.C.V.). In electronics goods distribution, the
ABMS model predicts 940 freight trips for all retailers in Jaipur city. There is a
An Agent-Based Approach to Evaluate Freight Consolidation Center … 23

Table 6 ABMS model results and validation

Building hardware Electronics
Trips/Deliveries Counts Counts
Nos. of deliveries/day extrapolated for all retailers (M) 1318 940
Trips counts/day from the primary survey (C) 1256 1030
Trips variation (%) 4.9 −8.7
GEH 2.4 4.5
Source Author

variation of 4.9% from modal count (M) and observed count (C) in building hardware
distribution. The model is over predicting the total freight deliveries by 4.9%. In the
case of electronics goods distribution, there is a variation of -8.7%, and the model is
under predicting the total freight deliveries by -8.7%. The G.E.H. statistic for model
validation has a value of 2.4 in the case of building hardware distribution and 4.5 in
electronics goods distribution. G.E.H. value less than 5 is acceptable for validation.

7 Evaluation of Consolidation Centre Locations

Table 7 shows the impact of new consolidation centres (strategy) with business
as usual (B.A.U.) scenario evaluated for building hardware (B.H.) and electronics
goods (E.M.) distribution in the case city. In the case of building hardware goods
distribution, the Kanauta location is best suitable in terms of total V.K.T., which has
observed a 34% increase in V.K.T. compared to the B.A.U. scenario and Chaksu is
the least preferable location where only an 8% increase in V.K.T. has been observed.
Kanuata is the best preferable location, followed by Kukus location for building
hardware distribution for wholesalers due to increased business deliveries. It is also
preferable for transport operators due increase in travel time and fleet efficiency
to maximize their profit. An increase in V.K.T. for the Kanuata location will have
a positive impact on emissions which is not desirable from an emission point of
view. Chaksu location is most preferred in case emission is concerned for local
policymakers due to minimum V.K.T. by transporters involved in building hardware
distribution.
In the case of electronics goods distribution, it can be observed Kukas followed
by Chaksu are the best preferable locations for t for transport operators to maximize
their efficiency. In the case of building hardware, distribution transport operators
can maximize the economic efficiency by an increase in V.K.T. and fleet usage, but
wholesalers have a reduction in the number of business deliveries compared to the
B.A.U. scenario and all five locations are not desirable locations for wholesalers
of electronics goods distribution. Exiting location of wholesalers in Jaipur city is
more desirable for wholesalers compared to new locations in Jaipur city. Chaksu (3)
 24

Table 7 Impact of logistics space strategy on urban goods distribution (KPI per week)
Location Commodity VKT % Chg (%) TKT % Chg (%) T.T. (Hrs) % Chg (%) DLV (nos) % Chg (%) Fleet usage (%) % Chg (%)
Atish Mkt BAU (BH) 1737 4623 58 607 58
Jayanti Mkt BAU (EM) 1134 1343 57 439 49
Kukas BH 2268 31 3221 31 76 30 727 20 70 12
Kanauta BH 2334 34 3314 35 77 33 740 22 72 14
Chaksu BH 1878 8 2667 8 62 8 715 18 61 3
Sez BH 2178 25 3093 26 73 25 721 19 68 10
Nh52/48 BH 2208 27 3135 27 74 27 718 18 68 10
Kukas EM 1464 29 1734 29 73 28 408 −7 60 11
Kanauta EM 1266 9 1500 12 64 12 403 −5 53 4
Chaksu EM 1446 25 1620 21 68 20 412 −4 57 8
Sez EM 1446 22 1710 27 72 26 407 −4 60 11
Nh52/21 EM 1200 5 1422 6 60 5 401 −5 51 2%
Source Author
P. Kant and S. Gupta
An Agent-Based Approach to Evaluate Freight Consolidation Center … 25

location is best suited for wholesalers for maximum business deliveries. NH52/48(5)
location is best suited if emissions are the priority of policy planners due to the least
V.K.T. by freight carriers. The study results confirm a trade-off between stakeholders
financial benefits and externalities associated with freight transport. It can also be
observed from the results that a location for consolidation centres strategy has a
different impact on key performance indicators across commodity distribution.

8 Conclusion

Urban freight strategies are vital decisions in the sustainability of urban goods distri-
bution. This paper investigated the impacts of consolidation centre strategy on freight
stakeholders involved in urban goods distribution across two commodities in Jaipur
city. The impacts of consolidation centre space strategy are evaluated in the ABMS
modelling framework with the help of key performance indicators of urban goods
distribution.
The research results indicate that consolidation centre space strategy impacts
vary across commodity distribution in urban areas and a single location for multiple
commodities is not desirable across commodity distribution for heterogeneous
freight stakeholders. Research results also show that there is a trade-off between
the economic efficiency of urban freight stakeholders and externalities associated
with freight transport in the selection of the location of consolidation centre spaces.
Proposed research study framework can be extended for other urban commodities
to develop the prepare freight mobility plan to assess the impacts on urban freight
stakeholders. Due diligence is required in the evaluation of the location for consol-
idation centre space strategy according to the commodity or group of commodities
in the master planning exercise in the context of Indian cities.
Wide spectrum of commodities for the evaluation of the consolidation centre
spaces in urban areas needs to be evaluated in case of Indian cities. The logistics
cost model needs to be integrated into the modelling framework for the evaluation of
monetary impacts on freight stakeholders. A joint model for shipment size and mode
choice could be explored in the ABMS modelling framework for the evaluation of
freight strategies.

Acknowledgements AnyLogic company permitted to use their AnyLogic 8.4 educational version
for research purposes in sustainable urban goods distribution project to the School of Planning and
Architecture New Delhi India.
26 P. Kant and S. Gupta

References

Anand N, Van Duin R, Tavasszy L (2016) Framework for modelling multi-stakeholder city logistics
domain using the agent-based modelling approach. Transp Res Proc 16:4–15
BESTUFS II (2006) Best urban freight solutions II. Retrieved March 21, 2018,
from http://www.bestufs.net/download/BESTUFS_II/key_issuesII/BESTUFS_II_results_dataco
llection/BESTUFS_II_data_collection_synthesis_report.pdf
Browne M, Sweet M, Woodburn AG, Allen J (2005) Urban freight consolidation centres. University
of Westminster, London
Browne M, Allen J (2011) Enhancing the sustainability of urban freight transport and logistics.
Transp Commun Bull Asia Pac 80:80
CIVITAS (2015) civitas.eu. Retrieved 06 12, 2019, from ‘Making urban freight logistics more
sustainable’, Civitas Policy Note, pp 1–63. http://www.eltis.org/resources/tools/civitas-policy-
note-making-urban-freight-logistics-more-sustainable
Chwesiuk K, Kijewska K, Iwan S (2010) Urban consolidation centres for medium-size touristic
cities in the Westpomeranian Region of Poland. Proc Soc Behav Sci 2:6264–6273
Comprehensive mobility plan Jaipur. Retrieved July 15, 2019, from http://transport.rajasthan.gov.
in
Dablanc L, Giuliano G, Holliday K, O’Brien T (2013) Best practices in urban freight management:
lessons from an international survey. TRB, Transp Res Rec (T.R.R.) 29–38
de Assis Correia V, de Oliveira LK, Guerra AL (2012) Economic and environmental analysis of an
urban consolidation centre for Belo Horizonte city (Brazil). Proc Soc Behav Sci 39:770–782
Elvsaas M (2020) https://www.ntnu.edu/. Retrieved from https://www.ntnu.edu/documents/341714
296/0/Nordt%C3%B8mme+presentation/146f1d52-27a0-4991-83dc-6db1482e65df
Faure L, Burlat P, Marquès G (2016). Evaluate the viability of urban consolidation centre with
regards to urban morphology. Trans Res Proc 12:348–356
Gudmundsson H, Hall RP, Marsden G, Zietsman J (2015) Sustainable transportation—indicators,
frameworks, and performance management. Samfundslitteratur, Frederiksberg
Jaipur development authority. Retrieved May 13, 2019, from http://jda.urban.rajasthan.gov.in/
Kijewska K, Iwan S, Kaczmarczyk T (2012). Technical and organizational assumptions of applying
U.C.C.s to optimize freight deliveries in the seaside tourist resorts of the West Pomeranian Region
of Poland. Proc Soc Behav Sci 39:592–606
Leonardi J, Browne M, Allen J (2012) Before-after assessment of a logistics trial with clean urban
freight vehicles: a case study in London. Soc Behav Sci 39:146–147
Lindholm M (2010) A sustainable perspective on urban freight transport: factors affecting local
authorities in the planning procedures. Proc Soc Behav Sci 2:6205–6216
Litman T (2019) developing indicators for sustainable and livable transport planning. Retrieved
May 9, 2019, from www.vtpi.org/wellmeas.pdf
Martin R, Hegi P, Haefeli U, Matti D, Schultz B, Rütsche P (2012) Sustainable goods supply and
transport in conurbations: freight strategies and guidelines. Proc Soc Behav Sci 39:116–133
MoUD (2014) https://smartnet.niua.org/. Retrieved March 2018, from https://smartnet.niua.org/
sites/default/files/resources/National%20Urban%20Transport%20Policy.pdf
OECD (2003) Delivering the goods: 21St century challenges to urban goods transport. Retrieved
April 12, 2019, from https://www.itf-oecd.org/sites/default/files/docs/03deliveringgoods.pdf
Patier D, Routhier J-L (2008) Hoe to improve the capacity of urban goods movement data. In:
Annecy: 8th international conference on survey methods in transport
Quak H (2014) https://www.researchgate.net. Retrieved from https://www.researchgate.net/pub
lication/254805169_Sustainability_of_Urban_Freight_Transport_Retail_Distribution_and_
Local_Regulations_in_Cities
Sharma S, Shelton J, Warner V (2017) Methodologies and models for selection of optimal truck
freight management strategies. In: Transportation research board (TRB)
U.S. DoT (2012) Assessing the effectiveness of transportation management plan (T.M.P.) strategies.
Federal Highway Administration, Washington, DC 20590
An Agent-Based Approach to Evaluate Freight Consolidation Center … 27

van Duin JHR, Quak H (2010) New challenges for urban consolidation centres: a case study in The
Hague. Proc Soc Behav Sci 2:6177–6188
Van Duin R, Slabbekoorn M, Tavasszy LA, Quak H (2018). Identifying Dominant Stakeholder
Perspectives On Urban Freight Policies: A Q-Analysis On Urban Consolidation Centres In The
Netherlands. Transport, 33(4), 867-880
van Heeswijk W, Larsen R, Larsen A (2019). An urban consolidation centre in the city of
Copenhagen: a simulation study. Int J Sustain Transp 13(9):675–691
van Rooijena T, Quak H (2010) Local impacts of a new urban consolidation centre—the case of
Binnenstadservice.nl. Proc Soc Behav Sci 2:5967–5979
Verlindea S, Macharis C, Witlox F (2012). How to consolidate urban flows of goods without setting
up an urban consolidation centre? Proc Soc Behav Sci 39:687–701
Zehra Akgün E, Monios J, Fonzone A (2019) Supporting urban consolidation centres with urban
freight transport policies: a comparative study of Scotland and Sweden. Int J Logistics Res Appl
2–21
A Proposed Framework for Designing
Blockchain Solutions for Logistics
in post-Covid Scenario and Future
Pandemics

Javed Aslam, Aqeela Saleem, Nokhaiz Tariq Khan, and Yun Bae Kim

1 Introduction

Logistics companies, that are participating in the distribution, storage, and movement
of goods, have been widely affected by the Covid-19 pandemic and it is also exposed
that traditional logistics is not adaptable, agile, and resilient enough to manage a
pandemic or another widespread disaster. During the Covid-19 pandemic, it is essen-
tial for all countries to save human lives. Unfortunately, many countries are unable to
provide a smooth supply of essential items such as food, medical diagnostics equip-
ment, and medicine (Han et al. 2021; Singh et al. 2021a, b; Thompson & Anderson
2021). The reason for this problem is the traditional logistics systems. Because the
traditional logistics system is highly reliable on the human interaction and the use of
the latest technology is limited (Cigolini et al. 2004; Dunn 1995; Fleischmann et al.
2000). Moreover, logistics companies are facing many challenges in logistics activ-
ities such as transparency, traceability, visibility, immutability, auditing, trust (Dutta
et al. 2020; Litke et al. 2019; Rejeb et al. 2020). These issues can be resolved by
using emerging technologies such as blockchain, the internet of things (IoT), artifi-
cial intelligence, cloud computing, and machine learning by transforming traditional
logistics into digital or smart logistics (Ahmad et al. 2021; Chen and Ho, 2021; Su,

J. Aslam · A. Saleem · Y. B. Kim (B)

Department of System Management Engineering, Sungkyunkwan University, Seoul, South Korea
e-mail: [email protected]
J. Aslam
e-mail: [email protected]
A. Saleem
e-mail: [email protected]
N. T. Khan
Information Technology University, Lahore, Pakistan
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 29

P. Golinska-Dawson et al. (eds.), Smart and Sustainable Supply Chain
and Logistics—Challenges, Methods and Best Practices, EcoProduction,
https://doi.org/10.1007/978-3-031-15412-6_3
30 J. Aslam et al.

2012; Wang and Sarkis 2021). Digital logistics infrastructure is required to improve
the firm’s logistics performance (Kersten et al. 2017). Blockchain technology is one
of the most effective proposals for digital logistics networks because this technology
has distinctive features; transparency, traceability, cybersecurity, real-time informa-
tion sharing, and visibility which are highly crucial for the future of logistics (Aslam
et al. 2021). This study is providing the initial framework for convergence of logistics
with the help of blockchain technology. In doing so, this study highlights the major
issues in traditional logistics and identifies the need for digital logistics. This study
guides the logistics companies to understand the relationship between digital logistics
and Blockchain technology as a solution for handling these pre-and-post-pandemic
challenges and the future of logistics management.
For the remainder of the study, we distributed our study into three parts. Firstly, this
study criticizes traditional logistics and its weaknesses. Secondly, describes smart
logistics and highlights its advantages. Thirdly, this study develops the relation-
ship between smart logistics and Blockchain technology as a solution for improving
logistics performance. This paper is designed as per the following sections: Sect. 2 is
based on a brief literature review and Sect. 3 explained the proposed framework. The
discussion is expressed in Sect. 4. Section 5 is devoted to the conclusion, implication,
and future recommendation.

2 Literature Review

2.1 Logistics Activities During Pandemic Covid-19

To prevent the spread of covid-19 few policies such as work-from-home, lockdowns,

and social distancing are being used globally. Due to the implementation of safety
protocols like lockdowns, and border closures, the logistics activities are mostly
suspended. The operational efficiency was affected and it lead to the shortage of
supply, delivery delays, and higher freight rates (Farooq et al. 2021). Moreover, the
Covid-19 has hit the logistics businesses very hard as firms were not familiar with
backup operations plans, many firms also lack the technology and had no proper infor-
mation to follow the health guidelines e.g., disinfecting deliveries. Apart from this,
logistics firms have contributed very well during Covid-19 and will be an important
part of the economic recovery, however, need to upgrade the way logistics operations
work, to handle the post-pandemic scenario and any similar problem in the future.
A Proposed Framework for Designing Blockchain Solutions … 31

2.2 Traditional Versus Digital Logistics Networks

Logistics management is a key attribute of the supply chain management process,

and it deals with the flow of products, information, and capital at the same time.
The fundamental aim of logistics management is to deliver the right amount of mate-
rial/product at the right time and right place (Ghoumrassi and T, igu 2017). Traditional
logistics network is based on less integrated systems e.g., manually data input, lack
of upgraded IT system, paper-based documents, and using follow up calls to track
the delivery status, less agile and responsive, and poor flexibility, thus, these methods
are less effective and efficient (Mears-Young and Jackson 1997; Ouyang et al. 2019;
Zielske and Held 2021). Therefore, logistics organizations need to reform their func-
tions for becoming highly integrated with providing better services in today’s era,
where the emerging technologies provide the main platform for the digital infras-
tructure for logistics companies (Oluyisola et al. 2021). The adoption of these tech-
nologies into logistics enhances the integration transform the logistics to agile and
resilient this phenomenon lead to the emergence of a new concept called “digital
logistics” or “smart logistics”(Burroughs and Burroughs 2020; Hofmann and Oster-
walder 2017; Issaoui et al. 2021). Digital logistics provide smart tools and methods
such as cloud-based data computing systems, end-to-end visibility, and traceability
for planning and controlling logistics operations (Barykin et al. 2021; Issaoui et al.
2019; Moldabekova et al. 2021; Sergi et al. 2021). Digital logistics focus on more
integrated techniques, and it allows the firms to adopt vertical integration (integrated
with numinous IT systems inside the organization); horizontal integration (collab-
oration with inter-department), and end-to-end engineering integration (collabora-
tion among product, machines, and stakeholders) (Bag et al. 2020). In the broader
perspective, the investment of technology will allow the firm to reduced third-party
logistics (3PL) services and labor costs using autonomous vehicles, drones, and
robotic (Strandhagen et al. 2017).

2.3 Blockchain for Logistics

Logistics has been amidst a tech-driven revolution. In recently, logistics 4.0 widely
discussed supply chain and logistics management. Logistics 4.0 is the combination
of emerging technologies used to optimize logistics processes. These emerging tech-
nologies are based on the inclusion of artificial intelligence, machine learning, the
internet of things (IoT), cloud computing, big data analytics, blockchain, and addi-
tive manufacturing or 3D printing (Koh et al. 2019; Qu et al. 2019; Singh et al.
2021a, b). Being an emerging technology, the Blockchain provides a decentralized
platform for storing and transmitting information securely and more transparently.
Due to the involvement of multiple activities such as various delivery points, inven-
tory planning, delivery tracking, visibility of materials, and high level of data sharing
are the major challenges for logistics companies (Chen et al. 2021). To cope with
32 J. Aslam et al.

these challenges, blockchain has effective features to support the logistics operations
such as cyber-security, real-time information system, reliability, transparency, visi-
bility, and traceability or trackability (Aslam et al. 2021; Behnke and Janssen 2020;
Cole et al. 2019; Helo and Hao 2019; Phadnis 2018). These Blockchain features
help to reduce the leads time, improve Inter-Intra firm privacy, secure the finan-
cial/transaction information, improve the forecasting, better control on inventory,
and effective utilization of firm resources (Hellani et al. 2021). Literature proposed
serval theories allow the implementation of Blockchain technology into the supply
chain and logistics management. These theories are the resource-based view (RBV),
principal-agent theory (PAT), network theory (NT), and transaction-cost theory, also
called transaction-cost analysis (TCA) (Treiblmaier 2018). Traditional logistics oper-
ations are fully connected via human physical interaction, and the covid-19 pandemic
is not allowed this business style due to prevent the spread of the pandemic by imple-
menting safety regulations like social distance and lockdowns. In the absence of
technology, the logistics firms have no electronic visibility, traceability, and tracka-
bility of inventory as well as in-transit materials and it’s created the biggest problem
for logistics firms as well as many businesses due to unavailability of product infor-
mation. Therefore, the logistics firm needs to upgrade its business structure and adopt
the technology which is beneficial for the firm to uplift the logistics activities. Among
other emerging technologies, blockchain is closely related to upgrading the logis-
tics and supply chain operation because this technology application is its interface
with the physical world. Additionally, it provides significant potential for improving
information and communication at all levels of the logistics network.

3 Proposed Framework

In this study, we highlighted the major problem in traditional logistics networks, and
these are negatively affecting the firm performance and market reputation. For the
solution to these problems, we proposed the adoption of emerging technologies into
logistics and update into a new domain as digital logistics. The literature suggests that
the blockchain is one of the most effective and efficient technology for digital logistics
because this technology has state-of-the-art features which are highly synchronized
with logistics needs. Figure 1 expressed the transformation from traditional to digital
logistics networks.

4 Discussion

Logistics is the main pillar of supply chain management, and the particular impor-
tance of logistics was realized during the recent covid-19 pandemic where delivery
and the supply of many essential products were either cut or delayed. Traditional
logistics have many problems which are expressed in Fig. 1. This paper proposed the
A Proposed Framework for Designing Blockchain Solutions … 33

Fig. 1 Framework for transformation towards digital logistics using Blockchain

framework for the transformation towards digital logistics as adoption of emerging

technologies. The reason behind this transformation is that digital logistics have
an important attribute that is highly needed for 21st-century logistics and will be
required for post-covid to recover the economy. This study provided a relation-
ship among emerging technologies with digital logistics from literature. The adop-
tion of all these emerging technologies is quite difficult because this concept for
prior practices may be costly. This study proposed Blockchain technology for this
adoption because this technology has serval key features such as traceability, track-
ability, real-time information sharing, secure financial platform, and transparency
which are highly recommended for logistics networks. From the pandemic perspec-
tive, blockchain technology is the best solution for performed logistics activity with
limited human-to-human interaction because blockchain provides a state-of-the-art
technology structure for track and trace the inventory in few electronic clicks. On
the other hand, blockchain has some advantages among other emerging technologies
such as it provides the best security platform to support the next industrial revolution
which is known as industry 5.0. Blockchain is also a useful security tool for the adop-
tion of the Internet of things (IoT). Furthermore, the implementation of blockchain
will provide generate a lot of data that can be used to re-design the warehouse and
optimization the delivery choosing the best route for transportation.

5 Conclusion

This paper provides very interesting findings, firstly, this study provides the actual
picture of where traditional logistics stand and highlighted its main problems.
Secondly, this study develops the relevancy among emerging technologies and digital
34 J. Aslam et al.

logistics and suggests that digital logistics is the solution to the pandemic and future
challenges of logistics. Thirdly, among the serval emerging technology, this study
recommends that the blockchain is the most relevant technology for digital logistics
because blockchain has unique features which are highly recommended for digital
logistics networks. Consequently, Blockchain technology is an important factor for
the 5th industrial revolution or Industry 5.0 due to it provides highly secure infor-
mation networks and supports the implementation of IoT. Moreover, this study also
expressed that how traditional logistics are not suitable for pandemics and this study
guides managers of logistics companies to understand the weakness of traditional
logistics and the use of Blockchain to modernize the firm logistics function. This
study guides the logistics companies to understand the relationship between digital
logistics and Blockchain technology as a solution for handling this type of pandemic.
It is very difficult to adopt all the latest emerging technologies of small and medium
enterprises (SME’s) as well as giants because it is very costly but due to the need
for future of logistics firms should be evaluated their technology need and alliance
with these technologies. The main limitation of this study is based on the litera-
ture only and could not be explained empirically. The real-time implementation of
Blockchain technology for logistics can open a new horizon of research for handling
the post-pandemic and future challenges of logistics networks.

References

Ahmad RW, Hasan H, Jayaraman R, Salah K, Omar M (2021) Blockchain applications and archi-
tectures for port operations and logistics management. In: Research in transportation business
and management. https://doi.org/10.1016/j.rtbm.2021.100620
Aslam J, Saleem A, Khan NT, Kim YB (2021) Factors influencing blockchain adoption in supply
chain management practices: a study based on the oil industry. J Innov Knowl 6(2):124–134.
https://doi.org/10.1016/j.jik.2021.01.002
Bag S, Gupta S, Luo Z (2020) Examining the role of logistics 4.0 enabled dynamic capabilities
on firm performance. Int J Logistics Manage 31(3):607–628. https://doi.org/10.1108/IJLM-11-
2019-0311
Barykin SY, Kapustina IV, Kalinina OV, Dubolazov VA, Esquivel CAN, Alyarovna NE, Sharapaev
P (2021) The sharing economy and digital logistics in retail chains: opportunities and threats.
Acad Strateg Manag J 20(SpecialIssue2):1–14
Behnke K, Janssen MFWHA. (2020). Boundary conditions for traceability in food supply chains
using blockchain technology. Int J Inf Manage 52(2019):101969. https://doi.org/10.1016/j.ijinfo
mgt.2019.05.025
Burroughs B, Burroughs WJ (2020) Digital logistics: enchantment in distribution channels. Technol
Soc 62. https://doi.org/10.1016/j.techsoc.2020.101277
Chen M-C, Ho PH (2021) Exploring technology opportunities and evolution of IoT-related logistics
services with text mining. Complex Intell Syst 0123456789. https://doi.org/10.1007/s40747-021-
00453-3
Chen CL, Deng YY, Weng W, Zhou M, Sun H (2021) A blockchain-based intelligent anti-switch
package in tracing logistics system. J Supercomput 77(7). https://doi.org/10.1007/s11227-020-
03558-7
A Proposed Framework for Designing Blockchain Solutions … 35

Cigolini R, Cozzi M, Perona M (2004) A new framework for supply chain management: conceptual
model and empirical test. Int J Oper Prod Manag 24(1–2):7–41. https://doi.org/10.1108/014435
70410510979
Cole R, Stevenson M, Aitken J (2019) Blockchain technology: implications for operations and
supply chain management. Supply Chain Manag 24(4):469–483. https://doi.org/10.1108/SCM-
09-2018-0309
Dunn SC (1995) Environmentally responsible logistics systems. Int J Phys Distrib Logist Manag
25(2):20–38. https://doi.org/10.1108/09600039510083925
Dutta P, Choi TM, Somani S, Butala R (2020) Blockchain technology in supply chain opera-
tions: applications, challenges and research opportunities. Transp Res Part E Logist Transp Rev
142(September):102067. https://doi.org/10.1016/j.tre.2020.102067
Farooq MU, Hussain A, Masood T, Habib MS (2021) Supply chain operations management in
pandemics: a state-of-the-art review inspired by covid-19. Sustainability (Switzerland) 13(5):1–
33. https://doi.org/10.3390/su13052504
Fleischmann M, Krikke HR, Dekker R, Flapper SDP (2000) A characterisation of logistics networks
for product recovery. Omega 28(6):653–666. https://doi.org/10.1016/S0305-0483(00)00022-0
Ghoumrassi A, T, igu G (2017) The impact of the logistics management in customer satisfaction.
Proc Int Conf Bus Excel 11(1):292–301. https://doi.org/10.1515/picbe-2017-0031
Han J, Zhang X, He S, Jia P (2021) Can the coronavirus disease be transmitted from food? A review
of evidence, risks, policies and knowledge gaps. Environ Chem Lett 19(1):5–16. https://doi.org/
10.1007/s10311-020-01101-x
Hellani H, Sliman L, Samhat AE, Exposito E (2021) On blockchain integration with supply chain:
overview on data transparency. Logistics 5(3):46. https://doi.org/10.3390/logistics5030046
Helo P, Hao Y (2019) Blockchains in operations and supply chains: a model and reference
implementation. Comput Ind Eng 136(July):242–251. https://doi.org/10.1016/j.cie.2019.07.023
Hofmann E, Osterwalder F (2017) Third-party logistics providers in the digital age: towards a new
competitive arena? Logistics 1(2):9. https://doi.org/10.3390/logistics1020009
Issaoui Y, Khiat A, Bahnasse A, Ouajji H (2019) Smart logistics: study of the application of
blockchain technology. Proc Comput Sci 160(2018):266–271. https://doi.org/10.1016/j.procs.
2019.09.467
Issaoui Y, Khiat A, Bahnasse A, Ouajji H (2021) Toward smart logistics: engineering insights and
emerging trends. Arch Comput Methods Eng 28(4):3183–3210. https://doi.org/10.1007/s11831-
020-09494-2
Kersten W, Blecker T, Ringle MC (2017) Digitalization in supply chain management and logistics.
In: Proceedings of the Hamburg international conference of logistics (HICL). https://doi.org/10.
15480/882.1442
Koh L, Orzes G, Jia F (2019) The fourth industrial revolution (Industry 4.0): technologies disruption
on operations and supply chain management. Int J Oper Prod Manag 39(6):817–828. https://doi.
org/10.1108/IJOPM-08-2019-788
Litke A, Anagnostopoulos D, Varvarigou T (2019) Blockchains for supply chain management:
architectural elements and challenges towards a global scale deployment. Logistics 3(1):5. https://
doi.org/10.3390/logistics3010005
Mears-Young B, Jackson MC (1997) Integrated logistics—call in the revolutionaries! Omega
25(6):605–618. https://doi.org/10.1016/S0305-0483(97)00032-7
Moldabekova A, Philipp R, Reimers HE, Alikozhayev B (2021) Digital technologies for improving
logistics performance of countries. Transp Telecommun 22(2):207–216. https://doi.org/10.2478/
ttj-2021-0016
Oluyisola OE, Bhalla S, Sgarbossa F, Strandhagen JO (2021) Designing and developing smart
production planning and control systems in the industry 4.0 era: a methodology and case study.
J Intell Manuf 0123456789. https://doi.org/10.1007/s10845-021-01808-w
Ouyang Q, Zheng J, Wang S (2019) Investigation of the construction of intelligent logistics system
from traditional logistics model based on wireless network technology. Eurasip J Wirel Commun
Netw 2019(1). https://doi.org/10.1186/s13638-018-1334-8
36 J. Aslam et al.

Phadnis S (2018) Internet of things and supply chains: framework for identifying oppertunities for
improvement and its application. Emerg Technol Supply Chain Manag
Qu YJ, Ming XG, Liu ZW, Zhang XY, Hou ZT (2019) Smart manufacturing systems: state of the
art and future trends. Int J Adv Manuf Technol 103(9–12):3751–3768. https://doi.org/10.1007/
s00170-019-03754-7
Rejeb A, Keogh JG, Zailani S, Treiblmaier H, Rejeb K (2020) Blockchain technology in the food
industry: a review of potentials, challenges and future research directions. Logistics 4(4):27.
https://doi.org/10.3390/logistics4040027
Sergi I, Montanaro T, Benvenuto FL, Patrono L (2021) A smart and secure logistics system based
on IoT and cloud technologies. Sensors 21(6). https://doi.org/10.3390/s21062231
Singh R, Gehlot A, Akram SV, Gupta LR, Jena MK, Prakash C, Singh S, Kumar R (2021a) Cloud
manufacturing, internet of things-assisted manufacturing and 3D printing technology: reliable
tools for sustainable construction. Sustainability (Switzerland) 13(13). https://doi.org/10.3390/
su13137327
Singh S, Kumar R, Panchal R, Tiwari MK (2021b) Impact of COVID-19 on logistics systems and
disruptions in food supply chain. Int J Prod Res 59(7):1993–2008. https://doi.org/10.1080/002
07543.2020.1792000
Strandhagen JO, Vallandingham LR, Fragapane G, Strandhagen JW, Stangeland ABH, Sharma
N (2017) Logistics 4.0 and emerging sustainable business models. Adv Manuf 5(4):359–369.
https://doi.org/10.1007/s40436-017-0198-1
Sun C (2012) Application of RFID technology for logistics on internet of things. AASRI Proc
1:106–111. https://doi.org/10.1016/j.aasri.2012.06.019
Thompson DDP, Anderson R (2021) The COVID-19 response: considerations for future human-
itarian supply chain and logistics management research. J Hum Logist Supply Chain Manag
11(2):157–175. https://doi.org/10.1108/JHLSCM-01-2021-0006
Treiblmaier H (2018) The impact of the blockchain on the supply chain: a theory-based research
framework and a call for action. Supply Chain Manag 23(6):545–559. https://doi.org/10.1108/
SCM-01-2018-0029
Wang Y, Sarkis J (2021) Emerging digitalisation technologies in freight transport and logistics:
current trends and future directions. Transp Res Part E Logist Transp Rev 148(March). https://
doi.org/10.1016/j.tre.2021.102291
Zielske M, Held T (2021) Application of agile methods in traditional logistics companies and
logistics startups: results from a German Delphi Study. J Syst Softw 177:110950. https://doi.org/
10.1016/j.jss.2021.110950
Application of Big Data Analytics
in Modern Logistics Solutions

Jozef Fras and Waldemar Osmólski

1 Introduction

Changing market needs and concentration of the entire supply chain on the customer
service level generate the need to constantly search for new solutions to improve
logistics processes. This trend leads to a transformation of the currently used solu-
tions, in the direction of intelligent supply chains. The use of modern technologies
is aimed at improving logistics processes at the operational level by reducing lead
times, minimizing bottlenecks and errors resulting from faulty information flow.
The encroaching robotization, mobility and automation into logistics processes in
our reality is a fact. Solutions using the latest achievements of electronics and robotics
largely contribute to lowering the costs of functioning of enterprises, speeding up
processes taking place in them or relieving people from tiresome and exhausting
work. This also applies to mobile solutions in which data is retrieved and transferred
in real time from mobile devices to enterprise IT systems, in particular based on
5G network capabilities. The pressure to automate is growing as companies look
to reduce labor costs. Many companies that are currently considering, for example,
relocating production to developed and economically stable countries may view
automation as a means to remain competitive with rising labor costs.
Restructuring efforts in integrated supply chains are increasingly aided by Artifi-
cial Intelligence (AI). As in other industries, AI will fundamentally expand human
performance in terms of reach, quality and speed by eliminating mundane and routine

J. Fras
Faculty of Management Engineering, Poznan University of Technology, Jacka Rychlewskiego 2,
60-965 Poznan, Poland
e-mail: [email protected]
W. Osmólski (B)
Poznan School of Logistics, Estkowskiego 6, 61-755 Poznan, Poland
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 37

P. Golinska-Dawson et al. (eds.), Smart and Sustainable Supply Chain
and Logistics—Challenges, Methods and Best Practices, EcoProduction,
https://doi.org/10.1007/978-3-031-15412-6_4
38 J. Fras and W. Osmólski

work. This approach allows logistics workers to focus on more meaningful issues
and impactful work. Another very important aspect is the inability to utilize the large
amount of data (Big Data) generated on a daily basis in integrated supply chains
in the traditional way. Many logistics companies around the world are currently
undergoing a digital transformation, moving away from legacy enterprise resource
planning systems to advanced analytics, increased automation, and hardware and
software robotics. This is helping the logistics industry to redefine today’s behaviors
and practices, taking actions from reactive to proactive, planning from predictive to
predictive, and processes from manual to autonomous.

2 Big Data and Its Analysis

In order to answer the question: what really is Big Data and its analysis, we must know
that Big Data generally refers to data that exceeds the typical storage, processing and
computational capabilities of conventional databases and data analysis techniques.
Most commonly, Big Data (Katal et al. 2013) is characterized as gigantic or highly
complex data sets, encompassing resources larger than an exabyte in size (Chaouni
Benabdellah 2016). On the other hand, Big Data describes how to collect, manage and
analyze large amounts of data, associated with the so-called 3V’s concept: Volume,
Velocity, Variety (Mikavica 2015; Robak et al. 2013). This model has been supple-
mented over time with another component, Value, i.e., verification, evaluation. In
the Polish language version, the Big Data model is defined as: Utilization, Inference,
Enrichment, Verification. Currently, it is also noted that Big Data is nothing more than
specific sets of information with a large volume, high variability and high diversity,
requiring new forms of processing to support new decision-making and discovery
of new phenomena, related to the optimization of various processes. However, to
fully understand the impact and application of Big Data in supply chain analysis, we
must first understand its nature and how it is collected. What is also important is the
fact that the importance of Big Data does not boil down solely to how much data a
company has, but most importantly how it is able to collect it, process it and on this
basis draw the appropriate conclusions. They are the basis for making the right and
fast decisions that enable dynamic development of any company. To be able to do
that we need to pay special attention to 4 basic aspects related to data management:
. Data must be accessible: Organizations must make it easy and convenient for
data owners at all levels to use data.
. Data must be high quality: Organizations are spending more time than ever
before searching for duplicates, errors, omissions, conflicts and inconsistencies.
. Data must be secure: Organizations will need to strive for compliance and put
strict data processes in place before using big data.
. Data must be analyzed using the right tools and platforms: Organizations must
find the right technology that works within their existing ecosystems and meets
their specific needs.
Application of Big Data Analytics in Modern Logistics Solutions 39

One of the most significant problems of Big Data is that it is often erroneous,
outdated or incomplete. Therefore, an extremely large amount of time should be
spent on their initial selection, eliminating errors and inaccuracies, resulting from
the way they are obtained. These actions, in consequence, lead to the reduction of
time allocated for the creation of innovative solutions (Chaouni Benabdellah 2016).
Poor data quality can also lead to higher supply chain operating costs, which in turn
is associated with lower efficiency and wasted resources (Mikavica 2015). Also, lack
of sufficient knowledge by data analysts forces data to be entered manually, which in
turn generates a higher probability of errors leading to disruptions in supply chains
(Wang 2016). Another of the elements that can introduce dysfunctions in the proper
functioning of supply chains is usually the lack of a functional and transparent user
interface for analyzing Big Data (Taniguchia 2015), or the amount of available data.
This can vary from region to region, for example, in small economies and rural areas,
only a small amount of data may be available as opposed to metropolitan cities and
urban areas (Chaouni Benabdellah 2016), which can have a significant impact on
logistics and supply chain activities in these regions. To overcome the above diffi-
culties, Big Data analytics should be carefully embedded in the operational realities
of the supply chain (Sahin and Robinson 2002) using adequate human resources on
both the analysis and operations side (Najafabadi et al. 2015). What is also important
is that Big Data requires tools and methods that can be used to analyze and extract
patterns from large-scale data (Katal et al. 2013). In other words, Big Data analytics is
nothing but a complex process of examining big data to discover information—such
as hidden patterns, correlations, market trends, and customer preferences that can
help organizations make informed business decisions. It is also a form of advanced
analytics that involves complex applications with elements such as predictive models
and statistical algorithms. What is extremely important is that the logistics sector is
ideally positioned to benefit from the technological and methodological advances
of Big Data. Today, logistics operators manage huge flows of goods while creating
gigantic data sets. As a result, there is an opportunity to eliminate the current divisions
and high competition in the service market, enabling the creation of interoperable
ecosystems, striving for the development of the Internet of Things (Osmólski et al.
2019). Many service providers are realizing that Big Data is a trend that is changing
the face of the logistics industry. In a recent survey on supply chain trends, sixty
percent of respondents said they plan to invest in Big Data analytics within the next
five years.

3 Application Areas of Big Data Analysis in Supply Chains

In order to effectively leverage Big Data analytics applications, it becomes necessary

to acquire clean and data from a variety of sources, which may include:
. social media data
. financial forecasts for enterprises
40 J. Fras and W. Osmólski

Fig. 1 Application areas for big data analytics in supply chains. Source Own elaboration

. traffic and weather data from sensors, monitors and forecasting systems installed
in vehicles
. vehicle diagnostics, driving patterns and location information
. enterprise data from operating systems.
As we have seen, there are many sources from which analytical systems can be
fed with the necessary information. All these data sources and skilful use of analyses
created on their basis lead to full optimization of production and warehouse processes,
optimization of inventory levels, delivery processes, verification of customer satis-
faction level or increase in financial effectiveness of enterprises. All this happens
thanks to cooperation and proper flow of information in integrated supply chains. An
important issue at this point becomes the question of sources of obtaining relevant
data, which are various types of sensors and devices installed in individual nodes of
the supply chain. Figure 1 shows the 5 major nodes of the supply chain that are the
main sources of Big Data generation.
Based on the data acquisition systems configured in this way, we are able to
identify 5 main areas of use for Big Data analytics, which are.

3.1 Optimization of Transport Processes

Based on the use of Big Data analytics technology, through RFID tags, GPS devices,
barcodes and more, logistics companies can track their vehicles in real time. These
systems can capture real-time traffic data, road network data and fleet data, which
entails the ability to optimize routes, planning and scheduling of deliveries. Based
Application of Big Data Analytics in Modern Logistics Solutions 41

on such in-depth analytics, it becomes possible to predict possible accidents or bad

weather conditions and track shipments and fleets from anywhere. Real-time delivery
status updates are also offered to customers. Customers receive automatic notifica-
tions in case of delivery delays. While there are other use cases for Big Data in
transportation, it’s not about the scope of the analysis, but the depth of the analysis.
The two basic requirements for good data analytics are data quantity and data quality.
An organization must be able to capture hundreds of thousands of events and ensure
their quality every time in order to make accurate predictions. It is also essential to
have the ability to stream analytics and analytics in real-time, especially in time-
sensitive data use cases. What is important is that it is critical for organizations to
perform analytics in a timely manner to highlight the true value of data in transit.
This is where the application of artificial intelligence (AI), or machine learning
(MI) to transportation processes comes to the rescue. Such a solution has been used
by DHL in its Smart Truck solution or Cargonnex for transportation planning, among
others. AI is also used in solutions for managing autonomous vehicles in both local
distribution and linehaul. Such a solution, called platooning, was tested, among
others, on the A9 freeway in Germany in cooperation between Schenker and MAN
during the transport of goods in a convoy of trucks.
Another example is the use of AI in air freight transport. Timely shipments here
are crucial to the economy, as they account for just 1% of global trade in terms of
tonnage, but as much as 35% in terms of value. DHL has developed a predictive
algorithm-based tool to predict delays during air freight. By analyzing 58 different
internal data parameters, the software is able to predict weeks in advance what
the expected average daily transit time is for planned connections. In addition, the
solution is able to identify the most important factors affecting shipment delays.

3.2 Optimization of Warehouse Processes

Warehouse management in the traditional way, using legacy systems or ERP systems
is already highly inefficient. Today, millions of customers are interested in receiving
real-time updates on product orders, getting up-to-date knowledge of product avail-
ability before making a purchase decision, and have instant access to product manu-
facturing details. Moreover, customers can make purchases from anywhere in the
world. With Big Data, you can understand how customer behavior is changing and
what expectations they often have of manufacturers and supply chain managers. You
can also gain detailed insights into the process of loading, moving, unloading and
delivering goods. As a result, you can better plan routes and delivery schedules to
increase safety and reduce operating costs for your business.
A very important path forward for logistics, especially warehouse logistics, is
robotization, using not only AI solutions, but also Internet of Things and Indus-
trial Internet of Things or Cloud Computing. According to the U.S. Association for
Advancing Automation, robot orders in 2018 increased nearly 16% over 2017 to
28,400 units. The largest increase in automation was in the food sector last year.
42 J. Fras and W. Osmólski

Robot shipments to companies in the electronic components manufacturing industry

increased by half, and by 13% in the metal manufacturing and processing industry.
Intelligent robotic sorting is an efficient and fast solution for sorting letters, pack-
ages and even palletized shipments. UK-based Ocado has reduced the time it takes
to prepare an order consisting of 50 lines from two hours to five minutes by using
robots in the picking process. Ocado currently already has 2500 such robots. The
Finnish ZenRobotics ZRR2 system, which uses a combination of embedded vision
and machine learning algorithms, is used for sorting and collecting recyclables from
moving conveyor belts. Autonomous AGVs have been used in warehouse processes
and drones or vision technologies have been used in goods inventory. These are only
a few examples confirming the use and development of the latest technologies in the
area of logistics.

3.3 Inventory Management System Optimization

Stock management system is considered as one of the most essential elements of

warehouse processes. Managing inventory in a seamless manner is not possible
by using conventional forms such as analyzing historical inventory data or sales
values. To operate effectively, retailers and supply chain managers need to obtain
information about customer behavior, product turnover, store performance, supplier
relationships, replenishment planning, etc. Inventory management must go beyond
traditional methods such as analyzing historical sales and inventory data. Algorithms
can explore patterns and relationships between various data elements and business
decisions. This gives retailers unparalleled insight into consumer behavior, supplier
relationships, product performance, offline and online store performance, replen-
ishment planning and more. Among other things, connecting multiple outlets can
provide essential information for forecasting optimal inventory levels. Real-time
data gives companies the ability to predict future sales of products that have no
previous sales history. Big Data provides valuable information that can help compa-
nies leverage related product details used to predict the sales potential of a new item,
seasonal product or help uncover new merchandising opportunities. Also, machine
learning and real-time data can help companies determine the average price for their
products by quickly assessing multiple factors, including available supplies, costs,
competitor prices and overall product value. Big Data also provides the ability to
predict demand for specific products. Knowing which products are bestsellers and
which are underperforming gives companies the ability to address any potential prob-
lems in real time. In particular, the real-time data obtained helps companies reduce
the number of backlogged items. Big Data uses predictive analytics to anticipate
consumer demand. These insights help improve planning and give manufacturers
more peace of mind and confidence when communicating with customers. This is
especially important in the financial aspect, where Big Data analytics can help create
strategic alliances with product suppliers by providing important real-time informa-
tion about orders, shipments and backlogs, so companies can gain valuable insights
Application of Big Data Analytics in Modern Logistics Solutions 43

into productivity and financial health. Understanding how to use Big Data in supplier
relationships can lead to better customer service and increased customer satisfaction.
As an example, e-commerce giant Walmart is among the large companies that
use analytics to keep their warehouses running smoothly. The company is present
in 27 countries and has an estimated 11,766 stores currently in operation, which
requires a lot of maintenance of warehouses in these countries to ensure a smooth
flow of goods to end users. The company relies on Big Data analytics to monitor
merchandise flows and labor productivity in real-time at both stores and distribution
centers. With a huge pool of data from all sources, it has the ability to accurately
plan the delivery of specific assortments to the appropriate points. This is achieved
by monitoring customer preferences, shopping patterns, which speeds up decision
making on how to stock store shelves and display merchandise. Big Data also helps in
obtaining detailed information on new products, discontinued products and specific
brands that are in higher demand.

3.4 Predictive Maintenance of Machinery and Equipment

Predictive maintenance of machinery and equipment is based on IoT/IIoT sensors

placed in the respective equipment systems of the monitored equipment. The Internet
of Things collects data in real time and sends it back to a central system for moni-
toring the status of individual systems. The relevant software takes the monitoring
data showing the regular operation of the system and automatically analyzes it to
establish a baseline for proper functioning. The predictive maintenance system then
uses machine learning to quickly process and analyze new data packets coming in
from IoT sensors all the time, monitoring the condition of the devices based on
anomalies and generating an alert when needed. Combining this data with static
information such as maintenance schedules makes it easy to gain contextual insights
into the status of monitored systems. Using predictive analytics based on the infor-
mation captured, companies are able to identify patterns based on which they can
predict deteriorating performance, inconsistencies and future failures. This allows
them to keep their equipment in good functional condition, thereby increasing their
productivity by avoiding costly downtime. What is also important to realize is that
not all failure modes can be predicted by predictive maintenance techniques. Some
failure modes may give no warning of impending failure, others may give some
warning but not enough to avoid the consequences of failure, and for others the cost
of detecting impending failure may outweigh the benefits. However, for all predictive
maintenance programs, it is true that in order to create an effective failure prediction
system, it is necessary to know the causes (modes) of failure to be avoided, as a kind
of benchmark necessary to verify ongoing analyses. At this point, it is important to
emphasize that for the prediction of failures against which systems must be protected,
requiring the use of more advanced techniques (such as those that use complex data
analysis and complex modeling techniques), it must be economically justified. In
44 J. Fras and W. Osmólski

other words, advanced predictive technologies must be focused on the significant

economic benefits of their use, not simply driven by technical need.
Predictive maintenance techniques have been used, among other things, to predict
the failure of Typhoon fighter components. The failure rate of aircraft components
varies with changing conditions due to the operating location and, therefore, the
location of the air base. In total, about 2000 components on each aircraft are monitored
for wear, with each verification generating a 40-megabyte output that becomes the
input for the predictive model. This runs various iterations of the model to predict
the impact of trends and environmental changes on the proper functioning of the
tested components. Using this type of approach has enabled decisions to be made
on whether to replace components before they fail, from an airbase or headquarters
location based in the UK. As a result, aircraft uptime has improved significantly—
for example, the downtime of an aircraft on the ground waiting for spare parts has
improved from double digits to 4%. The expected benefits are a savings of $3.1
billion in maintenance over 25 years for 10 aircraft.

3.5 Improve Customer Service

In order for a company to be successful in an ever-changing marketplace, it becomes

essential to know the exact needs of individual customers and, most importantly, to
meet those needs in a way that allows the company to be recommended to future
potential customers. Unfortunately, companies are not always able to grow their
business and keep up with each individual customer, which very often leads to their
partial loss. To be able to prevent this process in an effective way, Big Data analytics is
starting to be used on a large scale, thus enabling insights into the needs of individual
customers, learning about their preferences and habits. Big Data analytics provides
logistics companies with the right set of data about their customers. The data collected
from multiple customer-product interactions helps build statistical models that can
accurately predict customer actions. With them, they can apply historical and predic-
tive analytics models and understand how to create a sense of customer loyalty and
improve their relationship with the company. In-depth analytics also help predict
the emergence of potential problems and enable the organization to address them
earlier. Such initiatives help improve performance in terms of building and deep-
ening a sense of satisfaction among customers. What is also extremely important is
that based on Big Data analytics, revenue growth can be accelerated through precise
customer segmentation and targeted marketing efforts, increasing the company’s
sales potential.
Brand awareness is another way in which Big Data can have a significant impact
on building a company’s relationship with its customers. The Aberdeen Group’s
Data-Driven Retail study found that “data-driven retailers enjoy a 2.7 times greater
annual increase in brand awareness (20.1% vs. 7.4%) compared to everyone else.”
The breadth of customer data insights allows marketers to present customized content
when and where it is most effective for brand recognition and recall. Among other
Application of Big Data Analytics in Modern Logistics Solutions 45

things, this also leads to more effective retention and acquisition of new customers.
A McKinsey study found that “users who make heavy use of Big Data analytics
tools to learn about customer behavior are 23 times more effective in their efforts
than competitors who do not.” Big Data can help companies leverage real-time data
collected in cloud solutions. Big Data’s ability to acquire, process and analyze real-
time data quickly and accurately enough to take immediate and effective action
cannot be matched by any other technology. This is critical when analyzing data
from GPS, IoT sensors, website clicks or other data received in real-time. It provides
business intelligence that results in time and cost savings by optimizing marketing
efforts.
As an example of how Big Data can retain customers and their interests, let’s look
at Walmart. Walmart is building 40 (PB) Petabyte or 40 million Gigabyte of data in
the cloud to track the endless number of products they ship worldwide. Whereas this
retail giant in the old days had to wait days or months to know the status of their
sales in a selected sector or area and then make adjustments accordingly, now they
can do it instantly. This gives the achievement of a huge competitive advantage to
the company.

4 Conclusions

Using Big Data in logistics can improve the efficiency of procurement, distribu-
tion and manufacturing networks. According to a recent study, 91.6% of Fortune
1000 companies are increasing investments related to Big Data and AI. This study
clearly indicates that many companies have started using Big Data analytics, real-
izing the benefits it can provide and the impact it can have on their organizations.
By providing real-time insights to solve ongoing problems, Big Data analytics can
help improve warehousing, handling, transportation and other logistics processes
within a company. Thus, logistics operators can gather vital information that is highly
beneficial to a company’s accountability, visibility and customer service. Logistics
companies are aware of these opportunities and are trying to make decisions based
more on accurate Big Data analysis, which the authors have clearly highlighted in
the following article. In conclusion, we can state unequivocally that without accurate
Big Data analysis, modern companies are unable to create a competitive advantage
in the market, becoming, in part, outsiders slowly disappearing from the economic
space. However, using the benefits of data solutions can, among other things:
1. Create more effective marketing and sales
Big Data is increasingly being used for advanced consumer segmentation,
automating product personalization, tailoring communications during the sales
cycle or leveraging new sales opportunities.
2. Increase the ability to forecast demand more accurately
Sales teams in today’s enterprises now have access to vast and complex amounts
of data from a variety of sources, including suppliers and customers. Advanced
46 J. Fras and W. Osmólski

analytics tools can be used to integrate data from multiple systems, and when
combined with external factors such as weather forecasts, competitive behavior
and pricing, will greatly enhance the ability to forecast demand.
3. Help with route planning and traffic management
Big Data analytics is the ideal way to plan routes—avoiding congestion as much
as possible—and having absolute control over routes or times.
4. Reduce and optimize costs
Capturing transportation fleet data not only allows you to have more control over
the information, but also to tailor it better. From there, it becomes easier to make
decisions.
5. Reduce environmental impact
Today, more and more companies are trying to control and develop strategies to
reduce their “environmental footprint”. These actions will be much more effective
if they are developed in real time based on data obtained during the company’s
daily operations.
6. Solve more complex distribution network problems
An increasingly complex distribution network consisting of suppliers, manufac-
turers, distributors, or logistics operators faces increasing challenges in managing
the flow of goods, information, or changing demand patterns. Companies can deal
with this complexity more easily than in the past by leveraging Big Data analytics,
which provides the ability to solve highly complex problems by modeling results
in more detailed scenarios than ever before.
7. Develop better supply chain collaboration
Increased supply chain data should be seen as an opportunity to improve the
management of more complex supplier networks and to develop better collabo-
ration. Getting the most benefit from Big Data requires investment in technology,
of course, but also a change in organizational culture. Employees in different
departments of companies need to be involved in the process of determining
which data is useful to them and which should be ignored. Recognizing and
rewarding employees who can effectively use the analytical capabilities of the
data collected is just as important as investing in the latest analytical techniques.

References

Chaouni Benabdellah A (2016) Big data for supply chain management: opportunities and challenges.
Int J Sci Eng Res 7(11)
Katal A, Wazid M, Goudar RH (2013) Big data: issues, challenges, tools and good practices. In:
IEEE sixth international conference on contemporary computing (IC3), pp 404–409
Mikavica B (2015) Big data: challenges and opportunities in logistics systems. In: 2nd logistics
international conference, Serbia
Najafabadi MM, Villanustre F, Khoshgoftaar TM, Seliya N, Wald R, Muharemagic E (2015) Deep
learning applications and challenges in big data analytics. Springer J Big Data 2(1)
Application of Big Data Analytics in Modern Logistics Solutions 47

Osmólski W, Voronina R, Koliński A (2019) Verification of the possibilities of applying the prin-
ciples of the physical internet in economic practice, LogForum, Nr 1 rok 2019, Poznań, str.
7–17
Robak S, Franczyk B, Robak M (2013) Applying big data and linked data concepts in supply
chains management. In: 2013 federated conference on computer science and information systems
(FedCSIS), pp 1215–1221, 201
Sahin F, Robinson EP (2002) Flow coordination and information sharing in supply chains: review,
implications, and directions for future research. Decis Sci 33(4):505–536
Taniguchia E (2015) New opportunities and challenges for city logistics. In: 9th international
conference on city logistics, Canary Islands
Wang G (2016) Big data analytics in logistics and supply chain management: certain investigations
for research and applications. Int J Prod Econ 98–110. http://www.elsevier.com
Case Study Validation of a Predictive
Maintenance Implementation
Framework

Marcel André Hoffmann

1 Introduction

Industrial companies are faced with many tasks in efficiency, sustainability, and
customer requirements (Mack et al. 2015). To solve these multiple tasks, reliable
processes in production and logistics are inevitable. In a strategy paper of the Euro-
pean Commission, smart and adaptive manufacturing systems play a crucial role
in facing the challenges in supply chains (European Commission 2021). This also
includes the adaption of smart solutions in the management of reliability and mainte-
nance of production systems, accompanied by large amounts of data to be processed
(Feng and Shanthikumar 2018).
The European Standard DIN EN 13306 divides maintenance strategies into four
categories. In the reactive maintenance strategy, measures are only implemented
after failure. This leads to more extended downtimes and no planning or forecasting
opportunities. In a time-based preventive strategy, the useful lifetimes of parts are
defined and documented in a maintenance schedule, and parts are replaced according
to this plan. This strategy is characterized by higher maintenance activity and spare
parts consumption since the maximum lifetime is usually not reached (Erbe et al.
2005). The condition-based preventive maintenance follows the concept of either
constant or discrete observation of the object’s condition state. The monitoring can
be achieved by sensors or manual inspections (Prajapati et al. 2012).
Consequently, parts can be replaced before a breakdown occurs, and the wear
reserve can be used to a high extension. The predictive maintenance (PdM)
strategy extends the CBM, whereas a prognosis of the residual useful life (RUL)
allows good planning opportunities for scheduling maintenance activities and

M. A. Hoffmann (B)
Chair of Business Management, Esp. Logistics, Technische Universität Dresden, 01062 Dresden,
Germany
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 49

P. Golinska-Dawson et al. (eds.), Smart and Sustainable Supply Chain
and Logistics—Challenges, Methods and Best Practices, EcoProduction,
https://doi.org/10.1007/978-3-031-15412-6_5
50 M. A. Hoffmann

deploying resources. Furthermore, this strategy aims to inhibit failures and, therefore,
production downtimes (CEN 2017; Lei et al. 2018; Ansari et al. 2019).
Although PdM shows high potential in maintenance optimization, the number of
implementations in the industrial environment is still comparatively low (Haarman
et al. 2018). Therefore, Hoffmann and Lasch (2020) developed a roadmap for a
structural implementation process of this maintenance strategy based on literature
findings, which has not been applied yet. This paper aims to validate this framework
based on a practical case study. The following research questions are considered:
RQ1: Are all relevant aspects included in the implementation framework of a
predictive maintenance strategy?
RQ2: In which way and order are the proposed steps put into practice by a
maintenance expert team?
The remainder of the article is structured as follows: Sect. 2 provides a short
introduction to the roadmap of Hoffmann and Lasch (2020) and gives an overview
of the proposed methodology of the validation study. Section 3 shows the practical
realization of the analysis phase, Sect. 4 includes the decision-making process, and
in Sect. 5, the implementation process is shown. A conclusion is drawn in Sect. 6.

2 Predictive Maintenance Implementation Framework

and Validation Methodology

2.1 PdM Implementation Framework

The paper of Hoffmann and Lasch (2020) proposes a generic framework to implement
a predictive maintenance strategy in industrial processes. Therefore, technical criteria
of this maintenance strategy are considered as well as management-related tasks such
as cost–benefit-considerations or decision-making support. The overall aim is to give
a structured roadmap to the maintenance management on implementing PdM in an
industrial environment that does not require extensive know-how and splits up this
assignment into several packages of strategic and tactical tasks.
The general structure is divided into three phases: analysis, decision making, and
implementation. This is accompanied by a feedback loop, as shown in Fig. 1. A
detailed description of the process is given in chapters three to five.

2.2 Validation Methodology

Since the paper of Hoffmann and Lasch (2020) provides a literature-based theoretical
framework for PdM implementation, a practical realization is missing as yet. This
paper aims to research the feasibility of the given model to be applied in a practical
Case Study Validation of a Predictive Maintenance Implementation … 51

Analysis Decision Making Implementation

(t) Technical Analysis

(implementation Software
costs) (data warehouse,
RUL prediction
(s) PdM Decision toolbox)
(t) Analysing Main- (cost-benefit (t) PdM (t)
tenance Management (t)
analysis, make or Techniques Implementation
buy decision)
Hardware
(t) Process Analysis (sensors, edge
(implementation devices)
benefits) Service Provider

(s) strategic (t) tactical information feedback-loop

Fig. 1 PdM implementation roadmap according to Hoffmann and Lasch (2020)

industrial use case. Therefore, a co-validation based on Yin and McKay (2018) was
conducted. The validation team consists of maintenance experts in an automotive
industry company as model users and maintenance researchers. This approach was
chosen to enable an interdisciplinary validation process and gain new findings for
the model based on the practical implementation.
A case study evaluation approach based on the paper of Offermann et al.
(2009) was chosen to validate the roadmap. Therefore, the analysis and decision-
making phase was conducted in cooperation with the maintenance department of
the mentioned automotive industry company. The validation of the implementation
phase is based on a maintenance expert interview.
The considered assembly line consists of eleven maintenance objects shown in
Table 3, whereas most perform joining processes. The remaining machines execute
handling and testing processes. In the following, the proposed steps according to the
roadmap of Hoffmann and Lasch (2020) are performed.

3 Analysis Phase

3.1 Maintenance Management

The initial step starts with the analysis of the existing maintenance management.
Therefore, the current mix of applied maintenance strategies and their strategic roles,
the structure of the maintenance costs, and failure causes are examined. Applied key
performance indicators and the existing infrastructure of the information system are
further evaluation criteria.
Own employees carry out most maintenance measures, and only a few of them
are conducted by machine suppliers due to service contracts. The company follows
a preventive time-based maintenance strategy. Therefore, a period of 90 min every
52 M. A. Hoffmann

week and one maintenance shift every month are scheduled. Maintenance activities
are carried out based on a maintenance plan with fixed intervals and work instructions
documented based on an ISO 9001 quality management certification.
The accounting department of the company records direct maintenance costs. One
problem in this context is that those costs are assigned to working stations instead of
classification to the single maintenance objects. The indirect maintenance costs are
not captured. This is caused by the management’s high priority of delivery reliability.
Therefore, the maintenance intensity is high and parts are changed before reaching
the wear limit. However, conclusions about the indirect maintenance costs can be
partly drawn based on the machine downtimes documented by the maintenance
management.
The failure causes are recorded in a detailed manner by the maintenance depart-
ment. The mean time to repair (MTTR) and up- and downtime monitoring are
currently the only implemented KPI considering the maintenance performance.
The overall equipment effectiveness (OEE) is introduced as a relevant performance
indicator.
The IT infrastructure is based on the SAP PM module. The in-house IT department
develops specific solutions for plant condition monitoring. It should be noted that
this condition monitoring captures the overall state (operating or down) of a machine
but not maintenance-specific features such as vibration.

3.2 Process Analysis

The second step is process analysis, which aims to identify crucial steps within an
industrial process that show high potential benefits for implementing PdM. In this
context, the process configurations are of particular interest. Parallel and redundant
production processes with several machines for the same purpose have a lower risk of
a complete process shutdown than serial configurations, where unplanned downtimes
can have devastating impacts on the process’s reliability and safety. Further potentials
of PdM are given at high storage costs for spare parts or production facilities, of which
a failure has value-reducing impacts. High failure frequency, downtimes, and failure
impacts on output and product quality are analyzed criteria within this step too.
Quantification of the analysis criteria as a sub-process is necessary to conduct
the process analysis. Therefore, the expert team defines a weighting of the following
features that are necessary for maintenance purposes.
The direct maintenance costs (MC) contain expenses for maintenance measures
such as parts, personnel costs, and equipment, based on the accounting department’s
documentation. The failure frequency (FF) measures the number of failures of a
machine per year, whereas the downtime (DO) includes the annual cumulated hours
of downtime caused by machine failures. The criterion machine value (MV) measures
the negative impacts of a failure on the overall object value.
Indirect maintenance costs are included in the analysis by considering the influ-
ence on the production output (IO) and product quality (IQ). There are no documented
Case Study Validation of a Predictive Maintenance Implementation … 53

Table 1 Weighted criteria

Nr. Criteria Weight in %
process analysis
1 Maintenance cost (MC) 20
2 Failure frequency (FF) 30
3 Downtime (DO) 30
4 Machine value (MV) 5
5 Influence on output (IO) 5
6 Influence on quality (IQ) 10

values available for those criteria. Therefore, the expert team estimates the values
based on their experience. The weights of the criteria are shown in Table 1.
In this case study, the team considers downtime and failure frequency as the
most relevant criteria, whereas machine value, quality, and output are not of high
priority. Based on the absolute values of each maintenance object’s criteria, a scale
is defined by the team, which allows for an assignment to a value from 1 to 10. An
exemplary scale for the failure frequency is shown in Table 2. Similar scales exist
for the remaining criteria, whereas a higher scale value represents a higher potential
factor. As shown in Table 3, the scale values are weighted and summed up afterwards
to calculate a value for the PdM potential within the process analysis.
Exemplary calculation of the potential factor of object 1:

= 2 ∗ 0.2 + 10 ∗ 0.3 + 7 ∗ 0.3 + 8 ∗ 0.05 + 8 ∗ 0.05 + 8 ∗ 0.1 = 7.1

Table 2 Exemplary scale for FF

Value from to Scale value
0.0 7 1
7.1 14 2
14.1 21 3
21.1 28 4
28.1 35 5
35,1 42 6
42.1 49 7
49.1 56 8
56.1 63 9
63.1 70 10
54 M. A. Hoffmann

Table 3 Values for PdM potential

{
Criteria Object type MC FF DO MV IO IQ
Weight (%) 20 30 30 5 5 10
Object 1 Welding 2 10 7 8 8 8 7.1
Object 2 Welding 1 5 3 8 8 7 4.1
Object 3 Welding 3 7 10 10 8 7 7.3
Object 4 Welding 2 7 10 10 5 6 6.9
Object 5 Welding 2 7 10 10 5 6 6.9
Object 6 Forming 1 1 1 4 10 4 1.9
Object 7 Welding 2 7 3 10 10 4 4.8
Object 8 Rework 1 1 1 2 9 5 1.9
Object 9 Handling robot 10 4 9 6 10 2 6.9
Object 10 Leak test 1 3 5 4 10 9 4.2
Object 11 Optical testing 1 1 1 8 1 10 2.3

3.3 Technical Analysis

The technical analysis concludes the analysis phase, aiming to determine the effort
and cost of PdM implementation. If condition monitoring systems are already imple-
mented at the maintenance object, an extension to a predictive system can be installed
with comparatively low effort. Also, some existing sensors in the system might be
used for PdM purposes. If those two prerequisites are not fulfilled, retrofitted sensors
can compensate for that shortcoming. In this context, it must be analyzed whether
such an installation of sensors and communication can be done with a moderate
effort.
At first, the relevant criteria to estimate the effort of implementing PdM are
defined. Usable sensor data (US), which enable RUL prediction, allow minimal
implementation effort. If those are not available, sensor costs (SC) and the usable
installation space (IS) must be considered. The data transfer (DT) expresses the
efforts to extend the IT infrastructure for linking new sensors. Retrofitting (RF)
contains necessary hardware changes to fit new sensors to an existing machine.
Detecting measuring faults (MF) can be necessary when false signals are delivered
due to interferences in the system or damaged sensors.
The priority and weights of the individual criteria are defined by the maintenance
experts and shown in Table 4. In this case, the installation space and possible measure-
ment faults are considered the most critical aspects. The scales are configured similar
to those in Sect. 3.2, whereas a higher scale value represents a higher effort.
The evaluation considering the implementation effort for PdM of the eleven main-
tenance objects is done for every criterion and summed up similarly to the process
analysis. The results can be seen in Table 5.
Case Study Validation of a Predictive Maintenance Implementation … 55

Table 4 Weighted criteria technical analysis

Nr. Criteria Weight in %
1 Usable sensors (US) 20
2 Sensor costs (SC) 5
3 Installation space (IS) 30
4 Data transfer (DT) 5
5 Retrofitting (RF) 10
6 Measuring faults (MF) 30

Table 5 Values for PdM implementation effort

{
Criteria US SC IS DT RF MF
Weight (%) 20 5 30 5 10 30
Object 1 10 5 9 1 5 8 7.9
Object 2 10 5 9 1 9 1 6.2
Object 3 7 7 2 1 2 2 3.2
Object 4 10 3 5 1 6 6 6.1
Object 5 10 3 5 1 6 6 6.1
Object 6 10 5 3 1 3 10 6.5
Object 7 10 8 5 1 5 3 5.4
Object 8 10 5 5 1 3 10 7.1
Object 9 10 8 3 1 5 6 5.7
Object 10 10 5 3 1 3 10 6.5
Object 11 10 10 8 1 8 10 8.8

4 Decision-Making Phase

4.1 Decision-Making Matrix

The second phase addresses decision-making. Therefore a 3 × 3 decision support

matrix is proposed by Hoffmann and Lasch (2020), in which the results of the process
analysis (PdM potential) and technical analysis (implementation effort) are filled in.
Depending on the category the maintenance object is assigned to, recommendations
about implementing PdM are given based on a cost–benefit consideration. The calcu-
lated values for every maintenance object considering PdM potential (Sect. 3.2) and
implementation effort (Sect. 3.3) are filled into the decision-making matrix in Fig. 2.
Maintenance object 3 shows a high potential for PdM with a comparatively low
effort. Therefore, the implementation is highly recommended for this machine. The
main reason is a high number of failures and long downtimes. Furthermore, a missing
redundancy makes this machine crucial for the whole production process. On the
56 M. A. Hoffmann

10,0
Fig. 2 Decision-Making
matrix
high A B C
1
9 4
3 5
6,7

Potential
medium B C D
7 10
3,3
2

11

low C D 6 8 D

0,0
0,0 low 3,3 medium 6,7 high 10,0

Effort

other hand, necessary sensors for condition monitoring are already implemented,
enabling a cost-efficient use of the already continuously measured values. The main-
tenance objects 4, 5, and 9 are assigned to class B, which implies an advisable
implementation of PdM. The potential benefits are very high, but the effort is signifi-
cantly higher than for the object in the first category. Machines 4 and 5 have a parallel
configuration and can be described as redundant. However, they have a high failure
frequency, and cost-intensive spare parts are necessary. Object 9 is a handling robot
and, therefore, a critical part of the process since a breakdown can also lead to a
complete shutdown.
A clear recommendation cannot be given about PdM implementation for those
objects assigned to class C. However, a selective consideration is necessary to avoid
overfitting. In this case study, maintenance objects 1, 2, 7, and 10 are classified
in this category. Machine 1 shows high failure frequency and downtimes. Further-
more, a failure has comparatively high impacts on the output and product quality.
Consequently, a PdM consideration seems useful for better process quality, but the
implementation is not recommended in the first step under cost-considerations. The
remaining objects of class C are characterized by medium potential and effort values,
which leads to a lower priority for the PdM implementation.
Objects 6, 8, and 11 have an inconvenient effort-potential ratio and are therefore
not suitable for monitoring by a predictive maintenance strategy since this would
lead to adverse effects considering the overall maintenance costs. Following this
decision-making process, the company’s maintenance management decided to start
a pilot project to implement PdM for objects 3 and 9.
Case Study Validation of a Predictive Maintenance Implementation … 57

4.2 Service Provider and PdM Techniques

A strategic decision about outsourcing specific PdM tasks is included in the decision-
making phase. To gain further know-how in this field, the expert team makes a
strategic decision against outsourcing maintenance tasks in this context. The manage-
ment assesses that the existing maintenance team can execute all relevant mainte-
nance measures since there is no significant change in operational maintenance tasks
to be expected.
Following the proposed roadmap of Hoffmann and Lasch (2020), the decision
about the RUL prediction method is to be made at this point of the process. Deviating
from this recommendation, the maintenance team decides to postpone the decision
about the PdM techniques to be used since the choice depends on the used software.
The planning steps for this implementation pilot project are now explained based on
the maintenance team expert interview.

5 Implementation Phase

5.1 Hardware Implementation

Within the third phase, the actual implementation is conducted. Therefore, the
required software and hardware must be selected simultaneously for a secure and
reliable data transfer and cannot be considered separately from each other.
The most critical task in hardware implementation is choosing the relevant features
to be measured to ensure high data quality for condition monitoring. For those main-
tenance objects where sensors for condition monitoring are installed already (e.g.,
object 3), an evaluation is necessary whether the monitored features can be used
for PdM purposes. To ensure this, a consultation with the machine manufacturer is
required. This is also relevant for machines without any installed sensors for condition
monitoring purposes yet to gain information about relevant thresholds that indicate
the end of life for wearing parts and necessary maintenance measures. The company’s
sensor supplier and expert is consulted to assess which sensors to select for values
to be measured and how to position them.

5.2 Software Implementation

Depending on the existing information system infrastructure and the extent of the
expected amount of data, PdM tasks can either be implemented in an existing ERP
system or cloud computing solutions combined with IoT-based sensors might be
more suitable due to more scalable and flexible options.
58 M. A. Hoffmann

In this particular case, two options are considered. The first variant is a predic-
tive maintenance toolbox offered by the existing ERP system provider as an add-on.
Therefore, pre-defined algorithms predict the RUL and schedule necessary mainte-
nance measures. This variant reduces the effort of developing algorithms such as
artificial intelligence solutions and can be implemented rapidly without extensive
knowledge in data engineering.
The second variant contains an extensive in-house development of software for
PdM purposes. This requires more effort in the field of data evaluation. Based on the
framework of Lei et al. (2018), the process of health prognosis evaluation is split into
four steps: data acquisition, health indicator construction, health state division, and
RUL prediction. An own software solution requires superior know-how in feature
engineering and the configuration of artificial intelligence algorithms frequently used
for RUL prediction.

5.3 Implementation in Maintenance Management

A predictive strategy might require adaptions considering the operational targets

and organizational structures. Therefore, the implementation phase is concluded by
introducing this strategy into maintenance management. To realize the full potential
of PdM, a holistic approach that considers production planning and purchasing claims
is proposed.
Since the pilot project does not affect the whole production line, only minor
changes are expected considering the maintenance management. The weekly
timespan for maintenance activities is still used for inspections at the other objects.
It is not clear yet, whether the maintenance measures for objects 3 and 9 are still
executed in the monthly maintenance shift, or they can be done during the operation
time due to the excellent planning opportunities based on the estimated RUL.
Regarding software-variant 1, synergies are expected by including other ERP
modules such as production scheduling or material management, which can purchase
needed spare parts according to the demand.
A feedback loop is planned to be installed to improve the implemented
PdM strategy, aiming for constant optimization and adaptation of the mainte-
nance processes. The feedback information will be embedded into the continual
improvement process of the quality management system.

6 Conclusion

Many experts see a contribution of predictive maintenance to more efficient and

reliable production processes in the industrial environment. However, the number of
successful implementations is still low. Based on a practical case study and expert
Case Study Validation of a Predictive Maintenance Implementation … 59

consultation, this article gives industrial companies a structured suggestion to start

making their own experiences and gain know-how in the field of PdM.
The validation study shows that the underlying framework applies to a structured
PdM implementation process with few adjustments to the initial version.
In the considered case, the analysis revealed two maintenance objects with high
potential and moderate implementation effort, which are feasible for the maintenance
management department to start a PdM pilot project.
Regarding RQ1, the expert team recommends implementing a sub-process for
quantifying the analysis criteria in phase one. Besides this, the co-validation team sees
all relevant criteria included in the framework to give a structured recommendation
for action.
Regarding the execution and order of the proposed steps, the decision-making of
the PdM techniques is recommended to be shift into the implementation phase since
this is dependent on the used software solution, which answers RQ2.
The validation study is done with the limitation that the practical process is
performed for the first two phases in the company, whereas the evaluation of the
third phase is based on maintenance expert consultation. Further research in prac-
tical PdM implementation is still necessary to unfold the advantages for producing
companies. Furthermore, plug-and-play solutions are required that offer practitioners
the opportunity to get started without high effort.

Acknowledgements The author would like to thank the participants of the research institution and
the automotive company for their contribution to the validation study. Furthermore, he would like
to express his thanks to the reviewers and editor who have contributed to the improvement of this
research paper.

References

Ansari F, Glawar R, Nemeth T (2019) PriMa: a prescriptive maintenance model for cyber-physical
production systems. Int J Comput Integr Manuf 32(4–5). Taylor & Francis. https://doi.org/10.
1080/0951192X.2019.1571236
CEN (2017) DIN EN 13306: maintenance—maintenance terminology. Beuth Verlag, Berlin
Erbe H et al (2005) Infotronic technologies for e-maintenance regarding the cost aspects. In: IFAC
proceedings volumes (IFAC-PapersOnline). IFAC. https://doi.org/10.3182/20050703-6-cz-1902.
01384
European Commission (2021) Updating the 2020 new industrial strategy: building a stronger single
market for Europe’s recovery. https://ec.europa.eu/info/sites/default/files/communication-indust
rial-strategy-update-2020_en.pdf
Feng Q, Shanthikumar JG (2018) How research in production and operations management may
evolve in the era of big data. Prod Oper Manag 27(9). https://doi.org/10.1111/poms.12836
Haarman M et al (2018) Beyond the hype: PdM 4.0 delivers results. Predictive Maintenance 4.0.
https://www.pwc.de/de/industrielle-produktion/pwc-predictive-maintenance-4-0.pdf
Hoffmann MA, Lasch R (2020) Roadmap for a successful implementation of a predictive mainte-
nance strategy. In: Smart and sustainable supply chain and logistics—trends, challenges, methods
and best practices. EcoProduction (Environmental issues in logistics and manufacturing).https://
doi.org/10.1007/978-3-030-61947-3_29
60 M. A. Hoffmann

Lei Y et al (2018) Machinery health prognostics: a systematic review from data acquisition to RUL
prediction. In: Mechanical systems and signal processing. Elsevier Ltd, p 104.https://doi.org/10.
1016/j.ymssp.2017.11.016
Mack O et al (2015) Managing in a VUCA world.https://doi.org/10.1007/978-3-319-16889-0
Offermann P et al (2009) Outline of a design science research process. In: Proceedings of the
4th international conference on design science research in information systems and technology,
DESRIST‘09, June 2014. https://doi.org/10.1145/1555619.1555629
Prajapati A, Bechtel J, Ganesan S (2012) Condition based maintenance: a survey. J Qual Maint Eng
18(4). https://doi.org/10.1108/13552511211281552
Yin C, McKay A (2018) Model verification and validation strategies and methods: an application
case study. In: ISCIIA and ITCA 2018—8th international symposium on computational intelli-
gence and industrial applications and 12th China-Japan international workshop on information
technology and control applications, May 2019
 Modeling, Simulation and Optimization
of Supply Chain, Production and Logistics
Operations
The Simulation for the Ideal Optimum
Site Model in the Actual Distance Method

Angela Yu Yen Chen and Yutaka Karasawa

1 Introduction

Our research is based on the real distance data for selecting optimum sites, there are
five types are taken consideration, ➀ Ideal type optimum site (without consideration
of available or unavailable locations), ➁ Present location plus α type optimum site
(present center plus one to four optimum delivery sites), ➂ Available location type
optimum site, ➃ Present location plus α available location type optimum site (present
center plus one to four available optimum locations), and ➄ Present location type
only, moreover, subject to those types, we set up locations from one, two, three, four,
to five places, for each type of the optimum sites selection model, there are 20 places
of the optimum places are selected, the reason for a maximum of five locations is that
the area only covers approximately 50 km in a small area of Japan, for integrating
discussion to find the best solution for rational problem and real problem 1 (Fig. 1).
As this is a new research project, we tried to divide the region into 35 areas by zip
code, but it did not really work, then we divided into 38 places for the delivery zone
of the research, by measuring the actual distances between the distribution center
and customer’s place, so-called the map 38 places of the actual distance method.
The basic approach of the whole research is based on analyzing the master data
from the distribution center, the area of the objects and analysis of delivery distribu-
tion properties in the map, the measurement of the center of gravity of the delivery

A. Y. Y. Chen (B)
Department of Marketing and Logistics, China University of Technology, Taipei City, Taiwan
(R.O.C.)
e-mail: [email protected]
Y. Karasawa
Department of Industrial Engineering, Kanagawa University, Yokohama, Japan
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 63

P. Golinska-Dawson et al. (eds.), Smart and Sustainable Supply Chain
and Logistics—Challenges, Methods and Best Practices, EcoProduction,
https://doi.org/10.1007/978-3-031-15412-6_6
64 A. Y. Y. Chen and Y. Karasawa

Fig. 1 Types of optimum locations simulations

zone and inspection of the error (the error inspection is carried out only for the approx-
imation distance method), as shown in Fig. 2. The basic data for the simulation is the
preparation for the distant table and lk (carrying one liter for one kilometer) table,
the implement for the simulations of the optimum sites are based on the ideal type by
actual distance method in the map (using the Google Maps Distance Matrix API), the
dividing the cost for delivery centers are based on estimation the delivery cost and
inventory cost and final evaluation for the optimum site of {lk, then our research
continues inspecting for the optimum site, selecting the best order 2, as shown in
Fig. 3.

2 Research Contents

2.1 Theory of Segmentation by City and Village

2.1.1 The Calculation of Gravity Points

The first step of the process of implementing simulation for an optimum site is to
decide how to divide the area as shown in the Fig. 4. After trying to place 500
customers’ addresses of one group by mapping, and making sure that there will be
one portion of land without any disruption, such as rivers and mountains, and the
range will not be too distant. However, some of the customers’ master data did not
meet the JIS (Japanese Industrial Standards) code on the map, moreover, some of the
addresses could not be found in the areas, so we modified the mistyped and mapped
again the area as 38 locations for selecting the optimum sites.
The Simulation for the Ideal Optimum Site Model in the Actual Distance … 65

Fig. 2 Master data and the types of simulation

Fig. 3 The concept of research process

66 A. Y. Y. Chen and Y. Karasawa

Fig. 4 The preparation of the customers’ distribution map (Remark there is not any enclave within
the land)

General speaking, taking the whole Japan areas, then you should divide the area
by prefectures, since our research is only focused on the small area in Japan, so we
group the areas by postcode for calculating the optimum sites.
We decided to use the Yahoo Open Local Platform (YOLP) system/software,
implementing geocoding, with the information from 13,012 destinations by longitude
and latitude, by taking the actual distance, though there were several incorrect data
which were modified by hand. Since there are 2500 times per day limitation of
using the Google Maps API. Therefore, the grouping was completed by mapping the
customers’ information, and each group included the information of longitude and
latitude of the customers’ addresses and delivery amount (l) are introduced for the
center of gravity coordinates (longitude and latitude).
The centers of gravity coordinates of each area are as follows:
{n {n
i=1 Ti x i Ti yi
x = {n , y = {i=1
n
i=1 Ti i=1 Ti

Remarks:
x i : longitude of customers’ address
yi : latitude of customers’ address
T i : delivery amount (l)
n: customers’ numbers in each area
The Simulation for the Ideal Optimum Site Model in the Actual Distance … 67

2.1.2 The Considerations and the Quantity of Dividing into 38 Areas

Trying to set up a standard about 500 customers’ addresses as one group by mapping,
and making sure that there will be one group, even the road divided into two parts, and
the size of the area not being too distant. However, some of the customers’ master data
did not meet the JIS (Japanese Industrial Standards) code on the map, moreover, from
the code, some of the addresses could not be found in the areas, finally, we adjusted
as 38 areas on the map, the detailed information of the 38 locations is shown in Table
1. Customers by cities, from P1 to P23 are Toyogawa-shi (city), from P24 to P30
are Toyahashi-shi, from P31 to P34 belongs to Sinsiro-shi, and Kita-sitara-gun is on
P35, from P36 to P38 are Okazaki-shi, as shown in Figs. 5, 6 and 7. We also looked
at the number of customers, and there are 13,012 households, with the total amount
of delivery being 1,146,760 L, which included 30 bottles of 10 L size, 8193 bottles of
20 L size, and 19,652 bottles of 50 L size. After deciding the 38 locations the actual
distance method is used for calculating the distance between the two places of these
38 areas, therefore, we used the master data to calculate the figure of the customers,
the demanding amount, and the distance of each location. Finally, we produced a
demand-distance table for simulation.
The solution of our research is to calculate the total cost (delivery fee plus center
operation fee) from the master data to produce the lk (unit price for carrying one
liter for one kilometer) and it is an appropriate tentative perspective from the model
by each zone. Then, the delivery expenses were introduced, finally, to put those
expenses into the delivery centers that were selected, then the simulation for the
optimum sites were analyzed. There are one to five places for consideration, the base
on the method of the present and unit price by zone, namely, to simulate by four
types of the optimum site model and the present center.

2.2 Preparation of the Distant Table and lk Table

The method of preparing the distance matrix tables between the cities and villages,
we used the actual distance and the approximate distance from the map, here we
adopted the actual distance method for simulation, moreover, for preparing the lk
matrix Table 4, which is based on the distance table of 38 areas, as shown in Table 2,
then the basic document for simulation is completed, as shown in Table 3. General
speaking, when it comes to selecting an optimum site model like a gravity model,
people usually used the tk (tone/kilometer) matrix method to actually calculate the
results of dividing, this research has more or less, 20 thousand customers scattered
over the eastern part of the Aichi Prefecture and one western part of the Shizuoka
prefecture in Japan. Therefore, we use the lk instead of the tk as the unit 3.
68 A. Y. Y. Chen and Y. Karasawa

Table 1 T38 areas on the map

No. Name of cities No. of user
1 Toyokawa-shi 8065
2 Toyohashi-shi 2765
3 Sinshiro-shi 1380
4 Kitashitara-gun 374
5 Okazaki-shi 391
6 Toyota-shi 30
7 Hamamatsu-shi (Shizuoka) 7
Total 13,012
Location Cities No. of Customers (%) Delivery Delivery
customers amount (l) amount (%)
P1 Toyokawa-shi ➀ 142 1.1 11,050 1.0
P2 Toyokawa-shi ➁ 667 5.1 54,370 4.7
P3 Toyokawa-shi ➂ 471 3.6 36,190 3.2
P4 Toyokawa-shi ➃ 346 2.7 30,390 2.7
P5 Toyokawa-shi ➄ 217 1.7 20,490 1.8
P6 Toyokawa-shi ➅ 129 1.0 9150 0.8
P7 Toyokawa-shi ➆ 519 4.0 44,140 3.8
P8 Toyokawa-shi ➇ 764 5.9 55,850 4.9
P9 Toyokawa-shi ➈ 316 2.4 25,050 2.2
P10 Toyokawa-shi ➉ 313 2.4 29,170 2.5
P11 Toyokawa-shi o
11 290 2.2 27,280 2.4
P12 Toyokawa-shi o
12 165 1.3 24,050 2.1
P13 Toyokawa-shi o
13 183 1.4 14,850 1.3
P14 Toyokawa-shi o
14 198 1.5 27,580 2.4
P15 Toyokawa-shi o
15 415 3.2 38,070 3.3
P16 Toyokawa-shi o
16 420 3.2 36,780 3.2
P17 Toyokawa-shi o
17 212 1.6 18,540 1.6
P18 Toyokawa-shi o
18 624 4.8 52,570 4.6
P19 Toyokawa-shi o
19 259 2.0 19,450 1.7
P20 Toyokawa-shi o
20 278 2.1 29,730 2.6
Location Cities No. of Customers (%) Delivery Delivery
customers amount (l) amount (%)
P21 Toyokawa-shi 21 209 1.6 20,250 1.8
P22 Toyokawa-shi 22 661 5.1 51,010 4.4
P23 Toyokawa-shi 23 267 2.1 23,850 2.1
P24 Toyohashi-shi ➀ 299 2.3 23,900 2.1
(continued)
The Simulation for the Ideal Optimum Site Model in the Actual Distance … 69

Table 1 (continued)
Location Cities No. of Customers (%) Delivery Delivery
customers amount (l) amount (%)
P25 Toyohashi-shi ➁ 359 2.8 33,940 3.0
P26 Toyohashi-shi ➂ 276 2.1 32,600 2.8
P27 Toyohashi-shi ➃ 608 4.7 49,380 4.3
P28 Toyohashi-shi ➄ 398 3.1 39,060 3.4
P29 Toyohashi-shi ➅ 303 2.3 17,800 1.6
P30 Toyohashi-shi ➆ 522 4.0 63,980 5.6
P31 Sinshiro-shi ➀ 461 3.5 54,900 4.8
P32 Sinshiro-shi ➁ 360 2.8 36,950 3.2
P33 Sinshiro-shi ➂ 441 3.4 32,690 2.9
P34 Sinshiro-shi ➃ 142 1.1 5660 0.5
P35 Kitashitara-gun 357 2.7 15,070 1.3
P36 Okazaki-shi ➀ 233 1.8 23,830 2.1
P37 Okazaki-shi ➁ 126 1.0 9490 0.8%
P38 Okazaki-shi ➂ 62 0.5 7650 0.7
Total 13,012 100.0 1,146,760 100.0

Fig. 5 P36 Okazaki-shi, 233

customers

Fig. 6 P37 Okazaki-shi, 126

customers
70 A. Y. Y. Chen and Y. Karasawa

Fig. 7 P38 Okazaki-shi, 62

The method of actual distance is taken from the Yahoo! Open Local Platform
(YOLP) system/software, implementing geocoding,1 with the information from
13,012 destinations by latitude and longitude, and there were several incorrect data
which were modified by hand. Also, it was not possible to use the Google Maps API,
since there is a limitation of 2500 times per day. Our research, therefore, decided to
use YOLP at this time. As previously mentioned, the grouping was completed by
mapping the customers’ information, and each group included the information of the
customers, and the latitude and longitude from the delivery amount (l), and thereby
calculated the gravity coordinates (latitude and longitude).
The method of approximate distance is to assume that the radius of the Earth, r =
6378.137 km (equatorial radius) is a ball shape, and the distance d is measured from
spot A (latitude × 1, longitude y1) to spot B (latitude × 2, longitude y2), then the
program is as follows;

d = r arccos{sin y1 sin y2 + cos y1 cos y2 cos(x2 − x1 )}

Unfortunately, the Earth is an「oval shape」, and consequently it caused a toler-

ance. Therefore, the method of drawing up the matrix distance between the two points
of the cities should be considerate of both actual distance and approximate distance
on the map (Fig. 8).

2.3 The Optimum Location Simulation

For finding a realistic solution for the horizontal optimum location simulation, we
devised a new type of model so that we can support realistic management deci-
sions and decision-making, not just using the number of conventional locations. In
addition, the multi-step algorithm is based on a wide variety of algorithms such as

1Geocoding is the process of transforming a description of a location, such as a pair of coordinates,

an address, or a name of a place, to a location on the earth’s surface.
Table 2 Table of the actual distance method on the map for 38 areas (example)
Location Arrival point P1 P2 P3 P4 P5 P6
Latitude 34.8629707 34.84561071 34.8337699 34.81672573 34.85331179 34.80561814
Start Point Latitude Longitude 137.30131 137.3194779 137.3255714 137.3128 137.3394865 137.3208462
P1 34.8629707 137.3013095 0 2.877 4.417 8.045 4.58 9.279
P2 34.84561071 137.3194779 2.877 0 1.955 4.885 3.509 5.478
P3 34.8337699 137.3255714 4.057 1.595 0 3.429 3.677 4.022
P4 34.81672573 137.3128 7.208 4.746 3.11 0 6.915 1.979
P5 34.85331179 137.3394865 4.806 3.076 3.251 7.059 0 7.567
P6 34.80561814 137.3208462 8.332 5.87 4.234 2.379 7.398 0
P7 34.83734598 137.3401707 5.617 2.791 2.239 5.406 2.195 5.915
P8 34.80326089 137.3489893 9.406 6.581 5.469 5.089 7.739 3.641
P9 34.8003621 137.3620273 9.676 6.851 5.693 5.684 8.009 4.236
P10 34.82064251 137.3580456 8.52 5.694 4.725 6.071 4.786 6.579
P11 34.81644871 137.3749543 9.928 7.102 6.576 7.922 6.146 6.589
P12 34.80506781 137.3851106 12.144 9.319 8.146 9.697 8.541 7.149
P13 34.81754503 137.3906428 10.825 7.999 8.927 10.273 7.464 7.596
P14 34.81837196 137.4069214 14.861 12.035 10.862 12.414 8.759 9.865
The Simulation for the Ideal Optimum Site Model in the Actual Distance …

P15 34.82642438 137.3925224 11.585 7.922 7.988 9.334 6.987 9.843

P16 34.83316028 137.3642903 8.491 4.828 4.894 6.24 3.893 6.749
P17 34.83125618 137.402365 13.11 8.851 8.917 10.263 8.512 10.771
P18 34.83767594 137.3928127 11.811 7.651 7.717 9.063 7.213 9.571
P19 34.83295012 137.4176642 14.632 12.058 10.604 11.95 10.034 11.668
(continued)
71
Table 2 (continued)
72

Location Arrival point P1 P2 P3 P4 P5 P6

Latitude 34.8629707 34.84561071 34.8337699 34.81672573 34.85331179 34.80561814
Start Point Latitude Longitude 137.30131 137.3194779 137.3255714 137.3128 137.3394865 137.3208462
P20 34.84866582 137.3725186 9.076 5.972 5.419 7.62 4.478 8.128
P21 34.8539671 137.4057124 13.294 10.6 10.048 12.248 8.696 12.756
P22 34.853703 137.4269815 15.261 12.567 12.014 14.215 10.663 14.723
P23 34.87139649 137.4329877 15.911 13.217 12.665 14.865 11.313 15.373
P24 34.77821126 137.3444784 13.56 9.272 7.634 6.335 10.604 4.887
P25 34.78638958 137.3855037 12.668 9.843 8.67 10.221 11 7.625
P26 34.77912055 137.4222122 17.757 14.931 13.758 15.31 13.511 12.714
P27 34.79929553 137.4449543 18.749 15.924 14.751 16.302 15.923 13.706
P28 34.76418687 137.4264511 17.802 14.976 13.803 15.355 16.134 12.759
P29 34.83445434 137.4547471 19.045 18.39 17.217 18.769 14.447 16.173
P30 34.76605784 137.3712467 15.086 12.26 11.087 9.741 13.418 8.293
P31 34.87994115 137.4812778 21.92 19.226 18.674 20.874 17.322 21.383
P32 34.90793958 137.5159418 25.74 23.046 22.494 24.694 21.142 25.202
P33 34.93778212 137.5777741 33.815 31.121 30.569 32.769 29.217 33.277
P34 35.00936674 137.5339967 49.003 46.309 45.757 47.957 44.405 48.465
P35 35.07888126 137.6921871 59.772 57.077 56.525 58.725 55.173 59.234
P36 34.91646628 137.2189389 10.237 13.145 14.74 17.926 14.136 19.159
(continued)
A. Y. Y. Chen and Y. Karasawa
Table 2 (continued)
Location Arrival point P1 P2 P3 P4 P5 P6
Latitude 34.8629707 34.84561071 34.8337699 34.81672573 34.85331179 34.80561814
Start Point Latitude Longitude 137.30131 137.3194779 137.3255714 137.3128 137.3394865 137.3208462
P37 34.92218539 137.3005012 10.905 13.812 15.407 18.593 14.526 19.826
P38 34.99872367 137.2921787 24.531 27.439 29.034 32.22 28.153 33.453
Location Arrival point P7 P8 P9 P10 P11 P12 P13
Latitude 34.83734598 34.80326089 34.8003621 34.82064251 34.81644871 34.80506781 34.81754503
Start point Latitude Longitude 137.3401707 137.3489893 137.3620273 137.3580456 137.3749543 137.3851106 137.3906428
P1 34.8629707 137.3013095 5.411 9.214 9.393 8.17 9.302 12.061 10.852
P2 34.84561071 137.3194779 2.791 6.594 6.773 5.55 6.682 9.441 8.232
P3 34.8337699 137.3255714 1.92 5.152 5.331 4.237 6.206 7.999 8.209
P4 34.81672573 137.3128 5.168 5.089 7.034 5.941 7.91 9.703 9.913
P5 34.85331179 137.3394865 2.195 7.089 7.222 5.03 6.162 8.755 7.712
P6 34.80561814 137.3208462 5.651 3.641 4.236 5.219 6.618 7.188 7.596
P7 34.83734598 137.3401707 0 5.437 5.942 3.378 4.51 7.103 6.06
P8 34.80326089 137.3489893 5.293 0 2.013 3.107 3.355 5.087 5.297
P9 34.8003621 137.3620273 5.563 2.013 0 3.527 2.874 3.147 3.555
The Simulation for the Ideal Optimum Site Model in the Actual Distance …

P10 34.82064251 137.3580456 3.04 3.118 2.887 0 2.421 4.237 4.448

P11 34.81644871 137.3749543 4.51 3.355 2.903 2.097 0 1.888 2.039
P12 34.80506781 137.3851106 6.795 4.85 3.108 4.001 1.701 0 2.352
(continued)
73
Table 2 (continued)
74

Location Arrival point P7 P8 P9 P10 P11 P12 P13

Latitude 34.83734598 34.80326089 34.8003621 34.82064251 34.81644871 34.80506781 34.81754503
Start point Latitude Longitude 137.3401707 137.3489893 137.3620273 137.3580456 137.3749543 137.3851106 137.3906428
P13 34.81754503 137.3906428 5.827 5.373 3.555 4.448 1.784 2.066 0
P14 34.81837196 137.4069214 7.123 7.567 5.825 6.717 4.376 3.054 2.29
P15 34.82642438 137.3925224 5.578 5.819 5.132 4.393 2.757 3.643 1.671
P16 34.83316028 137.3642903 2.484 4.615 4.384 1.694 2.414 4.303 3.964
P17 34.83125618 137.402365 6.507 6.627 5.658 5.792 3.565 4.413 2.368
P18 34.83767594 137.3928127 5.307 7.513 6.114 4.592 3.225 5.939 2.824
P19 34.83295012 137.4176642 9.714 9.37 7.627 8.52 6.179 4.857 4.534
P20 34.84866582 137.3725186 3.627 6.997 6.766 4.076 4.744 6.632 5.76
P21 34.8539671 137.4057124 8.255 12.278 9.138 7.804 6.438 8.215 5.756
P22 34.853703 137.4269815 10.222 12.395 10.653 8.694 7.372 7.883 7.56
P23 34.87139649 137.4329877 10.873 14.895 12.709 11.282 10.441 9.939 9.616
P24 34.77821126 137.3444784 8.857 4.8 4.028 7.173 6.558 6.087 7.201
P25 34.78638958 137.3855037 8.555 5.402 3.65 6.332 3.987 2.548 4.677
P26 34.77912055 137.4222122 12.799 10.491 8.738 10.005 7.663 6.342 6.449
P27 34.79929553 137.4449543 13.792 11.483 9.731 10.998 8.656 7.334 7.034
P28 34.76418687 137.4264511 13.688 10.536 8.783 11.371 9.029 7.707 7.815
P29 34.83445434 137.4547471 14.127 13.95 12.197 13.464 11.122 9.8 9.5
(continued)
A. Y. Y. Chen and Y. Karasawa
Table 2 (continued)
Location Arrival point P7 P8 P9 P10 P11 P12 P13
Latitude 34.83734598 34.80326089 34.8003621 34.82064251 34.81644871 34.80506781 34.81754503
Start point Latitude Longitude 137.3401707 137.3489893 137.3620273 137.3580456 137.3749543 137.3851106 137.3906428
P30 34.76605784 137.3712467 10.972 6.2 6.067 8.749 6.996 5.558 7.687
P31 34.87994115 137.4812778 16.882 18.579 16.837 15.31 15.388 14.067 13.744
P32 34.90793958 137.5159418 20.702 23.601 21.859 21.112 17.554 19.089 18.766
P33 34.93778212 137.5777741 28.777 30.321 28.578 27.572 27.13 25.808 25.485
P34 35.00936674 137.5339967 43.965 47.987 46.758 44.374 43.533 43.987 43.664
P35 35.07888126 137.6921871 54.733 56.277 54.535 53.529 53.086 51.764 51.442
P36 34.91646628 137.2189389 15.291 19.094 19.273 18.05 19.182 21.941 20.732
P37 34.92218539 137.3005012 16.566 19.762 19.94 18.717 19.849 22.609 21.399
P38 34.99872367 137.2921787 30.193 33.388 33.567 32.344 33.476 36.235 35.026
Location Arrival point P14 P15 P16 P17 P18 P19 P20
Latitude 34.818372 34.82642438 34.83316028 34.83125618 34.83767594 34.83295012 34.84866582
Start point Latitude Longitude 137.406921 137.3925224 137.3642903 137.402365 137.3928127 137.4176642 137.3725186
P1 34.8629707 137.3013095 11.965 10.896 7.917 13.012 11.803 14.505 9.01
P2 34.84561071 137.3194779 9.346 8.276 5.297 10.541 9.331 12.033 6.116
The Simulation for the Ideal Optimum Site Model in the Actual Distance …

P3 34.8337699 137.3255714 10.8 7.815 4.368 10.081 8.871 11.573 5.655

P4 34.81672573 137.3128 12.504 9.491 6.513 10.324 9.411 11.904 8.062
P5 34.85331179 137.3394865 8.826 7.317 3.893 8.587 7.377 10.079 4.584
(continued)
75
Table 2 (continued)
76

Location Arrival point P14 P15 P16 P17 P18 P19 P20
Latitude 34.818372 34.82642438 34.83316028 34.83125618 34.83767594 34.83295012 34.84866582
Start point Latitude Longitude 137.406921 137.3925224 137.3642903 137.402365 137.3928127 137.4176642 137.3725186
P6 34.80561814 137.3208462 9.773 8.932 6.996 9.699 9.894 11.357 8.545
P7 34.83734598 137.3401707 7.173 6.104 3.125 8.369 7.159 9.861 3.943
P8 34.80326089 137.3489893 7.55 6.184 4.601 6.755 7.499 9.134 6.983
P9 34.8003621 137.3620273 5.732 4.891 5.163 5.658 6.114 7.316 7.545
P10 34.82064251 137.3580456 6.906 4.825 1.639 5.64 4.537 8.49 4.021
P11 34.81644871 137.3749543 4.606 2.852 2.414 3.692 3.222 5.325 4.143
P12 34.80506781 137.3851106 3.453 3.689 4.115 5.292 5.087 5.764 5.843
P13 34.81754503 137.3906428 2.341 1.601 3.732 2.368 2.824 3.723 5.143
P14 34.81837196 137.4069214 0 2.56 5.027 2.719 3.407 3.077 6.459
P15 34.82642438 137.3925224 2.423 0 3.169 1.414 1.725 3.371 3.655
P16 34.83316028 137.3642903 5.078 3.5 0 4.037 2.898 5.392 2.382
P17 34.83125618 137.402365 2.647 1.344 4.098 0 1.418 1.651 4.469
P18 34.83767594 137.3928127 3.458 1.799 2.898 1.329 0 2.683 3.17
P19 34.83295012 137.4176642 3.091 2.906 5.786 1.651 3.077 0 6.584
P20 34.84866582 137.3725186 6.509 3.896 2.382 4.38 3.17 5.735 0
P21 34.8539671 137.4057124 6.449 4.732 6.111 4.11 3.505 4.788 5.096
P22 34.853703 137.4269815 6.117 5.884 7 4.629 4.292 3.13 6.775
(continued)
A. Y. Y. Chen and Y. Karasawa
Table 2 (continued)
Location Arrival point P14 P15 P16 P17 P18 P19 P20
Latitude 34.818372 34.82642438 34.83316028 34.83125618 34.83767594 34.83295012 34.84866582
Start point Latitude Longitude 137.406921 137.3925224 137.3642903 137.402365 137.3928127 137.4176642 137.3725186
P23 34.87139649 137.4329877 8.173 8.735 9.589 6.93 7.509 5.633 8.879
P24 34.77821126 137.3444784 8.672 8.537 8.755 9.783 10.501 10.256 11.137
P25 34.78638958 137.3855037 4.938 6.014 6.401 6.897 7.615 9.094 8.129
P26 34.77912055 137.4222122 7.124 7.785 9.779 8.818 9.895 8.352 12.947
P27 34.79929553 137.4449543 5.294 6.814 9.281 6.693 8.967 6.041 12.473
P28 34.76418687 137.4264511 8.49 9.151 11.145 10.184 11.261 10.455 14.313
P29 34.83445434 137.4547471 7.76 7.319 10.199 6.064 7.491 4.565 10.997
P30 34.76605784 137.3712467 7.948 9.023 10.28 9.907 10.625 12.104 12.662
P31 34.87994115 137.4812778 12.301 12.116 13.616 10.86 10.908 9.361 13.392
P32 34.90793958 137.5159418 17.323 17.138 19.418 15.883 14.475 14.383 18.708
P33 34.93778212 137.5777741 24.042 23.857 25.879 22.602 23.17 21.103 25.654
P34 35.00936674 137.5339967 42.221 41.828 42.681 40.781 40.601 39.282 41.971
P35 35.07888126 137.6921871 49.998 49.814 51.835 48.558 49.127 47.059 51.61
P36 34.91646628 137.2189389 21.846 21.298 17.874 22.568 21.358 24.06 18.566
The Simulation for the Ideal Optimum Site Model in the Actual Distance …

P37 34.92218539 137.3005012 22.513 21.689 18.265 22.959 21.749 24.451 18.956
P38ˇ 34.99872367 137.2921787 36.14 35.316 31.892 36.585 35.376 38.078 32.583
(continued)
77
Table 2 (continued)
78

Location Arrival point P21 P22 P23 P24 P25 P26

Latitude 34.8539671 34.853703 34.87139649 34.77821126 34.78638958 34.77912055
Start point Latitude Longitude 137.4057124 137.4269815 137.4329877 137.3444784 137.3855037 137.4222122
P1 34.8629707 137.3013095 12.958 14.639 15.618 12.3 12.462 17.846
P2 34.84561071 137.3194779 10.486 12.168 13.147 9.266 9.843 15.226
P3 34.8337699 137.3255714 10.026 11.707 12.686 7.81 8.4 13.783
P4 34.81672573 137.3128 12.433 14.115 15.093 6.331 10.104 15.487
P5 34.85331179 137.3394865 8.532 10.214 11.192 10.588 10.291 13.505
P6 34.80561814 137.3208462 12.916 14.598 15.576 4.883 7.557 12.782
P7 34.83734598 137.3401707 8.314 9.996 10.974 8.936 9.011 11.853
P8 34.80326089 137.3489893 10.712 12.112 14.19 4.796 5.549 10.813
P9 34.8003621 137.3620273 10.791 10.294 12.398 3.806 3.516 8.741
P10 34.82064251 137.3580456 7.75 8.642 11.228 6.95 5.956 9.964
P11 34.81644871 137.3749543 6.608 8.304 10.612 6.195 3.987 7.614
P12 34.80506781 137.3851106 9.24 10.79 10.846 5.409 2.842 6.048
P13 34.81754503 137.3906428 6.21 6.701 9.007 6.847 4.57 6.449
P14 34.81837196 137.4069214 6.553 6.056 8.16 9.652 4.938 6.565
P15 34.82642438 137.3925224 5.111 6.35 7.786 8.424 6.148 8.457
(continued)
A. Y. Y. Chen and Y. Karasawa
Table 2 (continued)
Location Arrival point P21 P22 P23 P24 P25 P26
Latitude 34.8539671 34.853703 34.87139649 34.77821126 34.78638958 34.77912055
Start point Latitude Longitude 137.4057124 137.4269815 137.4329877 137.3444784 137.3855037 137.4222122
P16 34.83316028 137.3642903 6.111 7.003 9.589 8.444 6.401 9.757
P17 34.83125618 137.402365 4.271 4.629 6.528 9.484 6.917 8.681
P18 34.83767594 137.3928127 3.505 4.295 7.509 9.406 8.443 9.492
P19 34.83295012 137.4176642 4.729 3.13 5.98 9.928 9.094 8.673
P20 34.84866582 137.3725186 5.096 6.778 7.933 11.149 8.731 11.553
P21 34.8539671 137.4057124 0 2.803 5.58 13.286 10.719 12.422
P22 34.853703 137.4269815 2.986 0 3.107 12.954 12.182 11.761
P23 34.87139649 137.4329877 5.58 3.107 0 15.01 14.111 13.69
P24 34.77821126 137.3444784 13.731 13.234 15.338 0 4.971 9.554
P25 34.78638958 137.3855037 12.27 11.683 14.534 5.637 0 6.304
P26 34.77912055 137.4222122 12.265 10.942 13.115 9.854 6.304 0
P27 34.79929553 137.4449543 10.618 8.63 10.539 12.648 7.934 3.453
P28 34.76418687 137.4264511 13.631 13.045 14.853 9.558 6.233 2.016
P29 34.83445434 137.4547471 9.142 6.258 8.065 15.114 10.4 9.316
The Simulation for the Ideal Optimum Site Model in the Actual Distance …

P30 34.76605784 137.3712467 15.28 14.693 17.543 4.287 3.516 5.924

P31 34.87994115 137.4812778 10.106 7.394 6.071 22.182 17.469 16.298
P32 34.90793958 137.5159418 15.41 12.416 9.637 27.205 22.491 21.088
(continued)
79
Table 2 (continued)
80

Location Arrival point P21 P22 P23 P24 P25 P26

Latitude 34.8539671 34.853703 34.87139649 34.77821126 34.78638958 34.77912055
Start point Latitude Longitude 137.4057124 137.4269815 137.4329877 137.3444784 137.3855037 137.4222122
P33 34.93778212 137.5777741 21.864 19.135 18.24 33.924 29.21 27.807
P34 35.00936674 137.5339967 38.673 37.315 34.861 52.103 47.389 45.987
P35 35.07888126 137.6921871 47.821 45.092 44.196 59.88 55.167 53.764
P36 34.91646628 137.2189389 22.513 24.195 25.174 22.18 22.343 27.726
P37 34.92218539 137.3005012 23.366 25.316 25.983 22.847 23.01 28.393
P38ˇ 34.99872367 137.2921787 36.531 38.212 39.191 36.474 36.637 42.02
Location Arrival point P27 P28 P29 P30 P31 P32
Latitude 34.7992955 34.76418687 34.83445434 34.76605784 34.87994115 34.90793958
Start point Latitude Longitude 137.444954 137.4264511 137.4547471 137.3712467 137.4812778 137.5159418
P1 34.8629707 137.3013095 16.169 17.289 19.316 15.179 22.215 25.447
P2 34.84561071 137.3194779 13.549 14.669 16.845 12.56 19.209 22.976
P3 34.8337699 137.3255714 14.471 13.227 16.385 10.863 18.748 22.515
P4 34.81672573 137.3128 16.175 14.931 16.716 9.384 21.156 24.923
P5 34.85331179 137.3394865 15.968 15.118 14.891 13.009 17.789 21.022
P6 34.80561814 137.3208462 13.738 12.384 16.173 7.936 21.639 25.406
P7 34.83734598 137.3401707 11.377 13.838 14.673 11.728 17.037 20.804
P8 34.80326089 137.3489893 11.769 10.376 14.205 5.935 18.343 23.366
(continued)
A. Y. Y. Chen and Y. Karasawa
Table 2 (continued)
Location Arrival point P27 P28 P29 P30 P31 P32
Latitude 34.7992955 34.76418687 34.83445434 34.76605784 34.87994115 34.90793958
Start point Latitude Longitude 137.444954 137.4264511 137.4547471 137.3712467 137.4812778 137.5159418
P9 34.8003621 137.3620273 9.697 8.343 12.132 5.378 16.526 21.548
P10 34.82064251 137.3580456 10.92 10.783 13.355 8.673 16.217 21.057
P11 34.81644871 137.3749543 8.571 8.98 11.006 7.296 14.535 19.557
P12 34.80506781 137.3851106 7.04 7.414 9.476 6.152 16.134 21.157
P13 34.81754503 137.3906428 6.544 7.815 8.534 7.88 12.932 17.955
P14 34.81837196 137.4069214 5.375 7.931 7.667 8.248 14.469 19.491
P15 34.82642438 137.3925224 6.626 9.823 8.183 9.457 12.581 17.603
P16 34.83316028 137.3642903 9.281 11.123 10.203 10.17 14.578 19.418
P17 34.83125618 137.402365 6.698 10.047 6.462 10.227 10.86 15.883
P18 34.83767594 137.3928127 7.763 10.858 7.495 11.753 11.87 14.552
P19 34.83295012 137.4176642 6.041 10.455 4.963 12.403 9.361 14.383
P20 34.84866582 137.3725186 10.815 12.919 10.546 12.552 14.353 17.763
P21 34.8539671 137.4057124 9.79 14.204 8.734 14.029 10.378 13.061
P22 34.853703 137.4269815 9.128 13.543 6.188 15.491 7.832 12.854
The Simulation for the Ideal Optimum Site Model in the Actual Distance …

P23 34.87139649 137.4329877 11.058 15.472 7.994 17.42 6.498 9.637

P24 34.77821126 137.3444784 12.672 9.095 15.108 3.934 21.766 26.788
P25 34.78638958 137.3855037 7.934 5.926 10.37 3.816 17.028 22.05
(continued)
81
Table 2 (continued)
82

Location Arrival point P27 P28 P29 P30 P31 P32

Latitude 34.7992955 34.76418687 34.83445434 34.76605784 34.87994115 34.90793958
Start point Latitude Longitude 137.444954 137.4264511 137.4547471 137.3712467 137.4812778 137.5159418
P26 34.77912055 137.4222122 3.396 2.016 8.533 6.355 15.087 21.032
P27 34.79929553 137.4449543 0 5.042 5.957 9.437 12.511 18.456
P28 34.76418687 137.4264511 5.041 0 10.271 6.058 16.825 22.77
P29 34.83445434 137.4547471 6.683 11.395 0 13.709 10.037 15.059
P30 34.76605784 137.3712467 8.975 5.465 13.379 0 20.037 25.06
P31 34.87994115 137.4812778 13.666 18.377 9.966 20.778 0 5.869
P32 34.90793958 137.5159418 18.456 23.167 15.304 25.8 5.869 0
P33 34.93778212 137.5777741 25.175 29.886 22.023 32.519 12.588 7.56
P34 35.00936674 137.5339967 43.354 48.066 40.202 50.699 30.768 23.717
P35 35.07888126 137.6921871 51.131 55.843 47.979 58.476 38.545 33.71
P36 34.91646628 137.2189389 29.95 27.169 28.872 25.06 31.77 35.003
P37 34.92218539 137.3005012 30.34 27.837 29.263 25.727 32.045 35.812
P38ˇ 34.99872367 137.2921787 43.967 41.463 42.889 39.354 47.22 47.903
Location Arrival point P33 P34 P35 P36 P37 P38
Latitude 34.93778212 35.00936674 35.07888126 34.91646628 34.92218539 34.99872367
Start point Latitude Longitude 137.5777741 137.5339967 137.6921871 137.2189389 137.3005012 137.2921787
P1 34.8629707 137.3013095 31.868 48.093 59.51 10.17 11.075 24.492
(continued)
A. Y. Y. Chen and Y. Karasawa
Table 2 (continued)
Location Arrival point P33 P34 P35 P36 P37 P38
Latitude 34.93778212 35.00936674 35.07888126 34.91646628 34.92218539 34.99872367
Start point Latitude Longitude 137.5777741 137.5339967 137.6921871 137.2189389 137.3005012 137.2921787
P2 34.84561071 137.3194779 29.397 45.622 57.039 13.173 14.455 27.495
P3 34.8337699 137.3255714 28.936 45.161 56.578 14.425 15.707 28.747
P4 34.81672573 137.3128 31.344 47.568 58.986 17.496 18.778 31.818
P5 34.85331179 137.3394865 27.443 43.667 55.085 14.148 14.729 28.469
P6 34.80561814 137.3208462 31.827 48.051 59.469 18.62 19.902 32.942
P7 34.83734598 137.3401707 27.225 43.449 54.867 16.134 16.716 30.456
P8 34.80326089 137.3489893 30.085 46.665 56.073 19.109 20.391 33.431
P9 34.8003621 137.3620273 28.267 46.446 54.255 19.379 20.661 33.701
P10 34.82064251 137.3580456 27.478 43.703 53.947 18.725 19.307 33.047
P11 34.81644871 137.3749543 26.276 43.087 52.265 20.086 20.667 34.407
P12 34.80506781 137.3851106 27.876 46.055 53.864 21.847 23.129 36.169
P13 34.81754503 137.3906428 24.674 42.853 50.662 21.403 21.985 35.725
P14 34.81837196 137.4069214 26.21 44.39 52.198 22.698 23.28 37.02
P15 34.82642438 137.3925224 24.322 41.59 50.311 20.926 21.508 35.248
The Simulation for the Ideal Optimum Site Model in the Actual Distance …

P16 34.83316028 137.3642903 25.839 42.064 52.307 17.832 18.414 32.154

P17 34.83125618 137.402365 22.602 40.781 48.59 22.451 24.379 36.773
P18 34.83767594 137.3928127 23.611 39.984 49.599 21.152 23.08 35.474
(continued)
83
Table 2 (continued)
84

Location Arrival point P33 P34 P35 P36 P37 P38

Latitude 34.93778212 35.00936674 35.07888126 34.91646628 34.92218539 34.99872367
Start point Latitude Longitude 137.5777741 137.5339967 137.6921871 137.2189389 137.3005012 137.2921787
P19 34.83295012 137.4176642 21.103 39.282 47.091 23.973 25.901 38.295
P20 34.84866582 137.3725186 24.184 40.408 52.083 18.417 18.999 32.739
P21 34.8539671 137.4057124 22.119 38.055 48.108 22.635 23.366 36.957
P22 34.853703 137.4269815 19.573 37.753 45.561 24.602 25.333 38.924
P23 34.87139649 137.4329877 18.24 33.806 44.228 25.252 25.983 39.574
P24 34.77821126 137.3444784 33.507 51.687 59.495 23.263 24.545 37.584
P25 34.78638958 137.3855037 28.769 46.949 54.757 22.371 23.653 36.693
P26 34.77912055 137.4222122 27.751 45.93 53.739 27.459 28.741 41.781
P27 34.79929553 137.4449543 25.175 43.354 51.163 29.862 31.79 44.184
P28 34.76418687 137.4264511 29.489 47.668 55.477 27.505 28.786 41.826
P29 34.83445434 137.4547471 21.778 39.957 47.766 28.386 30.314 42.708
P30 34.76605784 137.3712467 31.779 49.958 57.767 24.788 26.07 39.11
P31 34.87994115 137.4812778 12.588 30.768 38.576 31.261 31.992 43.568
P32 34.90793958 137.5159418 7.591 23.749 33.738 35.081 35.812 44.405
P33 34.93778212 137.5777741 0 17.288 27.466 43.156 43.887 50.826
P34 35.00936674 137.5339967 17.288 0 34.445 53.898 47.177 44.281
P35 35.07888126 137.6921871 27.469 34.445 0 69.113 69.843 65.06
(continued)
A. Y. Y. Chen and Y. Karasawa
Table 2 (continued)
Location Arrival point P33 P34 P35 P36 P37 P38
Latitude 34.93778212 35.00936674 35.07888126 34.91646628 34.92218539 34.99872367
Start point Latitude Longitude 137.5777741 137.5339967 137.6921871 137.2189389 137.3005012 137.2921787
P36 34.91646628 137.2189389 41.424 57.146 69.066 0 10.269 16.391
P37 34.92218539 137.3005012 42.233 46.225 69.875 10.115 0 17.615
P38ˇ 34.99872367 137.2921787 54.324 44.281 65.06 16.382 17.615 0
The Simulation for the Ideal Optimum Site Model in the Actual Distance …
85
 86

Table 3 The lk Matrix table of the actual distance method on the map for 38 areas (example) (Unit:lkm)
P1 P2 P3 P4 P5 P6 P7 P8 P9 P10
P1 0 31790.85 48807.85 88897.25 50609 102533 59791.55 101814.7 103792.7 90278.5
P2 156422.5 0 106293.4 265597.5 190784.3 297838.9 151746.7 358515.8 368248 301753.5
P3 146822.8 57723.05 0 124095.5 133070.6 145556.2 69484.8 186450.9 192928.9 153337
P4 219051.1 144230.9 94512.9 0 210146.9 60141.81 157055.5 154654.7 213763.3 180547
P5 98474.94 63027.24 66612.99 144638.9 0 155047.8 44975.55 145253.6 147978.8 103064.7
P6 76237.8 53710.5 38741.1 21767.85 67691.7 0 51706.65 33315.15 38759.4 47753.85
P7 247934.4 123194.7 98829.46 238620.8 96887.3 261088.1 0 239989.2 262279.9 149104.9
P8 525325.1 367548.9 305443.7 284220.7 432223.2 203349.9 295614.1 0 112426.1 173526
P9 242383.8 171617.6 142609.7 142384.2 200625.5 106111.8 139353.2 50425.65 0 88351.35
P10 248528.4 166094 137828.3 177091.1 139607.6 191909.4 88676.8 90952.06 84213.79 0
P11 270835.8 193742.6 179393.3 216112.2 167662.9 179747.9 123032.8 91524.4 79193.84 57206.16
P12 292063.2 224122 195911.3 233212.9 205411.1 171933.5 163419.8 116642.5 74747.4 96224.05
P13 160751.3 118785.2 132566 152554.1 110840.4 112800.6 86530.95 79789.05 52791.75 66052.8
P14 409866.4 331925.3 299574 342378.1 241573.2 272076.7 196452.3 208697.9 160653.5 185254.9
P15 441041 301590.5 304103.2 355345.4 265995.1 374723 212354.5 221529.3 195375.2 167241.5
P16 312299 177573.8 180001.3 229507.2 143184.5 248228.2 91361.52 169739.7 161243.5 62305.32
P17 243059.4 164097.5 165321.2 190276 157812.5 199694.3 120639.8 122864.6 104899.3 107383.7
P18 620904.3 402213.1 405682.7 476441.9 379187.4 503147.5 278989 394958.4 321413 241401.4
P19 284592.4 234528.1 206247.8 232427.5 195161.3 226942.6 188937.3 182246.5 148345.2 165714
P20 269829.5 177547.6 161106.9 226542.6 133130.9 241645.4 107830.7 208020.8 201153.2 121179.5
P21 269203.5 214650 203472 248022 176094 258309 167163.8 248629.5 185044.5 158031
(continued)
A. Y. Y. Chen and Y. Karasawa
Table 3 (continued)
P1 P2 P3 P4 P5 P6 P7 P8 P9 P10
P22 778463.6 641042.7 612834.1 725107.2 543919.6 751020.2 521424.2 632269 543409.5 443480.9
P23 379477.4 315225.5 302060.3 354530.3 269815.1 366646.1 259321.1 355245.8 303109.7 269075.7
P24 324084 221600.8 182452.6 151406.5 253435.6 116799.3 211682.3 114720 96269.2 171434.7
P25 429951.9 334071.4 294259.8 346900.7 373340 258792.5 290356.7 183343.9 123881 214908.1
P26 578878.2 486750.6 448510.8 499106 440458.6 414476.4 417247.4 342006.6 284858.8 326163
P27 925825.6 786327.1 728404.4 804992.8 786277.7 676802.3 681049 567030.5 480516.8 543081.2
P28 695346.1 584962.6 539145.2 599766.3 630194 498366.5 534653.3 411536.2 343064 444151.3
P29 339001 327342 306462.6 334088.2 257156.6 287879.4 251460.6 248310 217106.6 239659.2
P30 965202.3 784394.8 709346.3 623229.2 858483.6 530586.1 701988.6 396676 388166.7 559761
P31 1203408 1055507 1025203 1145983 950977.8 1173927 926821.8 1019987 924351.3 840519
P32 951093 851549.7 831153.3 912443.3 781196.9 931213.9 764938.9 872057 807690.1 780088.4
P33 1105412 1017345 999300.6 1071219 955103.7 1087825 940720.1 991193.5 934214.8 901328.7
P34 277357 262108.9 258984.6 271436.6 251332.3 274311.9 248841.9 271606.4 264650.3 251156.8
P35 900764 860150.4 851831.8 884985.8 831457.1 892656.4 824826.3 848094.4 821842.5 806682
P36 243947.7 313245.4 351254.2 427176.6 336860.9 456559 364384.5 455010 459275.6 430131.5
P37 103488.5 131075.9 146212.4 176447.6 137851.7 188148.7 157211.3 187541.4 189230.6 177624.3
The Simulation for the Ideal Optimum Site Model in the Actual Distance …

P38 187662.2 209908.4 222110.1 246483 215370.5 255915.5 230976.5 255418.2 256787.6 247431.6
P11 P12 P13 P14 P15 P16 P17 P18 P19 P20
P1 102787.1 133274.1 119914.6 132213.3 120400.8 87482.85 143782.6 130423.2 160280.3 99560.5
P2 363300.3 513307.2 447573.8 508142 449966.1 287997.9 573114.2 507326.5 654234.2 332526.9
P3 224595.1 289483.8 297083.7 390852 282824.9 158077.9 364831.4 321041.5 418826.9 204654.5
(continued)
87
Table 3 (continued)
88

P11 P12 P13 P14 P15 P16 P17 P18 P19 P20
P4 240384.9 294874.2 301256.1 379996.6 288431.5 197930.1 313746.4 286000.3 361762.6 245004.2
P5 126259.4 179390 158018.9 180844.7 149925.3 79767.57 175947.6 151154.7 206518.7 93926.16
P6 60554.7 65770.2 69503.4 89422.95 81727.8 64013.4 88745.85 90530.1 103916.6 78186.75
P7 199071.4 313526.4 267488.4 316616.2 269430.6 137937.5 369407.7 315998.3 435264.5 174044
P8 187376.8 284109 295837.5 421667.5 345376.4 256965.9 377266.8 418819.2 510133.9 390000.6
P9 71993.7 78832.35 89052.75 143586.6 122519.6 129333.2 141732.9 153155.7 183265.8 189002.3
P10 70620.57 123593.3 129748.2 201448 140745.3 47809.63 164518.8 132344.3 247653.3 117292.6
P11 0 51504.64 55623.92 125651.7 77802.56 65853.92 100717.8 87896.16 145266 113021
P12 40909.05 0 56565.6 83044.65 88720.45 98965.75 127272.6 122342.4 138624.2 140524.2
P13 26492.4 30680.1 0 34763.85 23774.85 55420.2 35164.8 41936.4 55286.55 76373.55
P14 120690.1 84229.32 63158.2 0 70604.8 138644.7 74990.02 93965.06 84863.66 178139.2
P15 104959 138689 63614.97 92243.61 0 120643.8 53830.98 65670.75 128334 139145.9
P16 88786.92 158264.3 145795.9 186768.8 128730 0 148480.9 106588.4 198317.8 87609.96
P17 66095.1 81817.02 43902.72 49075.38 24917.76 75976.92 0 26289.72 30609.54 82855.26
P18 169538.3 312213.2 148457.7 181787.1 94573.43 152347.9 69865.53 0 141045.3 166646.9
P19 120181.6 94468.65 88186.3 60119.95 56521.7 1 12537.7 32111.95 59847.65 0 128058.8
P20 141039.1 197169.4 171244.8 193512.6 115828.1 70816.86 130217.4 94244.1 170501.6 0
P21 130369.5 166353.8 116559 130592.3 95823 123747.8 83227.5 70976.25 96957 103194
P22 376045.7 402111.8 385635.6 312028.2 300142.8 357070 236125.3 218934.9 159661.3 345592.8
P23 249017.9 237045.2 229341.6 194926.1 208329.8 228697.7 165280.5 179089.7 134347.1 21 1764.2
P24 156736.2 145479.3 172103.9 207260.8 204034.3 209244.5 233813.7 250973.9 245118.4 266174.3
(continued)
A. Y. Y. Chen and Y. Karasawa
Table 3 (continued)
P11 P12 P13 P14 P15 P16 P17 P18 P19 P20
P25 135318.8 86479.12 158737.4 167595.7 204115.2 217249.9 234084.2 258453.1 308650.4 275898.3
P26 249813.8 206749.2 210237.4 232242.4 253791 318795.4 287466.8 322577 272275.2 422072.2
P27 427433.3 362152.9 347338.9 261417.7 336475.3 458295.8 330500.3 442790.5 298304.6 615916.7
P28 352672.7 301035.4 305253.9 331619.4 357438.1 435323.7 397787 439854.7 408372.3 559065.8
P29 197971.6 174440 169100 138128 130278.2 181542.2 107939.2 133339.8 81257 195746.6
P30 447604.1 355600.8 491814.3 508513 577291.5 657714.4 633849.9 679787.5 774413.9 810114.8
P31 844801.2 772278.3 754545.6 675324.9 665168.4 747518.4 596214 598849.2 513918.9 735220.8
P32 648620.3 705338.6 693403.7 640084.9 633249.1 717495.1 586876.9 534851.3 531451.9 691260.6
P33 886879.7 843663.5 833104.7 785933 779885.3 845984.5 738859.4 757427.3 689857.1 838629.3
P34 246396.8 248966.4 247138.2 238970.9 236746.5 241574.5 230820.5 229801.7 222336.1 237555.9
P35 800006 780083.5 775230.9 753469.9 750697 781153.5 731769.1 740343.9 709179.1 777762.7
P36 457107.1 522854 494043.6 520590.2 507531.3 425937.4 537795.4 508961.1 573349.8 442427.8
P37 188367 214559.4 203076.5 213648.4 205828.6 173334.9 217880.9 206398 232040 179892.4
P38 256091.4 277197.8 267948.9 276471 270167.4 243973.8 279875.3 270626.4 291296.7 249260
P21 P22 P23 P24 P25 P26 P27 P28 P29 P30
P1 143185.9 161761 172578.9 135915 137705.1 197198.3 178667.5 191043.5 213441.8 167728
The Simulation for the Ideal Optimum Site Model in the Actual Distance …

P2 570123.8 661574.2 714802.4 503792.4 535163.9 827837.6 736659.1 797553.5 915862.7 682887.2
P3 362840.9 423676.3 459106.3 282643.9 303996 498806.8 523705.5 478685.1 592973.2 393132
P4 377838.9 428954.9 458676.3 192399.1 307060.6 470649.9 491558.3 453753.1 507999.2 285179.8
P5 174820.7 209284.9 229324.1 216948.1 210862.6 276717.5 327184.3 309767.8 305116.6 266554.4
P6 118181.4 133571.7 142520.4 44679.45 69146.55 116955.3 125702.7 113313.6 147983 72614.4
(continued)
89
Table 3 (continued)
90

P21 P22 P23 P24 P25 P26 P27 P28 P29 P30
P7 366980 441223.4 484392.4 394435 397745.5 523191.4 502180.8 610809.3 647666.2 517673.9
P8 598265.2 676455.2 792511.5 267856.6 309911.7 603906.1 657298.7 579499.6 793349.3 331469.8
P9 270314.6 257864.7 310569.9 95340.3 88075.8 218962.1 242909.9 208992.2 303906.6 134718.9
P10 226067.5 252087.1 327520.8 202731.5 173736.5 290649.9 318536.4 314540.1 389565.4 252991.4
P11 180266.2 226533.1 289495.4 168999.6 108765.4 207709.9 233816.9 244974.4 300243.7 199034.9
P12 222222 259499.5 260846.3 130086.5 68350.1 145454.4 169312 178306.7 227897.8 147955.6
P13 92218.5 99509.85 133754 101678 67864.5 95767.65 97178.4 11 6052.8 126729.9 117018
P14 180731.7 167024.5 225052.8 266202.2 136190 181062.7 148242.5 218737 211455.9 227479.8
P15 194575.8 241744.5 296413 320701.7 234054.4 321958 252251.8 373961.6 311526.8 360028
P16 224762.6 257570.3 352683.4 310570.3 235428.8 358862.5 341355.2 409103.9 375266.3 374052.6
P17 79184.34 85821.66 121029.1 175833.4 128241.2 160945.7 124180.9 186271.4 119805.5 189608.6
P18 184257.9 225788.2 394748.1 494473.4 443848.5 498994.4 408100.9 570805.1 394012.2 617855.2
P19 91979.05 60878.5 116311 193099.6 176878.3 168689.9 1 17497.5 203349.8 96530.35 241238.4
P20 151504.1 201509.9 235848.1 331459.8 259572.6 343470.7 321530 384081.9 313532.6 373171
P21 0 56760.75 112995 269041.5 217059.8 251545.5 198247.5 287631 176863.5 284087.3
P22 152315.9 0 158488.1 660783.5 621403.8 599928.6 465619.3 690828.4 315649.9 790195.9
P23 133083 74101.95 0 357988.5 336547.4 326506.5 263733.3 369007.2 190656.9 415467
P24 328170.9 316292.6 366578.2 0 118806.9 228340.6 302860.8 217370.5 361081.2 94022.6
P25 416443.8 396521 493284 191319.8 0 213957.8 269280 201128.4 351957.8 129515
P26 399839 356709.2 427549 321240.4 205510.4 0 1 10709.6 65721.6 278175.8 207173
P27 524316.8 426149.4 520415.8 624558.2 391780.9 170509.1 0 248974 294156.7 465999.1
(continued)
A. Y. Y. Chen and Y. Karasawa
Table 3 (continued)
P21 P22 P23 P24 P25 P26 P27 P28 P29 P30
P28 532426.9 509537.7 580158.2 373335.5 243461 78744.96 196901.5 0 401185.3 236625.5
P29 162727.6 111392.4 143557 269029.2 185120 165824.8 118957.4 202831 0 244020.2
P30 977614.4 940058.1 1122401 274282.3 224953.7 379017.5 574220.5 349650.7 855988.4 0
P31 554819.4 405930.6 333297.9 1217792 959048.1 894760.2 750263.4 1008897 547133.4 1140712
P32 569399.5 458771.2 356087.2 1005225 831042.5 779201.6 681949.2 856020.7 565482.8 953310
P33 714734.2 625523.2 596265.6 1108976 954874.9 909010.8 822970.8 976973.3 719931.9 1063046
P34 218889.2 211202.9 197313.3 294903 268221.7 260286.4 245383.6 272053.6 227543.3 286956.3
P35 720662.5 679536.4 666033.7 902391.6 831366.7 810223.5 770544.2 841554 723043.5 881233.3
P36 536484.8 576566.9 599896.4 528549.4 532433.7 660710.6 713708.5 647437.3 688019.8 597179.8
P37 221743.3 240248.8 246578.7 216818 218364.9 269449.6 287926.6 264173.1 277705.9 244149.2
P38 279462.2 292321.8 299811.2 279026.1 280273.1 321453 336347.6 317192 328100.9 301058.1
P31 P32 P33 P34 P35 P36 P37 P38
P1 245475.8 2811 89.4 352141.4 531427.7 657585.5 112378.5 122378.8 270636.6
P2 1044393 1249205 1598315 2480468 3101210 716216 785918.4 1494903
P3 678490.1 814817.9 1047194 1634377 2047558 522040.8 568436.3 1040354
P4 642930.8 757410 952544.2 1445592 1792585 531703.4 570663.4 966949
The Simulation for the Ideal Optimum Site Model in the Actual Distance …

P5 364496.6 430740.8 562307.1 894736.8 1128692 289892.5 301797.2 583329.8

P6 197996.9 232464.9 291217.1 439666.7 544141.4 170373 182103.3 301419.3
P7 752013.2 918288.6 1201712 1917839 2421829 712154.8 737844.2 1344328
P8 1024457 1304991 1680247 2606240 3131677 1067238 1138837 1867121
P9 413976.3 539777.4 708088.4 1163472 1359088 485444 517558.1 844210.1
(continued)
91
Table 3 (continued)
92

P31 P32 P33 P34 P35 P36 P37 P38

P10 473049.9 614232.7 801533.3 1274817 1573634 546208.3 563185.2 963981
P11 396514.8 533515 716809.3 1175413 1425789 547946.1 563795.8 938623
P12 388022.7 508825.9 670417.8 1107623 1295429 525420.4 556252.5 869864.5
P13 192040.2 266631.8 366408.9 636367.1 752330.7 317834.6 326477.3 530516.3
P14 399055 537561.8 722871.8 1224276 1439621 626010.8 642062.4 1021012
P15 478958.7 670146.2 925938.5 1583331 1915340 796652.8 818809.6 1341891
P16 536178.8 714194 950358.4 1547114 1923851 655861 677266.9 1182624
P17 201344.4 294470.8 419041.1 756079.7 900858.6 416241.5 451986.7 681771.4
P18 624005.9 764998.6 1241230 2101959 2607419 1111961 1213316 1864868
P19 182071.5 279749.4 410453.4 764034.9 915920 466274.9 503774.5 744837.8
P20 426714.7 528094 718990.3 1201330 1548428 547537.4 564840.3 973330.5
P21 210154.5 264485.3 447909.8 770613.8 974187 458358.8 473161.5 748379.3
P22 399510.3 655682.5 998418.7 1925781 2324067 1254948 1292236 1985513
P23 154977.3 229842.5 435024 806273.1 1054838 602260.2 619694.6 943839.9
P24 520207.4 640233.2 800817.3 1235319 1421931 555985.7 586625.5 898257.6
P25 577930.3 748377 976419.9 1593449 1858453 759271.7 802782.8 1245360
P26 491836.2 685643.2 904682.6 1497318 1751891 895163.4 936956.6 1362061
P27 617793.2 911357.3 1243142 2140821 2526429 1474586 1569790 2181806
P28 657184.5 889396.2 1151840 1861912 2166932 1074345 1124381 1633724
P29 178658.6 268050.2 387648.4 711234.6 850234.8 505270.8 539589.2 760202.4
P30 1281967 1603339 2033220 3196313 3695933 1585936 1667959 2502258
(continued)
A. Y. Y. Chen and Y. Karasawa
Table 3 (continued)
P31 P32 P33 P34 P35 P36 P37 P38
P31 0 322208.1 691081.2 1689163 2117822 1716229 1756361 2391883
P32 216859.6 0 280487.5 877525.6 1246619 1296243 1323253 1640765
P33 411501.7 247136.4 0 565144.7 897863.5 1410770 1434666 1661502
P34 174146.9 134238.2 97850.08 0 194958.7 305062.7 267021.8 250630.5
P35 580873.2 508009.7 413957.8 519086.2 0 1041533 1052534 980454.2
P36 757079.1 834121.5 987133.9 1361789 1645843 0 244710.3 390597.5
P37 304107.1 339855.9 400791.2 438675.3 663113.8 95991.35 0 167166.4
P38 361233 366458 415578.6 338749.7 497709 125322.3 134754.8 0
The Simulation for the Ideal Optimum Site Model in the Actual Distance …
93
94 A. Y. Y. Chen and Y. Karasawa

Fig. 8 Process of selecting optimum sites

optimal solutions and approximate solutions, but the Gravity Model is generally the
main one-step horizontal optimal location simulation tool. However, in this research,
there are 20 thousand delivery customers must be grouped and the distance of the
center of gravity must be estimated, and a simple location model that applies mapping
graph theory is applied to represent its validity. We adopted a method of verifica-
tion using a gravitational model (Gravity Model), using the Gravity Model to show
which area is the best for establishing a delivery center, with the lowest cost for the
firm, and knowing how important the optimum site selection affects the company’s
management and strategy. However, this paper is only focused on the actual distance
method, in the near future, we will produce the paper for the method of approximate
distance.

3 The Results of the Simulation for the Ideal Optimum Site

Type in the Actual Distance of 38 Places Mapping

Our research is introducing from longitude and latitude for selecting the ideal type
of optimum sites, generally speaking, the simulation of one phase of multi points
model uses the following gravity models:

{
n
{ ( )}
f (C) = T (Pi ) min K Pi , C j (1)
i=1

C j ∈ C = {C1 , C2 , . . . , Cm } ∈ P = {P1 , P2 , . . . , Pn } (2)

The Simulation for the Ideal Optimum Site Model in the Actual Distance … 95

However, T(Pi) is the demanding amount in the area Pi, and K(Pi, Cj) is the
distance between the area Pi and the delivery center Cj. At this moment, the objective
function f (C) is the minimum, namely, it becomes the min f (C), which the group C
puts the most optimum delivery centers together. For instance, when we search for
the most optimum delivery centers of two places from the whole area, first of all,
choosing the two candidate places C' = {C' 1, C' 2} of the delivery centers from the P,
and also making the P as a yardstick for the shortest distance, to allocate it for either
delivery center candidate area, we divided P into network N' 1, N' 2 (but N' 1 ∪ N' 2
= P) of the delivery destination area affiliated with each candidate delivery center
area, at this moment, C’1belong to (∈)N' 1, C' 2 belong to (∈)N' 2 and C' 1 and C' 2
are respectively the optimum sites in the network of N' 1, N' 2, When it is equivalent
to the center of gravity, C' 1 and C' 2 are respectively the optimum delivery centers in
the N' 1 and N' 2.
Regarding the above all with all of the combination of the C' (nC2 ways), and
attempting repeatedly, to find out the optimum delivery centers C = {C1, C2}by
the combination of the minimized {TK. Even the number of the optimum delivery
centers m is increasing, it can be accepted by similar logic.
The summary from the 38 map areas with the actual distance method, for the
simulation of ideal type optimum sites are for the one place is 9,576,888lk, for the
two places are 7,342,602lk, for the three places are 6,251,990lk, for the four places
are 5,291,167lk and for the five places are 4,614,433lk, those results ignored the
characters of the inventory fee and delivery fee, as shown in Table 4.

3.1 Simulation of 38 Locations for the Ideal Type

of Optimum Sites

We analyzed the 38 map areas with the actual distance method of the Ideal type, from
one to five places to be the delivery centers, the calculation time of the Ideal type
optimum sites from one to five places are 0.73 ms, 2.50 ms, 21.95 ms, 220.49 ms,
and 1,732.2 ms, as shown in Fig. 9, 10, 11, 12 and 13. The longest running time for
the Ideal type of optimum sites was taken 1732.2 ms.

3.2 Available Locations

There are four types of simulations are included in the simulation model, and one
to five places of optimum locations are, the Plus α type means the present center
plus one to four optimum locations, The Available type means the location is able
to be a delivery center by Japanese Law, the Ideal type means any area where
there is an Ideal place to set up a delivery center. There are five locations that are
96 A. Y. Y. Chen and Y. Karasawa

Table 4 The summary of ideal type of the optimum sites in actual distance of 38 map areas
{
Locations Delivery center lK Calculate time (ms)
Centers Address
1 Location P11 20, 4 Chome, Minamioodoori, 9,576,888 0.73
Toyogawa-shi, Aichi-ken
2 Locations P11 20, 4 Chome, Minamioodoori, 5,901,167 2.5
Toyogawa-shi, Aichi-ken
P32 100, Tomizawa, Sinshiro-shi, 1,441,435
Aichi-ken
Subtotal 7,342,602
3 Locations P2 46, Sinchou, Goyu-cho, 1,127,907 21.95
Toyogawa-shi, Aichi-ken
P11 20, 4 Chome, Minamioodoori, 3,682,648
Toyogawa-shi, Aichi-ken
P32 100, Tomizawa, Sinshiro-shi, 1,441,435
Aichi-ken
Subtotal 6,251,990
4 Locations P3 30, Toyonari, Koumachi, 1,803,880 220.49
Toyogawa-shi, Aichi-ken
P18 19, Hashimoto, Sanzougo-Cho, 1,433,466
Toyogawa-shi, Aichi-ken
P26 14–4, 3-Chome, Ushikawadoori, 842,229
Toyohashi-shi, Aichi-ken
P32 100, Tomizawa, Sinshiro-shi, 1,211,592
Aichi-ken
Subtotal 5,291,167
5 Locations P2 46, Sinchou, Goyu-cho, 1,074,196 1732.2
Toyogawa-shi, Aichi-ken
P9 43, Nishiura, Kosakaimachi, 609,491
Toyogawa-shi, Aichi-ken
P18 19, Hashimoto, Sanzougo-Cho, 1,090,883
Toyogawa-shi, Aichi-ken
P26 14–4, 3-Chome, Ushikawadoori, 628,272
Toyohashi-shi, Aichi-ken
P32 100, Tomizawa, Sinshiro-shi, 1,211,592
Aichi-ken
Sub total 4,614,433
Total 23,500,193 1,978
Remarks
1. The P11 center is near the present delivery center P9. Therefore, the present delivery center is the
optimum site, too. The present Delivery Center’s address is included on P9 area
2. MS Stands for Micro Second
The Simulation for the Ideal Optimum Site Model in the Actual Distance … 97

Fig. 9 One location for the ideal type of optimum sites

Fig. 10 Two locations for the ideal type of optimum sites

Fig. 11 Three locations for the ideal type of optimum sites

unavailable in our research, such as P14 (Location: Toyokawa-shi, Householder:

198 customers, Delivery Amount: 27,580l), as shown in Fig. 14, and P15 (Location:
Toyokawa-shi, Householder: 415 customers, Delivery Amount: 38,070l) as shown
in Fig. 15, P17 (Location: Toyokawa-shi, Householder: 212 customers, Delivery
98 A. Y. Y. Chen and Y. Karasawa

Fig. 12 Four locations for the ideal type of optimum sites

Fig. 13 Five locations for the ideal type of optimum sites

Amount: 18,540l), P18 (Location: Toyokawa-shi, Householder: 624 customers,

Delivery Amount: 52,570l), and P30 (Location: Toyohashi-shi, Householder: 522
customers, Delivery Amount: 63,980l), these places are near the Toyogawa Station,
where the r = 200 km is not allowed to establish a dangerous goods delivery center,
like gas. Even though, this paper is only focused on the Ideal type optimum sites in
the actual distance method.

3.3 Summary of Optimum Site

In the 38 mapping actual distance method, there are five patterns of optimum sites
in the Ideal type. It is true that the density of the customers is higher in the cities
than those that are scattered in the suburbs. It is easy to estimate the optimum sites,
between an overcrowded area and an underpopulated area of customer distribution,
when actually, with the result of our simulations, there are only two places, one is
in the overcrowded area, and the other is in the underpopulated area (Fig. 3). From
The Simulation for the Ideal Optimum Site Model in the Actual Distance … 99

Fig. 14 Location P14

Fig. 15 Location P15

the result of the calculation, we picked the optimum sites for the minimum total
liter-kilometer ({lk) places (Table 5).

3.4 Expenses for Delivery Center

There is a storage fee and a delivery fee for each delivery center, the former is the
expense for inventory, calculated by the liter base, the latter is for expenses of the
delivery goods, calculated by the liter and kilometer base. To allocate the Actual
Distance Method to the 38 areas map actual distance. Usually, when we calculate the
optimum inventory on the SKU (Stock Keeping Unit) base, the items of allocating
shelves, such as a product or service, and all attributes associated with the working
staff, equipment, and the frequency of storage and shipping times, and so on. Our
research is based on the master data, to allocate {l to produce the inventory fee.
The present {lk calculated the delivery fee. Usually, the delivery fee is calculated
for each optimum center by the delivery routes, which should be estimated by a
Table 5 lk of 38 area map actual distance method
100

Area Location Destination 10L (U) 20L (U) 50L (U) Demands (L)
P1 Mukaidani, Nagasawamachi, Toyogawa-shi, Aichi-ken 142 1 28 113 11,050
P2 46, Sinchou, Goyu-cho, Toyogawa-shi, Aichi-ken 667 0 245 422 54,370
P3 30, Toyonari, Koumachi, Toyogawa-shi, Aichi-ken 471 1 103 367 36,190
P4 11–53, Otuyama, Hiroishi, Mito-cho, Toyogawa-shi, Aichi-ken 346 0 89 257 30,390
P5 23, Komaba, Hiraomachi, Toyogawa-shi, Aichi-ken 217 0 41 176 20,490
P6 154, Oumazenden, Mitocho, Toyogawa-shi, Aichi-ken 129 0 30 99 9150
P7 100, Hongou, Yahatacho, Toyogawa-shi, Aichi-ken 519 0 136 383 44,140
P8 350–261, Minamiyamasinden, Inacho, Toyogawa-shi, Aichi-ken 764 0 201 563 55,850
P9 43, Nishiura, Kosakaimachi, Toyogawa-shi, Aichi-ken 316 1 73 242 25,050
P10 10–5, 6 Chome, Zoushi, Toyogawa-shi, Aichi-ken 313 0 65 248 29,170
P11 20, 4 Chome, Minamioodoori, Toyogawa-shi, Aichi-ken 290 0 55 236 27,280
P12 41, Shiroshita, Ushikubomachi, Toyogawa-shi, Aichi-ken 165 0 37 128 24,050
P13 84–31, Shidou, Furuzyucho, Toyogawa-shi, Aichi-ken 183 0 42 141 14,850
P14 49, Shimonishiura, Miyaharacho, Toyogawa-shi, Aichi-ken 198 0 51 147 27,580
P15 67–1 2 Chome„ Sakuragidoori, Toyogawa-shi, Aichi-ken 415 0 124 291 38,070
P16 2-Chome, Honohara, Toyogawa-shi, Aichi-ken 420 2 100 318 36,780
P17 2, 1-Chome, Shinyutakamachi, Toyogawa-shi, Aichi-ken 212 0 60 152 18,540
P18 19, Hashimoto, Sanzougo-Cho, Toyogawa-shi, Aichi-ken 624 0 170 454 52,570
P19 43, Teramae, Asouda-Cho, Toyogawa-shi, Aichi-ken 259 0 54 205 19,450
P20 84, 2-Chome, Misenryou, Toyogawa-shi, Aichi-ken 278 0 71 207 29,730
P21 301, Shimonawate, Ookimachi, Toyogawa-shi, Aichi-ken 209 0 53 156 20,250
(continued)
A. Y. Y. Chen and Y. Karasawa
Table 5 (continued)
Area Location Destination 10L (U) 20L (U) 50L (U) Demands (L)
P22 137, Zyooshinkiri, Ikkucho, Toyogawa-shi, Aichi-ken 661 0 151 510 51,010
P23 13–6, Warita, Uenagayama-Cho, Toyogawa-shi, Aichi-ken 267 0 87 181 23,850
P24 41–3, Kadou, Maeshibamachi, Toyohashi-shi, Aichi-ken 299 1 87 211 23,900
P25 55, Ikan, Shimojicho, Toyohashi-shi, Aichi-ken 359 0 91 268 33,940
P26 14–4, 3-Chome, Ushikawadoori, Toyohashi-shi, Aichi-ken 276 0 63 213 32,600
P27 22, Oota, Ishimakihoncho, Toyohashi-shi, Aichi-ken 608 1 197 410 49,380
P28 146, Iharacho, Toyohashi-shi, Aichi-ken 398 0 97 301 39,060
P29 74 Koutubo, Ishimakinishikawacho, Toyohashi-shi, Aichi-ken 303 0 112 191 17,800
P30 102, Araki, Hanadacho, Toyohashi-shi, Aichi-ken 522 0 144 378 63,980
P31 Ichisukitanakagawara, Sinshiro-shi, Aichi-ken 461 0 91 370 54,900
P32 100, Tomizawa, Sinshiro-shi, Aichi-ken 360 0 73 287 36,950
P33 2–4, Nagashinohiyake, Sinshiro-shi, Aichi-ken 441 0 131 310 32,690
P34 Tadajishionosawa, Sinshiro-shi, Aichi-ken 142 2 100 40 5660
P35 Nakashitara, Toueicho, Kitasitara-gun, Aichi-ken 357 16 222 119 15,070
P36 12–15, Harayama, Okazaki-shi, Aichi-ken 233 0 42 191 23,830
P37 105, Harashinden, Kashiyamacho, Okazaki-shi, Aichi-ken 126 0 38 88 9490
The Simulation for the Ideal Optimum Site Model in the Actual Distance …

P38 Chuicho, Okazaki-shi, Aichi-ken 62 0 17 45 7650

Area 10L (L) 20L (L) 50L (L) 10L (Bottle) 20L (Bottle) 50L (Bottle) 距離 (km)
P1 P2 P3 P4 P5
P1 60 1340 9650 6 67 193 0.00 2.88 4.42 8.05 4.58
P2 0 15,520 38,850 0 776 777 2.88 0.00 1.96 4.89 3.51
(continued)
101
Table 5 (continued)
102

Area 10L (L) 20L (L) 50L (L) 10L (Bottle) 20L (Bottle) 50L (Bottle) 距離 (km)
P1 P2 P3 P4 P5
P3 10 5580 30,600 1 279 612 4.06 1.60 0.00 3.43 3.68
P4 0 4740 25,650 0 237 513 7.21 4.75 3.11 0.00 6.92
P5 0 2340 18,150 0 117 363 4.81 3.08 3.25 7.06 0.00
P6 0 1300 7850 0 65 157 8.33 5.87 4.23 2.38 7.40
P7 0 7140 37,000 0 357 740 5.62 2.79 2.24 5.41 2.20
P8 0 8200 47,650 0 410 953 9.41 6.58 5.47 5.09 7.74
P9 10 3140 21,900 1 157 438 9.68 6.85 5.69 5.68 8.01
P10 0 3420 25,750 0 171 515 8.52 5.69 4.73 6.07 4.79
P11 0 2880 24,400 0 144 488 9.93 7.10 6.58 7.92 6.15
P12 0 1800 22,250 0 90 445 12.14 9.32 8.15 9.70 8.54
P13 0 1900 12,950 0 95 259 10.83 8.00 8.93 10.27 7.46
P14 0 2480 25,100 0 124 502 14.86 12.04 10.86 12.41 8.76
P15 0 5120 32,950 0 256 659 11.59 7.92 7.99 9.33 6.99
P16 20 4660 32,100 2 233 642 8.49 4.83 4.89 6.24 3.89
P17 0 3140 15,400 0 157 308 13.11 8.85 8.92 10.26 8.51
P18 0 7720 44,850 0 386 897 11.81 7.65 7.72 9.06 7.21
P19 0 2700 16,750 0 135 335 14.63 12.06 10.60 11.95 10.03
P20 0 3480 26,250 0 174 525 9.08 5.97 5.42 7.62 4.48
P21 0 2600 17,650 0 130 353 13.29 10.60 10.05 12.25 8.70
P22 0 6760 44,250 0 338 885 15.26 12.57 12.01 14.22 10.66
P23 0 3500 20,350 0 175 407 15.91 13.22 12.67 14.87 11.31
(continued)
A. Y. Y. Chen and Y. Karasawa
Table 5 (continued)
Area 10L (L) 20L (L) 50L (L) 10L (Bottle) 20L (Bottle) 50L (Bottle) 距離 (km)
P1 P2 P3 P4 P5
P24 10 4840 19,050 1 242 381 13.56 9.27 7.63 6.34 10.60
P25 0 3840 30,100 0 192 602 12.67 9.84 8.67 10.22 11.00
P26 0 3400 29,200 0 170 584 17.76 14.93 13.76 15.31 13.51
P27 10 8120 41,250 1 406 825 18.75 15.92 14.75 16.30 15.92
P28 0 4460 34,600 0 223 692 17.80 14.98 13.80 15.36 16.13
P29 0 4100 13,700 0 205 274 19.05 18.39 17.22 18.77 14.45
P30 0 6780 57,200 0 339 1144 15.09 12.26 11.09 9.74 13.42
P31 0 4600 50,300 0 230 1006 21.92 19.23 18.67 20.87 17.32
P32 0 3200 33,750 0 160 675 25.74 23.05 22.49 24.69 21.14
P33 0 4940 27,750 0 247 555 33.82 31.12 30.57 32.77 29.22
P34 20 2640 3000 2 132 60 49.00 46.31 45.76 47.96 44.41
P35 160 5760 9150 16 288 183 59.77 57.08 56.53 58.73 55.17
P36 0 2780 21,050 0 139 421 10.24 13.15 14.74 17.93 14.14
P37 0 1940 7550 0 97 151 10.91 13.81 15.41 18.59 14.53
P38 0 1000 6650 0 50 133 24.53 27.44 29.03 32.22 28.15
The Simulation for the Ideal Optimum Site Model in the Actual Distance …

Area 距離 (km)
P6 P7 P8 P9 P10 P11 P12 P13 P14 P15 P16
P1 9.28 5.41 9.21 9.39 8.17 9.30 12.06 10.85 11.97 10.90 7.92
P2 5.48 2.79 6.59 6.77 5.55 6.68 9.44 8.23 9.35 8.28 5.30
P3 4.02 1.92 5.15 5.33 4.24 6.21 8.00 8.21 10.80 7.82 4.37
P4 1.98 5.17 5.09 7.03 5.94 7.91 9.70 9.91 12.50 9.49 6.51
(continued)
103
Table 5 (continued)
104

Area 距離 (km)
P6 P7 P8 P9 P10 P11 P12 P13 P14 P15 P16
P5 7.57 2.20 7.09 7.22 5.03 6.16 8.76 7.71 8.83 7.32 3.89
P6 0.00 5.65 3.64 4.24 5.22 6.62 7.19 7.60 9.77 8.93 7.00
P7 5.92 0.00 5.44 5.94 3.38 4.51 7.10 6.06 7.17 6.10 3.13
P8 3.64 5.29 0.00 2.01 3.11 3.36 5.09 5.30 7.55 6.18 4.60
P9 4.24 5.56 2.01 0.00 3.53 2.87 3.15 3.56 5.73 4.89 5.16
P10 6.58 3.04 3.12 2.89 0.00 2.42 4.24 4.45 6.91 4.83 1.64
P11 6.59 4.51 3.36 2.90 2.10 0.00 1.89 2.04 4.61 2.85 2.41
P12 7.15 6.80 4.85 3.11 4.00 1.70 0.00 2.35 3.45 3.69 4.12
P13 7.60 5.83 5.37 3.56 4.45 1.78 2.07 0.00 2.34 1.60 3.73
P14 9.87 7.12 7.57 5.83 6.72 4.38 3.05 2.29 0.00 2.56 5.03
P15 9.84 5.58 5.82 5.13 4.39 2.76 3.64 1.67 2.42 0.00 3.17
P16 6.75 2.48 4.62 4.38 1.69 2.41 4.30 3.96 5.08 3.50 0.00
P17 10.77 6.51 6.63 5.66 5.79 3.57 4.41 2.37 2.65 1.34 4.10
P18 9.57 5.31 7.51 6.11 4.59 3.23 5.94 2.82 3.46 1.80 2.90
P19 11.67 9.71 9.37 7.63 8.52 6.18 4.86 4.53 3.09 2.91 5.79
P20 8.13 3.63 7.00 6.77 4.08 4.74 6.63 5.76 6.51 3.90 2.38
P21 12.76 8.26 12.28 9.14 7.80 6.44 8.22 5.76 6.45 4.73 6.11
P22 14.72 10.22 12.40 10.65 8.69 7.37 7.88 7.56 6.12 5.88 7.00
P23 15.37 10.87 14.90 12.71 11.28 10.44 9.94 9.62 8.17 8.74 9.59
P24 4.89 8.86 4.80 4.03 7.17 6.56 6.09 7.20 8.67 8.54 8.76
P25 7.63 8.56 5.40 3.65 6.33 3.99 2.55 4.68 4.94 6.01 6.40
(continued)
A. Y. Y. Chen and Y. Karasawa
Table 5 (continued)
Area 距離 (km)
P6 P7 P8 P9 P10 P11 P12 P13 P14 P15 P16
P26 12.71 12.80 10.49 8.74 10.01 7.66 6.34 6.45 7.12 7.79 9.78
P27 13.71 13.79 11.48 9.73 11.00 8.66 7.33 7.03 5.29 6.81 9.28
P28 12.76 13.69 10.54 8.78 11.37 9.03 7.71 7.82 8.49 9.15 11.15
P29 16.17 14.13 13.95 12.20 13.46 11.12 9.80 9.50 7.76 7.32 10.20
P30 8.29 10.97 6.20 6.07 8.75 7.00 5.56 7.69 7.95 9.02 10.28
P31 21.38 16.88 18.58 16.84 15.31 15.39 14.07 13.74 12.30 12.12 13.62
P32 25.20 20.70 23.60 21.86 21.11 17.55 19.09 18.77 17.32 17.14 19.42
P33 33.28 28.78 30.32 28.58 27.57 27.13 25.81 25.49 24.04 23.86 25.88
P34 48.47 43.97 47.99 46.76 44.37 43.53 43.99 43.66 42.22 41.83 42.68
P35 59.23 54.73 56.28 54.54 53.53 53.09 51.76 51.44 50.00 49.81 51.84
P36 19.16 15.29 19.09 19.27 18.05 19.18 21.94 20.73 21.85 21.30 17.87
P37 19.83 16.57 19.76 19.94 18.72 19.85 22.61 21.40 22.51 21.69 18.27
P38 33.45 30.19 33.39 33.57 32.34 33.48 36.24 35.03 36.14 35.32 31.89
Area 距離 (km)
P17 P18 P19 P20 P21 P22 P23 P24 P25 P26 P27
The Simulation for the Ideal Optimum Site Model in the Actual Distance …

P1 13.01 11.80 14.51 9.01 12.96 14.64 15.62 12.30 12.46 17.85 16.17
P2 10.54 9.33 12.03 6.12 10.49 12.17 13.15 9.27 9.84 15.23 13.55
P3 10.08 8.87 11.57 5.66 10.03 11.71 12.69 7.81 8.40 13.78 14.47
P4 10.32 9.41 11.90 8.06 12.43 14.12 15.09 6.33 10.10 15.49 16.18
P5 8.59 7.38 10.08 4.58 8.53 10.21 11.19 10.59 10.29 13.51 15.97
P6 9.70 9.89 11.36 8.55 12.92 14.60 15.58 4.88 7.56 12.78 13.74
(continued)
105
Table 5 (continued)
106

Area 距離 (km)
P17 P18 P19 P20 P21 P22 P23 P24 P25 P26 P27
P7 8.37 7.16 9.86 3.94 8.31 10.00 10.97 8.94 9.01 11.85 11.38
P8 6.76 7.50 9.13 6.98 10.71 12.11 14.19 4.80 5.55 10.81 11.77
P9 5.66 6.11 7.32 7.55 10.79 10.29 12.40 3.81 3.52 8.74 9.70
P10 5.64 4.54 8.49 4.02 7.75 8.64 11.23 6.95 5.96 9.96 10.92
P11 3.69 3.22 5.33 4.14 6.61 8.30 10.61 6.20 3.99 7.61 8.57
P12 5.29 5.09 5.76 5.84 9.24 10.79 10.85 5.41 2.84 6.05 7.04
P13 2.37 2.82 3.72 5.14 6.21 6.70 9.01 6.85 4.57 6.45 6.54
P14 2.72 3.41 3.08 6.46 6.55 6.06 8.16 9.65 4.94 6.57 5.38
P15 1.41 1.73 3.37 3.66 5.11 6.35 7.79 8.42 6.15 8.46 6.63
P16 4.04 2.90 5.39 2.38 6.11 7.00 9.59 8.44 6.40 9.76 9.28
P17 0.00 1.42 1.65 4.47 4.27 4.63 6.53 9.48 6.92 8.68 6.70
P18 1.33 0.00 2.68 3.17 3.51 4.30 7.51 9.41 8.44 9.49 7.76
P19 1.65 3.08 0.00 6.58 4.73 3.13 5.98 9.93 9.09 8.67 6.04
P20 4.38 3.17 5.74 0.00 5.10 6.78 7.93 11.15 8.73 11.55 10.82
P21 4.11 3.51 4.79 5.10 0.00 2.80 5.58 13.29 10.72 12.42 9.79
P22 4.63 4.29 3.13 6.78 2.99 0.00 3.11 12.95 12.18 11.76 9.13
P23 6.93 7.51 5.63 8.88 5.58 3.11 0.00 15.01 14.11 13.69 11.06
P24 9.78 10.50 10.26 11.14 13.73 13.23 15.34 0.00 4.97 9.55 12.67
P25 6.90 7.62 9.09 8.13 12.27 11.68 14.53 5.64 0.00 6.30 7.93
P26 8.82 9.90 8.35 12.95 12.27 10.94 13.12 9.85 6.30 0.00 3.40
P27 6.69 8.97 6.04 12.47 10.62 8.63 10.54 12.65 7.93 3.45 0.00
(continued)
A. Y. Y. Chen and Y. Karasawa
Table 5 (continued)
Area 距離 (km)
P17 P18 P19 P20 P21 P22 P23 P24 P25 P26 P27
P28 10.18 11.26 10.46 14.31 13.63 13.05 14.85 9.56 6.23 2.02 5.04
P29 6.06 7.49 4.57 11.00 9.14 6.26 8.07 15.11 10.40 9.32 6.68
P30 9.91 10.63 12.10 12.66 15.28 14.69 17.54 4.29 3.52 5.92 8.98
P31 10.86 10.91 9.36 13.39 10.11 7.39 6.07 22.18 17.47 16.30 13.67
P32 15.88 14.48 14.38 18.71 15.41 12.42 9.64 27.21 22.49 21.09 18.46
P33 22.60 23.17 21.10 25.65 21.86 19.14 18.24 33.92 29.21 27.81 25.18
P34 40.78 40.60 39.28 41.97 38.67 37.32 34.86 52.10 47.39 45.99 43.35
P35 48.56 49.13 47.06 51.61 47.82 45.09 44.20 59.88 55.17 53.76 51.13
P36 22.57 21.36 24.06 18.57 22.51 24.20 25.17 22.18 22.34 27.73 29.95
P37 22.96 21.75 24.45 18.96 23.37 25.32 25.98 22.85 23.01 28.39 30.34
P38 36.59 35.38 38.08 32.58 36.53 38.21 39.19 36.47 36.64 42.02 43.97
Area 距離 (km)
P28 P29 P30 P31 P32 P33 P34 P35 P36 P37 P38
P1 17.29 19.32 15.18 22.22 25.45 31.87 48.09 59.51 10.17 11.08 24.49
P2 14.67 16.85 12.56 19.21 22.98 29.40 45.62 57.04 13.17 14.46 27.50
The Simulation for the Ideal Optimum Site Model in the Actual Distance …

P3 13.23 16.39 10.86 18.75 22.52 28.94 45.16 56.58 14.43 15.71 28.75
P4 14.93 16.72 9.38 21.16 24.92 31.34 47.57 58.99 17.50 18.78 31.82
P5 15.12 14.89 13.01 17.79 21.02 27.44 43.67 55.09 14.15 14.73 28.47
P6 12.38 16.17 7.94 21.64 25.41 31.83 48.05 59.47 18.62 19.90 32.94
P7 13.84 14.67 11.73 17.04 20.80 27.23 43.45 54.87 16.13 16.72 30.46
P8 10.38 14.21 5.94 18.34 23.37 30.09 46.67 56.07 19.11 20.39 33.43
(continued)
107
Table 5 (continued)
108

Area 距離 (km)
P28 P29 P30 P31 P32 P33 P34 P35 P36 P37 P38
P9 8.34 12.13 5.38 16.53 21.55 28.27 46.45 54.26 19.38 20.66 33.70
P10 10.78 13.36 8.67 16.22 21.06 27.48 43.70 53.95 18.73 19.31 33.05
P11 8.98 11.01 7.30 14.54 19.56 26.28 43.09 52.27 20.09 20.67 34.41
P12 7.41 9.48 6.15 16.13 21.16 27.88 46.06 53.86 21.85 23.13 36.17
P13 7.82 8.53 7.88 12.93 17.96 24.67 42.85 50.66 21.40 21.99 35.73
P14 7.93 7.67 8.25 14.47 19.49 26.21 44.39 52.20 22.70 23.28 37.02
P15 9.82 8.18 9.46 12.58 17.60 24.32 41.59 50.31 20.93 21.51 35.25
P16 11.12 10.20 10.17 14.58 19.42 25.84 42.06 52.31 17.83 18.41 32.15
P17 10.05 6.46 10.23 10.86 15.88 22.60 40.78 48.59 22.45 24.38 36.77
P18 10.86 7.50 11.75 11.87 14.55 23.61 39.98 49.60 21.15 23.08 35.47
P19 10.46 4.96 12.40 9.36 14.38 21.10 39.28 47.09 23.97 25.90 38.30
P20 12.92 10.55 12.55 14.35 17.76 24.18 40.41 52.08 18.42 19.00 32.74
P21 14.20 8.73 14.03 10.38 13.06 22.12 38.06 48.11 22.64 23.37 36.96
P22 13.54 6.19 15.49 7.83 12.85 19.57 37.75 45.56 24.60 25.33 38.92
P23 15.47 7.99 17.42 6.50 9.64 18.24 33.81 44.23 25.25 25.98 39.57
P24 9.10 15.11 3.93 21.77 26.79 33.51 51.69 59.50 23.26 24.55 37.58
P25 5.93 10.37 3.82 17.03 22.05 28.77 46.95 54.76 22.37 23.65 36.69
P26 2.02 8.53 6.36 15.09 21.03 27.75 45.93 53.74 27.46 28.74 41.78
P27 5.04 5.96 9.44 12.51 18.46 25.18 43.35 51.16 29.86 31.79 44.18
P28 0.00 10.27 6.06 16.83 22.77 29.49 47.67 55.48 27.51 28.79 41.83
P29 11.40 0.00 13.71 10.04 15.06 21.78 39.96 47.77 28.39 30.31 42.71
(continued)
A. Y. Y. Chen and Y. Karasawa
Table 5 (continued)
Area 距離 (km)
P28 P29 P30 P31 P32 P33 P34 P35 P36 P37 P38
P30 5.47 13.38 0.00 20.04 25.06 31.78 49.96 57.77 24.79 26.07 39.11
P31 18.38 9.97 20.78 0.00 5.87 12.59 30.77 38.58 31.26 31.99 43.57
P32 23.17 15.30 25.80 5.87 0.00 7.59 23.75 33.74 35.08 35.81 44.41
P33 29.89 22.02 32.52 12.59 7.56 0.00 17.29 27.47 43.16 43.89 50.83
P34 48.07 40.20 50.70 30.77 23.72 17.29 0.00 34.45 53.90 47.18 44.28
P35 55.84 47.98 58.48 38.55 33.71 27.47 34.45 0.00 69.11 69.84 65.06
P36 27.17 28.87 25.06 31.77 35.00 41.42 57.15 69.07 0.00 10.27 16.39
P37 27.84 29.26 25.73 32.05 35.81 42.23 46.23 69.88 10.12 0.00 17.62
P38 41.46 42.89 39.35 47.22 47.90 54.32 44.28 65.06 16.38 17.62 0.00
The Simulation for the Ideal Optimum Site Model in the Actual Distance …
109
110 A. Y. Y. Chen and Y. Karasawa

simulation of the optimum route, here, we estimated from {lk to produce the direct
delivery cost.
We analyzed the 38 map areas with the actual distance method with the Ideal
type, from one to five places to be the delivery centers. The yardstick of selecting the
optimum sites is the cost minimum strategies, which is to allocate the delivery fee
and inventory fee of the delivery center. Namely, the optimum sites mean to allocate
the total expenses into the inventory fee and delivery fee from one to five places
to choose the minimum cost center, which is referred to as the so-called optimum
delivery center. The process is to allocate the inventory fee and delivery fee to each
center, then to make a ranking for each center from the total cost to choose the most
minimum one.
On the other hand, if the district area of one place is common, then the Ideal type
five optimum delivery centers sites are P2, P9, P18, P26 and P32. The result of the
one location optimum site of a delivery center, the Ideal type has P11, the P11 site
is near the present delivery center P9. Therefore, the present center is the optimum
site, too. Due to the fact that the unavailable areas are small, the present delivery
center is P9 which is near Toyogawa city. Finally, all the P2 and P26 of five locations
optimum sites are overlapped. This is the balance for two places, such as P11 and
P9 that are in an overcrowded area, and P32 and P33 that are in an underpopulated
area. The Ideal type five locations of optimum sites is {lk4,614,433 (Table 6), this
numerical data is only calculated on the lk base, and it should be aware that this does
not mean the cost minimum. However, when we discuss the optimum sites, it means
the minimum cost for the firm.

4 Examination

Regarding all of delivery centers where were simulated in {lK, and deciding for the
layout of warehouse and scale, for calculating a storage cost, including the operational
cost, moreover, about all delivery centers, we estimate the average delivery charges,
setting a delivery route and simulating, and working out the all delivery-related
delivery expenses, it is necessary for calculating the storage costs of each centers.
In our study, we also select the optimum sites based on the amount of money-based,
after calculating the cost (Table 7). As a results, the three places of delivery centers
became the first place conclusively, and, as for the economical expectation effect, it
is clear for reducing the current cost about 30%.

5 Conclusion

Our research is significant supporting systems for the real decision making of the
enterprises, the proposal of a moderate price in a small area for a distribution center
in Japan, the distribution center has adopted a mono price policy for the delivery
The Simulation for the Ideal Optimum Site Model in the Actual Distance … 111

Table 6 Present {lk calculate delivery fee for ideal type of delivery center in 38 actual distance
Centers {lk Compare to Delivery fee
present {lk%
One location P11 First 9,576,888 80.1 15,000,206
Subtotal 9,576,888 80.1 15,000,206
Two locations P11 First 5,901,167 49.3 9,242,953
P32 Second 1,441,435 12.1 2,257,709
Subtotal 7,342,602 61.4 11,500,661
Three locations P2 First 1,127,907 9.4 1,766,632
P11 Second 3,682,648 30.8 5,768,103
P32 Third 1,441,435 12.1 2,257,709
Subtotal 6,251,990 52.3 9,792,444
Four locations P3 First 1,803,880 15.1 2,825,403
P18 Second 1,433,466 12.0 2,245,227
P26 Third 842,229 7.0 1,319,177
P32 Fourth 1,211,592 10.1 1,897,707
Subtotal 5,291,167 44.2 8,287,514
Five locations P2 First 1,074,196 9.0 1,682,505
P9 Second 609,491 5.1 954,641
P18 Third 1,090,883 9.1 1,708,642
P26 Fourth 628,272 5.3 984,058
P32 Fifth 1,211,592 10.1 1,897,707
Subtotal 4,614,434 38.6 7,227,553
Total 33,077,081 276.5 51,808,379
Remark 1 Delivery fee =({lk ÷ 11,961,020) × 18,734,454
Remark 2 Present delivery center’s address is included on P9 area

Table 7 Simulation results for the ideal type of optimum sites on 38 map locations
112 A. Y. Y. Chen and Y. Karasawa

pricing of its products, although with contradiction over the years and in other aspects,
with a delivery price to end users that has been substantially increasing, especially
because of the cost which has grown due to the increased fuel charges. As a result, a
reevaluation of the delivery price setting has been urgently required at this time by top
management and a research project. Therefore, establishing the acceptable delivery
price setting method has now started. A new price setting project is required to take
into considerations several factors, including the new price setting theory, innovative
price setting concept, contribution to the whole company’s business activities, and so
on. In order to justify and evaluate the newly proposed price, market price method,
tariff price, cost plus α method, and so on, respectively.
This research is focused on how the optimum site will affect the delivery unit
price in tentative research. Our simulation for an optimum site selection proposal
was analyzed on the computer to get final solution, multi-location type of optimal
distribution site selection as well as optimum site selection to get gravity points by
zone based, instead of using the simulation model easily, we incorporate the wisdom
that accorded in a model and a system practically, that was able to make introducing
the points clearly.
It is unexpected joy, if our research can contribute to a young researcher for a
bridge function between the theory and the practice. We also apologize for cutting
the details from relations of the space.

References

Chen AYY, Karasawa Y, Aiura N, Suzuki K, Wakabayashi K (2014) Literature study overseas on
SCM strategy with a state art SCM strategy model. In: Golinska P (ed) Logistics operations,
supply chain management and sustainability. Springer, pp 201–226
Chen AYY, Sato T, Karasawa Y, Wakabayashi K, Toyotani J (2016a) A basic research on a delivery
cost settings to end users supported by the optimum site selection model. Indonesia, ICLS. 47
Chen AYY, Sato T, Karasawa Y, Wakabayashi K, Toyotani J (2016b) A practical on optimum site
selection in real world. Indonesia, ICLS, 77
Chen AYY, Sato T, Karasawa Y, Toyotani J, Wakabayashi K (2016c) A basic research on a delivery
cost. In: Proceedings of 19th Nihon logistics system association, pp 136–138
Solving a Location-Allocation-Sizing
Problem Using Differential Evolution
Algorithm: A Case Study of Agricultural
Water Resources in North-Eastern
Thailand

Rerkchai Srivoramas , Ponglert Sangkaphet , Chutchai Kaewta,

Rapeepan Pitakaso , Kanchana Sethanan ,
and Natthapong Nanthasamroeng

1 Introduction

Drought is one of the global problems that tend to increase in priority every year.
The United Nations has revealed that many regions of the world have encountered
severe aridity due to climate change. In 2020, water scarcity affected more than
one billion people worldwide (United Nations 2020). There are three categories of
drought, namely, meteorological, hydrological, and agricultural drought (Li et al.
2021). Agricultural drought has mainly affected crop production, which contributes
to food security, social economy, and stability (Zhang et al. 2021a, b).

R. Srivoramas
Faculty of Engineering, Ubon Ratchathani University, Ubon Ratchathani, Thailand
e-mail: [email protected]
P. Sangkaphet · C. Kaewta · R. Pitakaso
Faculty of Computer Science, Ubon Ratchathani Rajabhat University, Ubon Ratchathani, Thailand
e-mail: [email protected]
C. Kaewta
e-mail: [email protected]
R. Pitakaso
e-mail: [email protected]
K. Sethanan
Faculty of Engineering, Khon Kaen University, Khon Kaen, Thailand
e-mail: [email protected]
N. Nanthasamroeng (B)
Artificial Intelligence Optimization SMART Laboratory, Faculty of Industrial Technology, Ubon
Ratchathani Rajabhat University, Ubon Ratchathani, Thailand
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 113
P. Golinska-Dawson et al. (eds.), Smart and Sustainable Supply Chain
and Logistics—Challenges, Methods and Best Practices, EcoProduction,
https://doi.org/10.1007/978-3-031-15412-6_7
114 R. Srivoramas et al.

Thailand is a country located in the tropical area of Southeast Asia. The proportion
of land used for agriculture is 46.56%. Moreover, farmers make up 38% of Thailand’s
population. Therefore, aridity has inevitably affected the agricultural productivity of
Thailand. The overall crop area in Thailand is equal to 23.8 billion square meters, and
43% of the country’s crop areas are in the north-eastern region (Ministry of Agricul-
ture and Cooperatives of Thailand 2020). Ubon Ratchathani, one of the large areas
in the region, has more than 8.56 billion square meters of cropland. The population
size of Ubon Ratchathani is 1,686,571 people, and 48.37% are in the agricultural
sector (National Statistics Organization of Thailand 2021). The Water Crisis Preven-
tion Centre of Thailand (2021) reported that 534,201 m2 in Ubon Ratchathani face
a medium to high risk of aridity.
Thailand’s government has supported many projects to help farmers keep water
on their lands. However, the appropriate location and size of the reservoirs need to
be considered. Determining the optimal size and location of agricultural reservoirs
will mitigate the aridity problem and reduce the government’s budget for drought
compensation. Thus, the objective of this research is to find the optimal solution for
both the location and size of agricultural water reservoirs in Ubon Ratchathani.
Location problems, known as facility location problems, are classic problems
in the field of operations research. Inappropriate location decisions increase capital
costs and affect customer satisfaction. Alfred Weber (1909) was the first researcher
to publish research on solving the warehouse location problem. After his publication,
many location-related problems were published in several research areas for facilities
such as factories, distribution centers, fire stations, etc. (Moellmann and Thomas
2019; Yao et al. 2019; Zhang et al. 2021a, b).
Later, the facility location problem was extended to combinatorial problems,
including location-routing, location-inventory, location-allocation, etc. The objec-
tive of the facility location-allocation problem (FLP) is to find the optimal subset
from a given set of candidate facilities and allocate demand nodes to the proper
facilities (Chandra et al. 2021). In addition, location-sizing is also a combinatorial
problem widely used in many applications, such as distributed generators in micro-
grids (Kizito et al. 2021), emergency medical service stations (Liu et al. 2019), and
warehouses (Kalfakakou and Tsouros 2001).
However, water irrigation problems have normally been the solution to agricul-
tural drought problems. Many researchers have tried to solve the aridity problem
using irrigation techniques (Li et al. 2020; Haavisto et al. 2019; Naghdi et al. 2021;
Xiea et al. 2018). In addition, Luís and Cabral (2021) applied the Geographical Infor-
mation System (GIS) with Analytic Hierarchical Process (AHP) to select appropriate
locations for small dams/reservoirs in Mozambique and found that only 7% of areas
were suitable for the construction of small dams.
In this research, the location-allocation-sizing problem was introduced. A real case
study in Ubon Ratchathani province was used to compare the effectiveness of the
proposed algorithms. There are 219 subdistricts in Ubon Ratchathani province, each
of which was not only a candidate node for the construction of an agricultural water
reservoir but also a demand node for receiving water from a constructed reservoir
(Fig. 1).
Solving a Location-Allocation-Sizing Problem Using Differential … 115

(a) Basin map (b) Cropland map

Fig. 1 Map of subdistricts and land used in Ubon Ratchathani province

Data collected from each node consisted of x and y coordinates, the height above
sea level, and the average crop water requirements. Estimated construction costs
for water reservoirs, irrigation systems, and water receiving systems were calcu-
lated following the criteria for calculating median prices of irrigation construction
(Comptroller General’s Department 2017). There are three sizes of water reservoirs
in this research: small, medium, and large. The capacity and limitation of irrigation
distance for each reservoir size are shown in Table 1
The algorithms used in previous research for solving location and sizing problems
were genetic algorithm (GA), particle swarm optimization (PSO), flower pollina-
tion algorithm, simulated annealing (SA), and improved harmony search algorithm
(Abdelaziz et al. 2016; Ali et al. 2016; García-Muñoz et al. 2021; Ghaffarinasab et al.
2018; Suman et al. 2021). The differential evolution algorithm (DE) was applied to
solve the LASP in this research. DE is a type of population-based metaheuristic

Table 1 Size, capacity, and

Size of reservoir Capacity (Million Distance limitation
limitation of reservoir
square metre) (Kilometre)
Small 10 10
Medium 50 60
Large 100 100
116 R. Srivoramas et al.

algorithm and is currently one of the most interesting and impressive evolutionary
algorithms. Normally, it has five general steps: (1) generate an initial solution; (2)
perform a mutation process; (3) perform a recombination process; (4) perform a
selection process; and then (5) repeat steps (2)–(4) until termination conditions, such
as computational time or maximum number of iterations, are met.
DE has been successfully applied in several fields, such as production scheduling
(Pitakaso 2015; Pitakaso and Sethanan 2015), manufacturing problems (López et al.
2003), vehicle scheduling problems (Liao et al. 2012), and vehicle routing problems
(Hou et al. 2010). We developed an efficient DE for solving the LASP to minimize
the cost of solving the aridity problem in the north-eastern region of Thailand, partic-
ularly in Ubon Ratchathani province. This paper is organized as follows. In Sect. 2,
the mathematical model representing the location-allocation-sizing problem (LASP)
is presented. In Sect. 3, the proposed methodologies are presented, and the computa-
tional framework and results are presented in Sect. 4. The conclusion is summarized
in Sect. 5.

2 Mathematical Formulation

In this section, the mathematical model of LASP for agricultural water resources is
introduced.
Sets
I Set of agricultural water resource nodes i
J Set of demand nodes j
K Set of water resource sizes k
Parameters
f k Cost of constructing agricultural water resources with size k (THB).
Uk Capacity of agricultural water resources with size k(m 3 )
v Cost of construct irrigation system per distance from agricultural water resources
i to demand node j (THB)
w j Crop water requirement at node j (m 3 )
bi j Hindrance value of water shed
di j Distance from agricultural water resource i to demand node j (m)
qi j Water quantity from water resource i to demand node j (m 3 )
m k Maximum distance of water flow from water resource with size k (m)
s Subsidiary cost per aridity risk unit (THB)
ri Aridity risk in area of node i
h i Altitude of node i (m)
h j Altitude of node j (m)
Solving a Location-Allocation-Sizing Problem Using Differential … 117

Decision Variables
(
1,if node i was selected to be agricultural water resources with size k
xik
0,otherwise
(
1,if water from node i was assigned to demand node j
yi j
0,otherwise
Objective Function
{{ {{ { {{
Min Z = f k xik + v di j yi j − s ri xik + yi j ci j (1)
i∈I k∈K i∈I j∈J i∈I i∈I j∈J

Subject to

qi j = yi j w j , i ∈ I, j ∈ J (2)

{{
xik ≤ 1 (3)
i∈I k∈K
{{
di j yi j ≤ m k xik , ∀i∀ j (4)
i∈I k∈K
{{
yi j ≥ 1 (5)
i∈I j∈J
{{ {{
qi j ≤ Uk xik (6)
i∈I j∈J i∈I k∈K

yi j h j ≤ yi j h i , i ∈ I, j ∈ J (7)

yi j ≤ bi j , i ∈ I, j ∈ J (8)

Objective function (1) is to minimize the cost of constructing the agricultural water
resources. The cost terms in objective function consisted of fixed cost of installation
of reservoir, cost of irrigation, subsidiaries cost for aridity risk, and receiver cost
respectively. Constraint (2) is used to ensure that the volume of water supplied from
node i to node j equals crop water requirements at node j. Constraint (3) is used to
limit each water resource i to only one size k. Constraint (4) is used to ensure the
desired distance of water supply from node i to node j does not exceed the maximum
distance of water flow from a water resource with size k. Constraint (5) is used
to ensure that each demand node j receives water from at least one supply node i.
Constraint (6) is used to ensure that the volume of water supply from node i to node j
does not exceed the capacity of the water resource with size k. Constraint (7) is used
to ensure that the water flows from a higher altitude to a lower altitude. Constraint (8)
is used to limit the flow if the watershed blocks the waterway from node i to node j.
118 R. Srivoramas et al.

3 Methodology

In this study, the differential evolution algorithm was modified to solve the LASP
problem. The differential evolution algorithm (DE) is composed of five general steps,
which are: (1) generate the initial vectors, (2) perform a mutation process, (3) perform
a recombination process, (4) perform a selection process, and (5) repeat steps (2)–
(4) until the termination condition is met. The DE used to solve the LASP can be
explained by the steps described below.

3.1 Generate a Set of Initial Vectors

The vector for the representation of the LASP was designed as a 1 × WP vector,
where WP is the number of work packages. A set of initial sample vectors was
encoded, as shown in Table 2, with 10 nodes of candidate locations and 5 WPs.
A decoding method was developed to obtain the solution of the proposed problem,
as explained below.
The decoding method
The decoding method is composed of five steps: (1) sort the probability of the WP
in ascending order; (2) select the value node with the lowest probability as the water
reservoir; (3) establish the probability criteria for assigning the size of the water
reservoir; (4) assign the size of the reservoir according to its probability by using
criteria established in step (3); and (5) assign demand nodes to the selected water
reservoir. The conditions that need to be considered include (1) the distance limitation
for irrigation, (2) the crop water requirements of assigned demand nodes, which must
not exceed the capacity of water reservoirs in supply nodes, and (3) the height above
sea level of supply nodes, which must be higher than that of the demand nodes.
Details of sample nodes, both supply and demand nodes, are shown in Table 3,
including crop water requirements (CRWs), drought risk, and the height above sea
level. CRWs of each node were calculated from the average CRWs of rice, which

Table 2 Example of vector used in the proposed method

Vector Node
1 2 3 4 5 6 7 8 9 10
1 0.10 1.00 0.43 0.23 0.49 0.72 0.50 0.92 0.30 0.23
2 0.41 0.07 0.48 0.50 0.57 0.86 0.28 0.21 0.71 0.68
3 0.88 0.91 0.48 0.66 0.93 0.13 0.21 0.21 0.43 0.02
4 0.84 0.60 0.56 0.42 0.87 0.13 0.56 0.77 0.13 0.34
5 0.36 0.34 0.65 0.30 0.77 0.88 0.56 0.85 0.54 0.68
Solving a Location-Allocation-Sizing Problem Using Differential … 119

Table 3 Detail of the candidate nodes

Node no. Crop water requirements (for demand node) Drought risk Height from sea level
1 25.77 0.85 144
2 32.67 0.85 184
3 52.81 0.85 149
4 58.24 0.85 190
5 34.90 0.85 183
6 31.65 0.65 164
7 47.02 0.65 164
8 24.17 0.65 148
9 19.89 0.65 164
10 20.63 0.65 135

Table 4 Detail of the water

Size of water Criteria Capacity Distance
reservoir
reservoir probability for (Million limitation
selection square metre) (Kilometre)
Small 0–0.33 10 10
Medium 0.34–0.66 50 60
Large 0.67–1.00 100 100

was the main crop in the area, multiplied by crop area. The drought risk of each node
was also obtained from the average aridity statistics data in the area.
Table 4 shows the details of the water reservoir size, including the probability
criteria for sizing selection, reservoir capacity, and distance limitation for irrigation.
All information was inferred from the irrigation system manual of the Royal Irrigation
Department, Ministry of Agriculture and Cooperatives.
From Table 5, nodes 2, 4, 7, and 8 were selected to be the water reservoirs. Node 2,
which has a probability value of more than 0.67, was assigned to be the large reservoir
and supplies water to node 2 (itself), node 5, and node 6. The construction of the
large water reservoir is 1580 million Baht. The irrigation cost is 7.91 million Baht
per kilometer. In this case, node 2 distributes its water to nodes 5 and 6 with a total
distance of 20.47 km; therefore, the irrigation cost is 161.92 million Baht. Subsidiary
cost is the money that the government pays to farmers affected by a drought disaster
in their area. The recent average drought subsidiary rate is 22.26 million Baht per
node. However, the amount of the subsidiary cost depends on the node’s drought risk,
which is calculated from the probability of drought in the area. In node 2, the drought
risk is 0.85, so the subsidiary cost is equal to 18.92 million Baht. Receiver cost is the
cost that nodes not selected to be water reservoirs have to pay for the construction
of the receiving pond and necessary equipment. For geological engineering reasons,
receiver cost is not the same for each node. In node 2, the receiver cost is 98.15. The
total cost of node 2 selected to be the large water reservoir and supply water to nodes
120 R. Srivoramas et al.

Table 5 Result of the assignment

Water Size Supply Construction Irrigation Subsidiaries Receiver Total
reservoir to node cost cost cost cost cost
2 Large 2, 5, 6 1580.00 161.92 18.92 98.15 1821.15
4 Large 4, 9, 10 1580.00 417.57 18.92 108.40 2087.05
7 Large 3, 7 1580.00 200.91 14.47 45.96 1812.40
8 Medium 1, 8 790.00 260.32 14.47 20.59 1056.44
Grand total cost 5530.00 1040.72 66.78 273.10 6777.04

5 and 6 is 1,821.15 million Baht, which is calculated by the sum of the construction
cost, irrigation cost, and receiver cost. Then, the subsidiary cost is subtracted from
the total. Nodes 4, 7, and 8 follow the same procedure explained above, and the
sum of the total cost for each node is the grand total cost of this sample solution.
The procedure for the decoding method used in this article is shown in Algorithm 1
(Fig. 2).

Fig. 2 Pseudo code of the decoding method

Solving a Location-Allocation-Sizing Problem Using Differential … 121

3.2 Perform Mutation Process

The transformation of the target vector into the mutant vector shown in Table 2 is
called the mutation process. The mutation process is implemented following Eq. (9):

Vi, j,G = X r1 , j,G + F(X r2 , j,G + X r3 , j,G ) (9)

where r 1 , r 2 , and r 3 are the indices of randomly selected vectors; F is a scaling factor,
which is set to 0.8 (Qin et al. 2009); i represents the vector number; i = 1, 2,…, NP;
and j is the position of a vector when j = 1, 2,…, D.

3.3 Perform the Recombination Process

The transformation of the mutant vector into the trial vector is performed by a recom-
bination process. In this process, Eq. (10) is applied, where Vi, j,G is the mutant vector,
X i, j,G is the target vector, and Ui, j,G is the trial vector. In this formula, which was
presented by Pitakaso and Sethanan (2015), rand bij1 is random number 1 of vector
i, position j, and randbij1 is random number 2 of vector i, position j.
(
Vi, j,G when j ≤ randbi, j,1 and j ≥ randbi, j,2
Ui, j,G = (10)
X i, j,G when randbi, j,1 < j < randbi, j,2

3.4 Perform the Selection Process

Finally, before the vector can proceed to the next iteration, the selection process
has to be carried out. The purpose of this process is to select the new target vector.
The candidate for the next target vector is the current target (X i, j,G ) and the current
trial vector (Ui, j,G ). The selection is executed using Eq. (11), where f (Ui, j,G ) and
f (X i, j,G ) are the objective functions of the trial vector and target vector, respectively.
(
pr e Ui, j,G if f (Ui, j,G ) ≤ f (X i, j,G )
X i, j,G+1 = (11)
X i, j,G otherwise
122 R. Srivoramas et al.

3.5 Repeat Steps (Sects. 3.2–3.4) Until Termination

Condition is Met

In this research, the termination condition was set as the computational time (5–
20 min, depending on the size of the test problem). The proposed differential
evolution algorithm procedure is shown in algorithm 2 (Fig. 3).
To compare the efficiency of the proposed DE algorithm, the genetic algorithm
(GA) was applied to solve the same LASP problem for the same real-world scenario.
The GA is composed of five general steps, which are: (1) generate an initial solution,
(2) perform a crossover process, (3) perform a mutation process, (4) perform an
evaluation process, and (5) repeat steps (2)–(4) until the termination condition is
met. The proposed GA procedure is shown in algorithm 3 (Fig. 4).

Fig. 3 Pseudo code of differential evolution algorithm

Solving a Location-Allocation-Sizing Problem Using Differential … 123

Fig. 4 Pseudo code of genetic algorithm

4 Results

In this research, differential evolution and genetic algorithms were coded with Python
and tested using an Intel(R) Core (TM) i7-7500U CPU @ 2.70 GHz, 2904 MHz,
two Core(s), and four logical processors. The algorithms were tested on a real-world
case study of 219 nodes of candidate locations for water reservoirs. The results of
the study are shown in Table 6. The termination condition of DE and GA was set
to be the computational time. The execution time was set to vary from 5 to 30 min,
124 R. Srivoramas et al.

Table 6 Computational result of the test instances

Number of Result obtained from GA Result obtained from DE
candidate Number of Average Total cost Number of Average Total cost
nodes water computational (Million water computational (Million
reservoirs time (sec) Baht) reservoirs time (sec) Baht)
(nodes) (nodes)
219 119 (S = 86.2365 120443.99 94 (M = 2323.36 163818.41
6, M = 79, 67, L =
L = 34) 27)

depending on the size of the problem. The results of the experiment are shown in
Table 6.
As indicated in Table 6, DE outperformed the GA in both the computational time
and objective function value. DE consumed 25.93% less computational time than
GA. The objective value of DE was also 16.44% better than that of GA. The total
number of nodes selected to be water reservoirs differed between DE and GA. The
DE resulted in a smaller number of water reservoirs than GA, which contributed to
a major proportion of the total cost and the cost of construction. In addition, both
algorithms tried to open more medium-sized water reservoirs than large ones because
the latter doubled the cost of construction. The performance of DE and GA is plotted
in Fig. 5.
The result from DE is plotted in the Google Earth map in Fig. 6, which shows
94 locations that were selected for opening agricultural water reservoirs with two
different sizes.
In Fig. 6, the location pin symbols represent the candidate locations selected for
the water reservoirs. Pins with ‘A’ represent a large reservoir, and ‘B’ represents a
medium reservoir. Thumbtacks are used to represent demand nodes. The numbers
on the pins and thumbtacks represent each node by subdistrict. The water reservoir
locations are dispersed around the province, but there are some clusters of receivers in
Muang Ubon Ratchathani and Warin Chamrab districts. This clustering phenomenon
occurred because these two districts have a low risk of drought, and a large proportion
of the area is not cropland; therefore, the cost of subsidiaries to subtract from the
total is less attractive compared with other nodes.

100 86.24 130000 120443.99

Objective value
Computational time

(Million Baht)

80 63.88 120000
60
110000 100642.05
(sec)

40
20 100000
0 90000
GA DE GA DE
Proposed algorithm Proposed algorithm

(a) Computational time of DE and GA (b) Objective value of DE and GA

Fig. 5 Comparison of performance between DE and GA

Solving a Location-Allocation-Sizing Problem Using Differential … 125

Fig. 6 Optimal result of the location-allocation-sizing problem

5 Conclusion

In this paper, a mathematical model of LASP for agricultural water reservoirs is

introduced. The DE and GA were applied to solve a real-world scenario involving
219 subdistricts of Ubon Ratchathani, Thailand. The computational results show that
both the objective value and computational time of the proposed DE are better than
those of GA. Compared with GA, the DE consumed 25.93% less computational time,
and its objective value was 16.44% better. The algorithm assigned 94 locations as
water reservoirs to supply water to 219 districts. Two sizes of reservoirs, medium and
large, were chosen and dispersed around the case study area. It can be concluded that
the factors predominantly influencing the location of water reservoirs are the node’s
CWRs and drought risk. Therefore, urban areas with less cropland and low drought
risk tend to be receiver nodes. Future extensions of this research can be carried out in
many aspects. For instance, (1) the subsidiary cost in this research was combined into
a single objective function, so in future research, the model could be transformed into
a multi-objective model, and (2) the area of the study can be narrowed to the micro-
level, from sub-districts to villages, which can be more practically implemented by
the local government organization.

Acknowledgements Authors would like to gratitude the supports from 1) Research Unit on System
Modelling for Industry (Grant No. SMI.KKU 64/06), Faculty of Engineering, Khon Kaen University,
Thailand, 2) Meta-heuristics for Logistics Optimization Laboratory, Faculty of Engineering, Ubon
Ratchathani University, and 3) Department of Logistics Management, Ubon Ratchathani Rajabhat
University.
126 R. Srivoramas et al.

References

Abdelaziz A, Ali E, Abd Elazim S (2016) Optimal sizing and locations of capacitors in radial
distribution systems via flower pollination optimization algorithm and power loss index. Eng Sci
Technol Int J 19(1):610–618. https://doi.org/10.1016/j.jestch.2015.09.002
Ali E, Abd Elazim S, Abdelaziz A (2016) Improved harmony algorithm and power loss index for
optimal locations and sizing of capacitors in radial distribution systems. Int J Electr Power Energy
Syst 80:252–263. https://doi.org/10.1016/j.ijepes.2015.11.085
Chandra S, Sarkhel M, Vatsa AK (2021) Capacitated facility location–allocation problem for
wastewater treatment in an industrial cluster. Comput Oper Res 132:105338. https://doi.org/
10.1016/j.cor.2021.105338
García-Muñoz F, Díaz-González F, Corchero C (2021) A novel algorithm based on the combination
of AC-OPF and GA for the optimal sizing and location of DERs into distribution networks. Sustain
Ener Grids Netw 27:100497. https://doi.org/10.1016/j.segan.2021.100497
Ghaffarinasab N, Motallebzadeh A, Jabarzadeh Y, Kara BY (2018) Efficient simulated annealing
based solution approaches to the competitive single and multiple allocation hub location problems.
Comput Oper Res 90:173–192. https://doi.org/10.1016/j.cor.2017.09.022
Haavisto R, Santos D, Perrels A (2019) Determining payments for watershed services by hydro-
economic modeling for optimal water allocation between agricultural and municipal water use.
Water Resour Econ 26:100127. https://doi.org/10.1016/j.wre.2018.08.003
Hou L, Zhou H, Zhao J (2010) A novel discrete differential evolution algorithm for stochastic
VRPSPD. J Comput Inform Syst 6(8):2483–2491
Kalfakakou R, Tsouros CC (2001) Determining the size and location of suppliers for a minimum
total distribution route problem. Int J Prod Econ 71(1–3):401–405. https://doi.org/10.1016/s0925-
5273(00)00144-4
Kizito R, Liu Z, Li X, Sun K (2021) Stochastic optimization of distributed generator location and
sizing in an islanded utility microgrid during a large-scale grid disturbance. Sustain Ener Grids
Netw 27:100516. https://doi.org/10.1016/j.segan.2021.100516
Li M, Fu Q, Singh VP, Liu D, Gong X (2020) Risk-based agricultural water allocation under multiple
uncertainties. Agric Water Manag 233:106105. https://doi.org/10.1016/j.agwat.2020.106105
Li Y, Lu H, Yang K, Wang W, Tang Q, Khem S, Yang F, Huang Y (2021) Meteorological and
hydrological droughts in Mekong River Basin and surrounding areas under climate change. J
Hydrol Reg Stud 36:100873. https://doi.org/10.1016/j.ejrh.2021.100873
Liao T, Egbelu P, Chang P (2012) Two hybrid differential evolution algorithms for optimal inbound
and outbound truck sequencing in cross docking operations. Appl Soft Comput 12(11):3683–
3697. https://doi.org/10.1016/j.asoc.2012.05.023
Liu K, Li Q, Zhang ZH (2019) Distributionally robust optimization of an emergency medical
service station location and sizing problem with joint chance constraints. Transp Res Part B
Method 119:79–101. https://doi.org/10.1016/j.trb.2018.11.012
López Cruz I, van Willigenburg L, van Straten G (2003) Optimal control of nitrate in lettuce by a
hybrid approach: differential evolution and adjustable control weight gradient algorithms. Comput
Electron Agric 40(1–3):179–197. https://doi.org/10.1016/s0168-1699(03)00019-x
Luís ADA, Cabral P (2021) Small dams/reservoirs site location analysis in a semi-arid region of
Mozambique. Int Soil Water Conserv Res 9(3):381–393. https://doi.org/10.1016/j.iswcr.2021.
02.002
Ministry of agricultural and cooperatives of Thailand (2020) Thailand’s essential agricultural and
cooperatives information 2020. Office of Planning, Ministry of agricultural and cooperatives of
Thailand
Moellmann J, Thomas VM (2019) Social enterprise factory location and allocation model: small
scale manufacturing for East Africa. Soc Econ Plann Sci 68:100694. https://doi.org/10.1016/j.
seps.2019.02.009
Solving a Location-Allocation-Sizing Problem Using Differential … 127

Naghdi S, Bozorg-Haddad O, Khorsandi M, Chu X (2021) Multi-objective optimization for alloca-

tion of surface water and groundwater resources. Sci Total Environ 776:146026. https://doi.org/
10.1016/j.scitotenv.2021.146026
National Statistics Organization of Thailand (2021) Ubon Ratchathani provincial statistical report.
http://ubon.nso.go.th/index.php?option=com_content&view=article&id=382&Itemid=507
Pitakaso R (2015) Differential evolution algorithm for simple assembly line balancing type 1
(SALBP-1). J Ind Prod Eng 32(2):104–114. https://doi.org/10.1080/21681015.2015.1007094
Pitakaso R, Sethanan K (2015) Modified differential evolution algorithm for simple assembly line
balancing with a limit on the number of machine types. Eng Optim 48(2):253–271. https://doi.
org/10.1080/0305215x.2015.1005082
Suman GK, Guerrero JM, Roy OP (2021) Optimisation of solar/wind/bio-generator/diesel/battery
based microgrids for rural areas: a PSO-GWO approach. Sustain Cities Soc 67:102723. https://
doi.org/10.1016/j.scs.2021.102723
United Nations (n.d.) 2010–2020: UN decade for deserts and the fight against desertification. UN
decade for deserts and the fight against desertification. Retrieved July 5, 2021, from https://www.
un.org/en/events/desertification_decade/whynow.shtml
Water Crisis Prevention Centre of Thailand (2021) Flood and drought data of Thailand. http://mek
hala.dwr.go.th/download-cate.php?txtdoccate=28
Weber A (1909) Über den Standort der Industrien, Tübingen. Theory of the location of industries,
University of Chicago Press (English translation by Friedrich CJ (1929)).
Xie Y, Xia D, Ji L, Huang G (2018) An inexact stochastic-fuzzy optimization model for agricultural
water allocation and land resources utilization management under considering effective rainfall.
Ecol Ind 92:301–311. https://doi.org/10.1016/j.ecolind.2017.09.026
Yao J, Zhang X, Murray AT (2019) Location optimization of urban fire stations: access and service
coverage. Comput Environ Urban Syst 73:184–190. https://doi.org/10.1016/j.compenvurbsys.
2018.10.006
Zhang S, Chen N, She N, Li K (2021a) Location optimization of a competitive distribution center
for urban cold chain logistics in terms of low-carbon emissions. Comput Ind Eng 154:107120.
https://doi.org/10.1016/j.cie.2021.107120
Zhang Y, Hao Z, Feng S, Zhang X, Xu Y, Hao F (2021b) Agricultural drought prediction in China
based on drought propagation and large-scale drivers. Agric Water Manag 255:107028. https://
doi.org/10.1016/j.agwat.2021.107028
Simulation-Based Application
for Improving Carton Production Process

Nara Samattapapong and Thiti Mhoraksa

1 Introduction

The industrial sector is continually undergoing changes at the moment. Regardless

of how much technology rapidly advances, cartons continue to be used as packaging
for electronics, automobile parts, and a variety of other industries. It continues to be
widely used today because of the qualities that can protect the goods from scratches
and damage, as well as its lightness. According to the case study factory’s analysis,
the company is a contract packaging company. The company receives raw materials
from suppliers and applies made-to-order models to cut them into pieces based on
the customer’s requirements. Each client has specific product requirements, so each
customer’s product is manufactured using a specific production technique. It means
that working systems are arranged in a variety of ways. As a result, the work system
must be structured methodically and the number of machines and employees has
to be suitably increased or decreased to keep production running and the project
has to finish on time. According to the company’s goal, “the company will produce
the package in accordance with the specific needs and inventing the evolution of
materials at a rapid rate in order to reduce the waiting time. The quality and punctu-
ality must be consistent. The objective is to keep all of the partners’ supply chains
running smoothly. New innovations are continuously introduced as effective tools for

N. Samattapapong
Institute of Engineering, School of Industrial Engineering, Suranaree University of Technology,
111 University Avenue, Suranari, Muang Nakhon Ratchasima, Nakhon Ratchasima 30000,
Thailand
e-mail: [email protected]
T. Mhoraksa (B)
Establishment Project Faculty of Integrated Engineering and Technology, Chanthaburi Campus,
Department of Industrial Engineering, Rajamangala University of Technology Tawan-Ok, 131
M.10 Bumratnaradoon Rd, Pluang, Khao Kitchakut, Chanthaburi 22210, Thailand
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 129
P. Golinska-Dawson et al. (eds.), Smart and Sustainable Supply Chain
and Logistics—Challenges, Methods and Best Practices, EcoProduction,
https://doi.org/10.1007/978-3-031-15412-6_8
130 N. Samattapapong and T. Mhoraksa

developing and creating something better, more flexible, lighter, and more environ-
mentally friendly. The research team collaborated together to design the work process
and system placement, so that they could operate more efficiently. It would have the
ability to increase production. The objective was to enhance efficiency and reduce
waste in the carton manufacturing process by integrating simulation techniques into
the process design.

2 Theoretical Framework

Definition of Simulation
There are researchers who have studied and discussed simulations in various
contexts, as follows:
Kelton et al. (2007) suggested that simulation is a method and application that
uses computer software to simulate the behavior of a real system. Pisuchpen (2008)
said that simulation is defined as the process of building a model of a real system
and then performing experiments to understand how the real system behaves. The
experiment’s results will next be evaluated before being used to solve problems
in real-life situations. Surachetkiati (2001) said that the simulation is the process
of designing a model of a real system and the experimental design in this model
and then understand system behavior and estimating the variables important to the
operation within the system.
3 Gen
It is a manufacturing philosophy in practice. This is a concept and practice that
has been established as a model for the manufacturing sector. Based on the prin-
ciple of “don’t trust anyone,” this approach can generate consistency between real
circumstances and scenarios in production. It consists of (1) GENBA is the actual
place, (2) GENBUTSU is the actual production piece, and (3) GENJITSU is the fact
(Chalermjirarat 2006).
7 Waste
Taiichi Ohno, the founder of the Toyota production system, divides waste into
seven categories: overproduction, waiting, transportation, inventory, movement,
unnecessary processes, and waste (Praison 2011).

3 Related Studies

From the relevant research studies, it was found that computer-based simulations
have been applied to the allocation of the number of employees to reduce labor costs
and to improve employee efficiency. Kengpol and Youngswaing (2015) have adopted
the principles of concurrent engineering and simulations used to allocate employees
and adjust the production process of the oil pipe production process. It was found
that the new production process can reduce the number of employees from 4 to 2
Simulation-Based Application for Improving Carton Production Process 131

people. Moreover, it also reduces the machines used in the production process from 2
machines to 1 machine. Therefore, the remaining machines and staff can be allocated
to support the new production line. As a result, there are benefits such as increasing
production capacity by 50%, lowering the investment cost of purchasing pipe bending
machines, and lowering the annual cost of recruiting employees. Samattapapong
(2018) has employed simulation modeling to investigate the problem and to develop
a better cassava starch packing process. The outcomes were compared to those of the
common practice. The best alternative was shown to be capable of reducing testing
time by 19.02%, sealing time by 17.99%, transit time by 47.18%, and moving time
by 34.10%. Furthermore, it was able to cut the workforce by three individuals.
The efficiency and management system have been improved by using a situational
model. Mhoraksa et al. (2020) have used the model to examine the production system
and enhance the drinking water manufacturing process. As a consequence, the infor-
mation and Flexsim® were utilized to construct a model. It was discovered that a
bottleneck existed in the filling process. As a result, three solutions were devised
to address the issue. After putting each alternative to the test, the findings revealed
a 36.61% reduction in manufacturing time. Ghiyasinasab et al. (2018) have also
employed lean approaches and simulations to improve their work. Kusoncum et al.
(2018) have use the model to improve mill yard management that aims to reduce the
time in the system for sugarcane transport vehicles. As a result, two solutions were
devised to address the issue. After putting each alternative to the test, the findings
revealed a 11.39% reduction in the system time. Klinlek et al. (2020) have use the
model to increase the quantity of products and the quality of the working process.
Jarernram and Samattapapong (2018) have used the model to production scheduling
and to search makespan optimization of parallel machine production scheduling, the
production process of this plant is parallel machine. Phanindra et al. (2019) have
used the model to improve its productivity, optimizing it, and increase the overall
efficiency of a Plant. The paper illustrates that the system can be optimized by high
work station utilization through managing bottlenecks with the addition of buffers.
Chanthakhot and Ransikarbum (2021) have used the simulation model that integrates
fire dynamics simulation coupled with agent-based evacuation simulation to eval-
uate the impact of smoke and visibility from fire on evacuee behavior. Ransikarbum
(2020) have used the agent-based simulation software to evaluate traffic problems at
the intersection. Huihui et al. (2016) have used the Flexsim simulation to realize the
storage for a simulation of the operation process and find out the bottleneck existing
in the system according to the simulation results, finally, the bottleneck problem to
optimize the model, and put forward improvement opinions and suggestions. Darayi
et al. (2013) have used the simulation model to study the capacity enhancement
scenarios in a tire manufacturing company located in Iran. Samattapapong (2017)
have used the model to increase efficiency in the warehouse operation. The result
for simulation analysis found that the conveyor belt was a bottleneck in the ware-
house operation. Therefore, many scenarios to improve that problem were generated
and testing through simulation analysis process. The result showed that an average
queuing time was reduced from 89.8% to 48.7% and the ability in transporting the
product increased from 10.2% to 50.9%. Thus, it can be stated that this is the test
132 N. Samattapapong and T. Mhoraksa

method for increasing efficiency in the warehouse operation. Nie and Wang (2019)
have used the Flexsim simulation to improve the operation efficiency of rail mounted
gantry crane and reduce the waiting time of container trains and trucks. Ishak et al.
(2020) have used the simulation model to find out whether the vise production process
time with the number of production targets can be met and is effective. In addition,
He and Hua (2018) have established a model of the enterprise operating system to
measure the service capabilities of the enterprise through the simulation of the model.
Computer-based simulation has been used increasingly in collaboration with lean
manufacturing techniques to manage the number of employees in the production
line and to optimize production processes. For example, Chandrakumar et al. (2016)
used Flexsim® simulation software and a lean manufacturing system to improve
the transfer process’s performance. The issues have been identified, particularly
wait time, line length, and employee idle time. Kumar et al. (2015) have used the
Flexsim for measuring and analysis of performance measures of Flexible manu-
facturing system is applied. And it has been found that the simulation techniques
are easy to analyze the complex flexible manufacturing system. Jarernram (2017)
used Flexsim® simulation software to optimize production scheduling and discover
the best appropriate overall production time for this issue. Furthermore, the issue
was NP-Hard, which meant that finding a solution required a lengthy time. As the
complexity of the issue became larger, it required exponentially longer to come up
with a solution. The findings demonstrated that simulation techniques may be used
to solve difficult and time- consuming issues. Luscinski and Ivanov (2020) have used
the simulation model to developed flexible manufacturing system, then the example
data were used to demonstrate the developed model applicability. “The Ontology
on Flexibility” was applied for evaluation of achieved flexibility of manufacturing
system.
Furthermore, Pawlak (2008) have proposes a new modelling framework for simu-
lating flows of people between suburban areas and the metropolis. The model is based
on a logit relationship used in researches of transport mode choice problems. Rodal-
wski (2006) presented the issue of building a model that reflects the real business
process, as well as simulating the behavior of that model in order to draw conclu-
sions about effectiveness and efficiency of a real business process. Cárdenas et al.
(2018) have used the model to management of a high circulation road connecting two
mainstream cities in Chile. Dobrzyński and Waszczur (2018) have used the model to
show the opportunities of the analysis of the process according to the scenarios and
variants developed in connection with the qualitative assessment process. Kluska
(2021) presented mechanism is the basis for the methodology of automatic simula-
tion modeling of warehouses. It allows significant reduction in simulation models
building duration, and thus a significant reduction in the time of projects consisting in
verifying the concept of spatial arrangement in various projects related to the storage
area. The proposed tool is innovative and useful for practitioners specializing in
simulation modeling and specialists in warehouse design. Due to the organization
and simplification of data structures, it can be implemented in various simulation
modeling environments. Also after implementation, it can be used by people who do
not have advanced simulation skills.
Simulation-Based Application for Improving Carton Production Process 133

4 Methodology

This research is a study of the carton manufacturing process A-01 of a case study:
an establishment in the packaging carton manufacturing industry. The methods for
conducting research were as follows:
(1) Studying the current situation to collect information by studying the procedures,
working methods, factory planning, machine position, production aids, and the
number of staff. The rule of 3 GEN was applied to explore the real workplace,
actual production, and real situation.
(2) Analyzing the information gathered in the first phase and then developing a
model for the current situation.
(3) Creating three alternatives and the model of each alternative.
(4) Proposing the most suitable approach to improvement.

4.1 Information About the Current System

The production of paper boxes A-01, the steps are shown in Fig. 1. The first step is
putting the raw materials into the paper cutter machine to get the required size. After
that, take the cutting pieces to the machine and run them to create an outline that
will be simple to fold into a box. The pieces will be checked again for evaluating
their completeness and they will be separated into three groups. The first group is
unfixable pieces; these products will be discarded. The second group is incomplete
workpieces: they will be repaired. And the last group is complete workpieces: they
can be formed and packaged into boxes then put into bags, 5 boxes per bag, which
can then be kept in a warehouse. The total number of employees is 6, divided into
six section (1 employee per section) that include working on a forklift to transport
and store raw materials or products that must be sent to customers, paper cutting
machine, creasing machine, slotting machine, box forming and packing, and storing.

Analysis of current operational data

From the analysis of the current operation, a large number of pieces in the waiting
line had occurred because there was only one staff in the process of forming the boxes
and adding foam into the box before delivering to the packing process which was a
time consumer. As a consequence, production process was delayed, and employees
who were in charge of keeping products faced waiting times. Furthermore, it was
discovered that the inspection station for cutting was located at the back of the factory,
causing unnecessary transportation, resulting in time wasted on transportation.

Hypothesis
Option 1: Increase the number of employees from one to two for folding and packing
the boxes. To reduce the waiting time for the piece, each piece of work that comes
out would be completed by the work of one person.
134 N. Samattapapong and T. Mhoraksa

Fig. 1 A-01 carton

manufacturing process

Option 2: Increase the number of employees from one to two. The first employee
is responsible for folding, while the second employee is responsible for packing.
The employees were assigned to do different tasks to be more proficient at that task.
Therefore, the employee could work faster and the work would be completed more
quickly.
Option 3: Modify the factory layout by relocating the checkpoints that would be kept
closer to the paper cutter machine. Then, the storage station would be closer to the
Simulation-Based Application for Improving Carton Production Process 135

Processor1 Queue5 Queue4

Queue1 Processor3
Source1
Processor5
Queue3

Processor4
Combiner1

Queue6
Processor2 Queue2 Rack1 Rack2 Source2

Queue7

Rack3

Fig. 2 System representative of the system in the simulation (Top View)

packing station. Thus, it would reduce transportation time and increase the number
of products.

Modeling
The modeling processes are the steps to perform various tasks. It can modify the
situation to try out new operating systems and it does not cause any impact on the
actual operation. It uses computer systems to model and generate alternatives in
order to test the most efficient process. Researchers utilized the Flexsim® software,
which uses real-life data, to build a current functioning model and alternatives. The
following information was utilized in the model’s design which was represented in
Fig. 2:

1. Source1 represents the arrival of raw materials into the production process.
2. Processor1 represents the paper cutter that cut the raw material to the specified
size.
3. Queue1 represents the waiting of pieces before moving to the creasing machine.
4. Processor2 represents the creasing machine. It is used to make a mark which
can be easy to form into a box.
5. Queue2 represents the waiting of pieces before moving to the slotting machine.
6. Processor3 represents the paper slotting machine. This will cut the paper into
grooves at the specified points.
7. Queue3, Queue4 represent the waiting of pieces before moving to the check-
point.
8. Processor4 represents the checking area for evaluating cutting quality.
9. Queue5 represents the paper-forming line.
10. Processor5 represents the area of forming the paper into a box and filling the
foam into the box.
136 N. Samattapapong and T. Mhoraksa

11. Combiner1 represents the packing of 5 boxes per pack.

12. Source2 represents the materials that will be used for packing 5 boxes.
13. Queue6 represents the area of the waiting of items before packing.
14. Queue7 represents the area of the warehouse before delivering to customers
(Fig. 2).

Verification and validation

Verification is concerned with building the model correctly, according to the
conceptual model and its assumptions.
Chi-Square Test for the relationship between the two variables from the number
of boxes produced each day from 30 days of data collection of the current system
and the generated model to verify that the generated model can be representative of
the system.
Figures 4 and 5 show that the generated model can be used to represent the system.

Fig. 3 System representative of the system in the simulation (Perspective View)

Fig. 4 Using Experfit in FlexSim for Chi-Square Test

Simulation-Based Application for Improving Carton Production Process 137

Fig. 5 Take the data obtained from the model when running 30 times and validate all 30 results,
and the result is not reject

Generate Alternative Model

According to the results of the research of the working process using the Flexsim®
software to build a model, the issue arose during the forming and packaging process
because there were numerous waiting lines. There were three different alternatives
to enhance it:
Option 1: Increase production capacity by adding one person to the box folding and
packing station, which would reduce the waiting time of the prior stage (shown in
Figs. 6 and 7).

Fig. 6 Adding one employee for option 1

138 N. Samattapapong and T. Mhoraksa

Fig. 7 System representative of the system in option 1

Fig. 8 Adding one employee in packing process for option 2

Option 2: Increase the number of employees in the packing area by enabling current
employees to fold the boxes. It allowed workers to work more quickly. (As shown
in Figs. 8 and 9).
Option 3: Change the layout of the factory by relocating the intersection check point
to the storage site, the workpiece’s transit distance was reduced (Shown as Figs. 10
and 11).

5 Results and Discussion

From the experiment using the Flexsim® , the test was run for 28,800 seconds that
is 8 hours with the following results: in the traditional system, an average of 701
jobs was submitted, but only 530 were completed, meaning that 74.96% of the total
number of jobs was completed and 177 jobs were still being processed (Table 1).
Simulation-Based Application for Improving Carton Production Process 139

Fig. 9 System representative of the system in option 2

Fig. 10 Factory layout adjustment for option 3

Fig. 11 System representative of the system in option 3

140 N. Samattapapong and T. Mhoraksa

Table 1 Results from the experiment with various scenarios using Flexsim® (tested for
28,800 seconds)
Option Status
Pieces from Source1 Pieces move to Work capacity Work in process
Queue7
Current system 707 530 74.96% 177
Option 1 701 605 86.31% 96
Option 2 681 585 85.90% 96
Option 3 719 555 77.19% 164
Source Own work

Option 1 (adding 1 person to the forming and packaging unit) resulted in an

increase of 11.35% in the number of jobs from the current system to 86.31%. Further-
more, there were 96 items of work in process, down from 45.76% at the present time
(shown in Table 1). For the option 2 (one employee in packing process and one
employee in forming process), this resulted in the number of jobs accounting for
85.90%, an increase of 10.94% from the current system, and 96 pieces of work in
progress, or a 45.76% decrease from the current status (as shown in Table 1). The
last option is to change the factory layout, which resulted in the current scenario in
which the number of jobs accounted for 77.19% and there were 2.23% more jobs
than before, with 164 in-process work items, which was 7.34% fewer than the normal
system (shown in Table 1).

6 Conclusions

The findings of this study on simulation and comparing work productivity and waiting
time for all three alternatives indicated that option 1 is the most suitable option by
adding 1 staff in the forming and packing section. This resulted in the number of
jobs representing 86.31% or an increase of 11.35% from the current situation and the
work in progress decreased by 45.76%, which is quite similar to the second option. In
terms of option 2, the development was planned by adding 1 employee for the packing
section. And the remaining staff could work in the folding and forming process. As
a consequence, the number of jobs increased by 10.94% from the regular operation,
while the work in progress dropped by 45.76% from the regular operation. It was
concluded that adding more staff to work in high-waiting areas would help to increase
the quality of production and to reduce the wasted time. Therefore, the simulation is
used to significantly improve the carton manufacturing process. It can also provide
better alternatives, resulting in more effective decision-making, the capability of the
process, and competitiveness of enterprises.
The results of this study can be applied to further studies in various aspects, such as
modifying the structure of the factory, technology modification, and the appropriate
Simulation-Based Application for Improving Carton Production Process 141

amount of staff and machines for each process. However, choosing the alternative for
the improvement also depends on other factors regarding the company’s situation.
To illustrate, the cost of improvements needs to be suitable with the budget of the
company and the number of employees need to be sufficient for the improvement
plan. Moreover, the improvement method should not affect the other products of
the company. Therefore, it can be concluded that using simulation modeling with
Flexsim® provides a clear visualization of the changes without having any impact
on the real operation.

Acknowledgements I wish to express my deepest gratitude to the case study company for dedi-
cating and providing useful information which is beneficial for this project and further studies. I
would also like to thank all the employees for their cooperation in conducting the research. Moreover,
I am deeply grateful for the support of the Department of Industrial Engineering, Project to estab-
lish the Faculty of Integrated Engineering and Technology, Rajamangala University of Technology
Tawan-Ok, Chanthaburi Campus which provided the funding to support the research publication.

Funding Source Declaration Rajamangala University of Technology Tawan-Ok

References

Cárdenas O, Valencia A, Montt C (2018) Congestion minimization through sustainable traffic

management. A micro-simulation approach. LogForum 14(1):21–31. https://doi.org/10.17270/J.
LOG.2018.260
Chalermjirarat W (2006) The QC problem solving approach solving workplace problems the
Japanese way. Technology Promotion Association (Thailand-Japan), Bangkok
Chandrakumar C, Gowrynathan J, Kulatunga AK, Sanjeevan N (2016) Incorporate lean and green
concepts to enhance the productivity of transshipment terminal operations. Proc CIRP 40:301–
306. https://doi.org/10.1016/j.procir.2016.01.042
Chanthakhot W, Ransikarbum K (2021) Integrated IEW-TOPSIS and fire dynamics simulation for
agent-based evacuation modeling in industrial safety. Safety 7(2):47, 1–21. https://doi.org/10.
3390/safety7020047
Darayi M, Eskandari H, Geiger CD (2013) Using simulation-based optimization to improve perfor-
mance at a tire manufacturing company. Qsci Connect 13:1–12. https://doi.org/10.5339/connect.
2013.13
Dobrzyński M, Waszczur P (2018) Simulation analysis of a production process with selected six
sigma ratios. LogForum 8(1):47–54. Retrieved Jan 03, 2021, from https://www.logforum.net/pdf/
8_1_6_12.pdf
Ghiyasinasab M, Lehoux N, Ménard S, Cloutier C (2018) Using lean techniques and simulation to
improve the efficiency of engineered wood production: a case study in a small factory. Ind Eng
Manag 7(4):1–11. https://doi.org/10.4172/2169-0316.1000269
He H, Hua Z (2018) Analysis of fast food service capability based on Flexsim modeling and
simulation. IOP Conf Ser Mater Sci Eng 394:1–6. https://doi.org/10.1088/1757-899X/394/5/
052005
Huihui S, Xiaoxia M, Xiangguo M (2016) Simulation and optimization of warehouse operation
based on flexsim. J Appl Sci Eng Innov 3(4):125–128. https://doi.org/10.35940/ijrte.B1137.078
2S419
142 N. Samattapapong and T. Mhoraksa

Ishak A, Zubair AF, Cendani AS (2020) Production line simulation in vise using the Flexsim
application. IOP Conf Ser Mater Sci Eng 1003:1–6. https://doi.org/10.1088/1757-899X/1003/1/
012103
Jarernram J (2017) Makespan optimization of parallel machine production scheduling through
simulation technique, Master’s thesis, School of Industrial Engineering, Suranaree University
of Technology. Retrieved Feb 10, 2021, from http://sutir.sut.ac.th:8080/jspui/handle/123456789/
7652
Jarernram J, Samattapapong N (2018) Parallel machine scheduling using simulation software. In:
4th international conference on mechanical engineering and industrial automation held on, pp
150–155
Kelton WD, Sadowski RP, Sturrock DT (2007) Simulation with arena. McGraw-Hill Education,
New York
Kengpol A, Youngswaing W (2015) Application of concurrent engineering and simulations in
productivity improvement of the pipe bending: a case study in an oil pipe manufacturer. J
KMUTNB 25(2):233–242. https://doi.org/10.14416/j.kmutnb.2015.01.005
Klinlek P, Samattapapong N, De Cadenet V (2021) Modeling, increasing efficiency and production
capacity in melamine dish production line. EasyChair preprint no. 4652, pp 1–6. Retrieved Apr
10, 2021, from https://easychair.org/publications/preprint/7rMW
Kluska K (2021) Automatic simulation modelling of warehouses. LogForum 17(1):59–69. https://
doi.org/10.17270/J.LOG.2021.547
Kumar BS, Mahesh DV, Kumar BS (2015) Modeling and analysis of flexible manufacturing system
with FlexSim. Int J Comput Eng Res 5(10):1–6. Retrieved Jan 03, 2021, from https://www.ijcero
nline.com/papers/Vol5_issue10/A05100106.pdf
Kusoncum C, Sethanan K, Putri EP, Neungmacha W (2018) Simulation-based approaches for
processes improvement of a sugar mill yard management system: a case study of the sugar
industry in the central region of Thailand. Eng Appl Sci Res 45(4):320–331. https://doi.org/10.
14456/easr.2018.44
Luscinski S, Ivanov V (2020) A simulation study of industry 4.0 factories based on the ontology
on flexibility with using flexsim® software. Manag Prod Eng Rev 11(3):74–83. https://doi.org/
10.24425/mper.2020.134934
Mhoraksa T, Samattapapong N, Yunyao S, Wongsakul P (2020) Simulation-based application for
improving drinking water production process: a case study of the drinking water factory in
Chanthaburi. RMUTL Eng J 5(2):36–42. https://doi.org/10.14456/rmutlengj.2020.11
Nie X, Wang L (2019) Simulation process design for scheduling mode of railway container terminals
based on Flexsim. IOP Conf Ser Mater Sci Eng 1176:1–5. https://doi.org/10.1088/1742-6596/
1176/5/052012
Pawlak Z (2008) An alternative application of logit modelling in management of metropolitan areas:
the case study of Poznań urban area. LogForum 4(4):1–14. Retrieved Jan 03, 2021, from https://
www.logforum.net/pdf/4_4_2_2008.pdf
Phanindra KD, Ratna PP, Kalamulla MD, Sai KP, Sai PBLN (2019) Analyze the production system
of an body-in-white system through modelling and perform bottleneck, optimization using simu-
lation software. Int J Rec Technol Eng 8(2S4):685–690. https://doi.org/10.35940/ijrte.B1137.078
2S419
Pisuchpen R (2008) Modeling with arena. SE-Education, Bangkok
Praison J (2011) Operation management. SE-Education, Bangkok
Ransikarbum K (2020) Analysis of traffic flow at a red light intersection using computer simulation
technique. Thai J Oper Res 8(2):1–14. Retrieved Mar 05, 2021, from https://ph02.tci-thaijo.org/
index.php/TJOR/article/view/225684/164532
Rodalwski B (2006) Simulation of logistics process (SIMPROCESS). LogForum 2(1):1–15.
Retrieved Mar 03, 2021, from https://www.logforum.net/pdf/2_1_4_06.pdf
Samattapapong N (2017) An efficiency improvement in warehouse operation using simulation
analysis. IOP Conf Ser Mater Sci Eng 273:1–7. https://doi.org/10.1088/1757-899X/273/1/012013
Simulation-Based Application for Improving Carton Production Process 143

Samattapapong N (2018) Productivity improvement of tapioca packing process through simulation

modeling analysis. In: Proceedings of 2018 5th international conference on industrial engineering
and applications (ICIEA), pp 453–457. https://doi.org/10.1109/IEA.2018.8387143
Suracherdkiati W (2001) Computer simulation. Skybook, Pathum Thani
Solving Multi-Echelon Location
Allocation Problem Using Modified
Differential Evolution Algorithm: Case
Study of Agricultural Products
Warehouse in Greater Mekong Subregion

Kiatisak Pranet, Ponglert Sangkaphet , Rapeepan Pitakaso ,

Natthapong Nanthasamroeng , Thanatkij Srichok ,
Kanchana Sethanan , and Peema Pornprasert

1 Introduction

This research aimed to design a transportation network for agricultural products’

international trade among Thailand, Laos, and Cambodia. To that end, firstly, agri-
cultural products must be delivered from the agricultural cooperatives (AC) that

K. Pranet
Meta-Heuristics for Logistics Optimization Laboratory, Faculty of Industrial Technology, Ubon
Ratchathani Rajabhat University, Ubon Ratchathani, Thailand
e-mail: [email protected]
P. Sangkaphet
Meta-Heuristics for Logistics Optimization Laboratory, Faculty of Computer Science, Ubon
Ratchathani Rajabhat University, Ubon Ratchathani, Thailand
e-mail: [email protected]
R. Pitakaso · T. Srichok
Meta-Heuristics for Logistics Optimization Laboratory, Faculty of Engineering, Ubon
Ratchathani University, Ubon Ratchathani, Thailand
e-mail: [email protected]
T. Srichok
e-mail: [email protected]
N. Nanthasamroeng
Meta-Heuristics for Logistics Optimization Laboratory, Faculty of Industrial Technology, Ubon
Ratchathani Rajabhat University, Ubon Ratchathani, Thailand
e-mail: [email protected]
K. Sethanan
Faculty of Engineering, Khon Kaen University, Khon Kaen, Thailand
e-mail: [email protected]
P. Pornprasert (B)
Faculty of Industrial Technology, Department of Logistics Management, Ubon Ratchathani
Rajabhat University, Ubon Ratchathani, Thailand
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 145
P. Golinska-Dawson et al. (eds.), Smart and Sustainable Supply Chain
and Logistics—Challenges, Methods and Best Practices, EcoProduction,
https://doi.org/10.1007/978-3-031-15412-6_9
146 K. Pranet et al.

Fig. 1 Framework of the proposed problem

collected the products from the farmers. The AC will transport the products to the
wholesaler agricultural market (WSAM), which will transport the products to the
border checkpoint (BC). After the BC, the trucks delivering the products from the
WSAM will pass through the border and deliver the products to the end-market in
the specified countries and regions. The framework of this process is shown in Fig. 1.
As shown in Fig. 1, the model of transportation can be formulated as a multi-
echelon transportation problem or multi-level network flow problem. The transporta-
tion problem for agricultural products has been widely studied by many researchers
such as Hsu et al. (2007), who presented the food distribution planning model when
taking into account the stochastic and time-dependent travel times and time-varying
temperature. Osvald and Stirn (2008) addressed the distribution of fresh vegeta-
bles while considering time-dependent travel times. In that research, the quality
degradation-based cost as an objective function was presented. Tarantilis and Kira-
noudis (2001, 2002) studied milk transportation with a fixed fleet. Sethanan and
Pitakaso (2016) presented an algorithm to resolve raw milk transportation while
considering the special case of a truck with multiple compartments used to transport
milk.
The agricultural product is one type of perishable product. Perishable products
need intensive care during transportation to the final destination. Doerner et al. (2008)
studied the Pickup and Delivery Problem (PDP) of blood products with strict time
windows. Hsu et al. (2007) modelled a food distribution planning problem with
stochastic and time-dependent travel times and time-varying temperatures. Osvald
Solving Multi-Echelon Location Allocation Problem Using Modified … 147

and Stirn (2008) not only addressed the distribution of fresh vegetables with time-
dependent travel times but also add a quality degradation-based cost function to the
objective function. Tarantilis and Kiranoudis (2001, 2002) studied the distribution
of fresh milk with a heterogeneous fixed fleet, and the distribution of fresh meat in a
multi-depot network, respectively. They considered strict time windows for delivery
of the products. Derigs et al. (2011), Mendoza et al. (2011), and Reis and Leal (2015)
proposed different shipping strategies, numbers of actors in the transportation chain,
and limitations of the transportation modes. In our study, the transportation time
is discussed as the time taken from the origin to the final destination needs to be
within a certain time. As our products need to be transported via border checkpoints,
this can delay the transportation time with a hold-up while the inter-trade document
is checked. We select borders as candidates for the shipping route based on which
borders are convenient for international trade shipping.
Transported agricultural products pass through different actors in the supply chain,
and thus it forms a transportation or service network. A service network is a structure
that brings together several entities to deliver a particular service. In our research, the
service network is formed to deliver the agricultural product from the origin to the
end-market. The network design problem has become more popular for researchers
in the last decade. The service network design (SND) problem has been modelled
and presented by Pedersen et al. (2009) and Andersen et al. (2009a, b, 2011). Moccia
et al. (2011) presented a rail and road transportation system with both consolidated
and dedicated services. Later, Thiongane et al. (2015) and Li et al. (2016) presented
a capacitated multi-commodity with heterogeneous assets and non-bifurcated hop-
constrained multi-commodity capacitated fixed-charge network design.
In this study, the network design for the transportation of the agricultural products
from the origin to the end-market involves passing through the borders of several
countries. The borders of different countries may have different regulations, which
can give the transportation a higher or lower cost and lead to a shorter or longer trans-
portation time, which can affect decisions on the transportation mode and methods.
We will present the mathematical model for the proposed problem and solve it heuris-
tically to obtain a promising solution that provides the highest profit throughout the
supply chain while keeping a low computational time.
The method that we will use to solve the proposed problem is a differential evolu-
tion algorithm. A differential evolution algorithm is a metaheuristics method that
iteratively improves the solution quality by changing the neighborhood structure
using the differential of two or more sets of numbers. This was firstly proposed by
Storn and Price (1997) and has been used to successfully solve many different types
of problems. The problems it has solved include the unmanned aerial vehicle multi-
tasking problem (Su and Wang 2021), reservoir production optimization problem
(Jianwei et al. 2021); parameter calibration problem (Luo et al. 2021; Srichok et al.
2020, 2021), scheduling problem (Pitakaso and Sethanan 2019), vehicle routing
problem (Dechampai et al. 2015), and employee transportation problem (Pitakaso
et al. 2019).
In this research, we modify the differential evolution algorithm so that DE has
better search behavior. We add one step into the original DE called the vector hybrid
148 K. Pranet et al.

reincarnation process (HRP). This step is added before the recombination process.
The defeated vector from the selection process will be randomly selected as the
trunked (TV) set of vectors. This set of vectors will be hybrid exchanged with the
best set of vectors (BV) to obtain the reincarnation vector (RV). The RV will be used
as one choice to form the trial vector.
This paper will be organized as follows. In Sect. 2, the mathematical model that
represents the network design will be presented. Sections 3 and 4 will show the
proposed methods and computational results, and the last section, Sect. 5, will present
the conclusions and outlook of this research.

2 Mathematical Model Formulation

Indices
i Wholesaler market i (i = 1, 2, 3, …, I).
j Cooperative j (j = 1, 2, 3, …, J)
k Border k (k = 1, 2, 3, …, K).
l End-market l (l = 1, 2, 3, …, L)
Parameters
I Number of wholesaler markets
J Number of agricultural cooperatives
K Number of border checkpoint
L Number of end-market s.
Dij Distance from wholesaler market i to cooperative j (km).
TS Transportation fuel cost ($/Km)
ACi Operating costs of wholesaler market i ($).
DMl The demand of end-market l (ton)
CPi Capacity of wholesaler market i (ton)
V Li Volume of central market I (ton).
EPik Distance from wholesaler market i to border k (km)
SPi Sale price of end-market l ($)
V Ckl Volume of goods delivered from border k to end-market l (ton).
NFj Product quantity of cooperative j (ton)
AMCk Operating costs of border k ($)
TT Vk Cargo quantity of border k (ton)
DCEkl Distance from border k to end-market l (Km)
V CCik Volume of goods delivered from wholesaler market i to border k
Decision variable
⎧
1 if wholesaler market i serves cooperativej
Xij
0 otherwise
Solving Multi-Echelon Location Allocation Problem Using Modified … 149
⎧
1 if wholesaler market i serves borderk
Zik
0 otherwise
⎧
1 if border k serves the end − marketl
Vkl
0 otherwise
⎧
1 if wholesaler market i is a collection point for agricultural products
Yi
0 otherwise

Objective function

{
K {
L {
K {
I {
I {
J
Max = SPl V Ckl − AMCk TT Vk − ACi V Li − Xij Dij
k=1 l=1 k=1 i=1 i=1 j=1

{
I {
K {
K {
L
+ TS.Zik EPik + TS.Vkl DCEkl (1)
i=1 k=1 k=1 l=1

Subject to

{
J
NFj Xij ≤ Yi CPi ∀i (i = 1, 2, 3, ..., I ) (2)
j=1

{
I
Xij = 1 ∀j (j = 1, 2, 3, ..., J ) (3)
i=1

{
I
V Li = Xij NFj ∀i (i = 1, 2, 3, ..., I ) (4)
i=1

V CCik ≤ Zik V Li ∀i (i = 1, 2, 3, ..., I )∀k (k = 1, 2, 3, ..., K ) (5)

Zik ≤ Yi ∀i (i = 1, 2, 3, ..., I )∀k (k = 1, 2, 3, ..., K ) (6)

{
K
V CCik = V Li ∀i (i = 1, 2, 3, ..., I ) (7)
k=1

{
I
TT Vk = V CCik ∀k (k = 1, 2, 3, ..., K) (8)
i=1

V Ckl ≤ Vkl TT Vk ∀k (k = 1, 2, 3, ..., K )∀i (i = 1, 2, 3, ..., I ) (9)

{
L
V Ckl = TT Vk ∀k (k = 1, 2, 3, ..., K ) (10)
l=1
150 K. Pranet et al.

{
K
V Ckl ≤ DMl ∀l (l = 1, 2, 3, ..., L) (11)
k=1

Equation 1 defines the objective function, attempting to maximize the total profit
of the network chain. The profit comprises the selling price of the agricultural product
at the end-market after subtracting all travel and administration costs of the actors in
the chain. Equation 2 controls the total number of products delivered from all coop-
eratives to a wholesaler market so that these do not exceed its capacity. Equation 3
shows that a single cooperative must not deliver to more than one wholesaler market.
Equations 4, 5, 6, and 7 define that the quantity of agricultural products traveling
from wholesaler i to border k must not exceed the products’ availability, while Eqs. 8,
9, and 10 control the numbers of products delivered from border k to end-market
l as these must not exceed the products that are available at the border checkpoint.
Equation 11 shows that the products delivered from all borders must not exceed the
capacity of the end-market.

3 The Proposed Method

To solve the proposed problem in this research, the differential evolution algorithm
(DE) will be used. In general, DE consists of five steps: (1) generate initial solution,
(2) perform mutation process, (3) perform mutation process, (4) perform selection
process, and (5) redo steps (2)–(4) until termination condition is met. In this research,
the modified differential evolution algorithm (MDE) will be developed to improve
the efficiency of the original DE. The hybrid reincarnation process will be added
after step (3) before the recombination process is performed. Although the MDE
proposed in this research consists of six steps, steps (1) to (3) are similar to those
of the traditional DE. Step (4) of the MDE is to perform the hybrid reincarnation
process, step (5) is to perform the recombination process, step (6) is to perform the
selection process, and step (7) is to redo steps (2)–(6) until the termination condition
is met. The modified DE can be explained stepwise as follows.

3.1 Generate the Initial Solution

Target vectors are generated in this step to obtain the vector that represents the
proposed problem. In this research, the indirect coding technique is used to represent
the problem, and the decoding process is applied to the coded vector to obtain the
solution to the proposed problem.
Solving Multi-Echelon Location Allocation Problem Using Modified … 151

Table 1 Five target vectors example

Vector Cooperative
1 2 3 4 5 6
1 642 973 610 116 498 629
2 866 158 974 152 465 121
3 157 694 816 466 626 346
4 921 569 265 959 152 520
5 127 612 780 645 552 472

3.1.1 Encode the Target Vector

Integer number coding is used to represent the problem. Table 1 represents the five
target vectors of the 1□D vector when D is the number of agricultural cooperatives.

3.1.2 Decode the Target Vector

Figure 2 displays the decoding method and presents the problem for cooperative
number 1 of vector 1. The coding number or value in position is 642. This number
can be interpreted to mean that cooperative 1 will deliver its product to wholesaler
market number 6, and then wholesaler number 6 will deliver the product to border
number 4, passing through to end-market number 2. If some actors in the chain have
more than 1 digit (e.g., there are 20 wholesaler markets), then the first two digits in
the position will represent the wholesaler number and the total digits in the position
will number four. In the same manner, the digits will run from 0 to 9 in every position
if there are less than 10 actors at the horizontal level of the actors. For example, our
system has three borders, but as can be seen in Fig. 1, it has the number in position
2 that runs from 0 to 9. This problem can be solved by dividing 10 digits into three
groups: 1–3 (border 1), 4–6 (border 2), and 7–10 (border 3). When the second digit
falls in one of these groups of numbers, it can be interpreted as corresponding to that
border.
If the problem has six cooperatives, four wholesalers, three borders, and four
end-markets, the digit presenting each actor is shown in Fig. 3.
Details of all actors are shown in Table 2.
The solution obtained from vector 1 is shown in Tables 3 and 4.

6 4 2

Label of the wholesaler Label of the border Label of the end-market

market checkpoint

Fig. 2 Example of the result represent the problem’s solution

152 K. Pranet et al.

Digit represent each wholesaler Digit represent each borders Digit represent each end-market

Digit position 1 Digit position 2 Digit position 3

0-2 wholesaler 1 0-2 border 1 0-2 end-market 1

3-4 wholesaler 2 3-6 border 2 3-4 end-market 2

5-6 wholesaler 3 7-9 border 3 5-6 end-market 3

7-9 wholesaler 4 7-9 end-market 4

Fig. 3 The interpretation of digit to problem’s solution

Table 2 Details (supply, capacity) of all actors in network chains

Cooperative Wholesaler Border End-market
No. Supply No. Capacity No. Capacity No. Capacity
(ton) (ton) (ton) (ton)
1 1200 1 4000 1 6000 1 4500
2 1000 2 3000 2 6000 2 4000
3 900 3 4000 3 5000 3 3000
4 1400 4 4500 4 4000
5 2000
6 1300

Table 3 Result of the

Wholesaler Border End-market
assignment result of the
decoding method 1 3 2 1
2 4 3 2
3 3 1 1
4 1 1 3
5 2 3 4
6 3 1 4

If the average prices per ton of the agricultural product at end-markets 1 to 4 are
$14, $12, $11, and $15, respectively, then the total revenue generated is $143,200. The
administration costs of wholesalers 1 to 4 are $2, $1, $2, and $1.50, respectively and
the administration costs at borders 1 to 1 are $2.30, $2.40, and $2.10, respectively. The
total administration cost is thus $20,960. The travel cost is calculated by multiplying
the total traveling distance by the traveling cost per kilometer, which is $0.10 per
kilometer. If this plan has a total distance of 8300 km, then it has a traveling cost of
$830. Therefore, the total profit generated from this plan is $104,410.
After the decoding method is performed, the next steps of the MDE are to perform
the mutation, recombination, hybrid reincarnation, and selection processes.
Solving Multi-Echelon Location Allocation Problem Using Modified … 153

Table 4 Result of the volume transported within the chain

Wholesaler
1 2 3 4
Cap 4000 3000 4000 5000
Total agricultural product 1400 2000 3400 1000
Border
1 2 3
Cap 6000 6000 5000
Total agricultural product 3600 1200 3000
End-market
1 2 3 4
Cap 4500 4000 3000 4000
Total agricultural product 4100 1000 1400 3300

3.2 Perform the Mutation Process

The mutation process is used to transform the target vector to the mutant vector.
The process uses the difference of two selected vectors as the indicator of the
transformation. Equation (12) represents the mutation operators.

Vi,j,G = Xr1 ,j,G + F(Xr2 ,j,G + Xr3 ,j,G ) (12)

where r1, r2, and r3 are the indices of the randomly selected vectors; F is the scaling
factor, which is set to 0.8 (Qin et al. 2009); i represents the vector number; i = 1,
2,…, NP; j is the position of a vector when j = 1, 2,…, D.

3.3 Perform the Hybrid Reincarnation Process (HR)

This process is a new process that we present in this paper. The reincarnated vector
(H i,j,G ) is formed in this process. To form the reincarnated vector, we use two sets
of vectors. The first set of vectors is known as trunked vectors (TV: M i,j,G ); this
comprises NP vectors that are randomly selected from the vectors that are trunked
in selection process of the last iterations. These vectors will be updated iteratively;
the system that is used to update the trunked vectors is one in which the vector that
stays on the list the longest is replaced first by a newly selected TV. The second set of
vectors used in this process is that of the best vectors (BV). NP best vectors obtained
for entire iterations will be collected in the best vector (BV: Bi,j,G ). The procedure
to generate the Hi,j,G has three steps: (1) randomly select a vector from the trunked
vectors (Hi,j,G ), (2) randomly select a vector from the best vectors.
154 K. Pranet et al.

(Bi,j,G ), and (3) use Eq. (12) to construct a reincarnated vector (RV : H i,j,G ). An
example of a reincarnated vector generator is shown in Fig. 4. CR is a predefined
parameter that is set to 0.65 in this research.
⎧
Mi,j,G when randi,j,G ≤ CR
Hi,j,G = (13)
Bi,j,G otherwise

3.4 Perform the Recombination Process

Obtaining the trial vector is the aim of performing the recombination process.
Equation (14) is used to transform the mutant vector into the trial vector.
⎧
⎨ Hi,j,G if randi,j,G < CR1
Ui,j,G = Vi,j,G if CR1 ≤ randi,j,G < CR2 (14)
⎩
Xi,j,G otherwise

where Ui,j,G is the trial vector, Vi,j,G is the mutant vector, rand i,j,G is a random
number [0,1], Xi,j,G is the target vector of vector i position j iterations G, and CR1
and CR2 are predefined numbers set to 0.35 and 0.65, respectively.

3.5 Perform the Selection Process

The selection process is executed to generate a new target vector. Equation (15) is
used to select the new target vector to use in the next iteration. When Xi,j,G+1 is
the target vector for the next iteration, f (Ui,j,G ) is the objective function of the trial
vector and f (Xi,j,G ) is the objective function of the current target vector.
⎧ ( ) ( )
Ui,j,G when if f Ui,j,G ≥ f Xi,j,G
Xi,j,G+1 = (15)
Xi,j,G otherwise

3.6 Redo Step (3.2–3.4) Until Termination Condition is Met

The stoppage criterion for this research is the computational time of the proposed
method in order for a fair comparison with the other compared heuristics. The pseudo
code of the proposed methods is shown in Fig. 5.
Solving Multi-Echelon Location Allocation Problem Using Modified … 155

Trunked Vector (Mi,j,G)

cooperative 1 2 3 4 5 6

Vector

1 904 394 995 852 280 649

2 501 193 956 447 806 366

3 898 342 814 878 968 261

4 156 389 189 611 690 303

5 313 221 799 882 480 378

Best Vector (Bi,j,G)

cooperative 1 2 3 4 5 6

Vector

1 138 894 765 388 643 450

2 850 541 313 485 225 378

3 767 716 775 128 430 597

4 681 576 436 645 560 131

5 104 940 792 607 270 951

cooperative 1 2 3 4 5 6

Vector

Mi,j,G (2) 501 193 956 447 806 366

Bi,j,G (4) 681 576 436 645 560 131

, , 0.41 0.76 0.18 0.41 0.89 0.34

CR 0.65 0.65 0.65 0.65 0.65 0.65

Hi,j,G 501 576 956 447 560 366

Fig. 4 Example of reincarnated vector generator

156 K. Pranet et al.

Fig. 5 Pseudo code of the proposed method

3.7 Compared Heuristics

In the experiment, we compare two other methods with the proposed method: (1) the
original differential evolution algorithm and (2) the original genetic algorithm. The
original DE is similar to the proposed method, but the original does not have the HS
process, while the GA is a method taken from Mitchell (1996) to use in our problem.
The pseudo code of GA used in our research is shown in Fig. 6.

4 Computational Results and Framework

We code the mathematical model in Lingo V.11 and code the modified DE, DE, and
GA with C++, executing them using a PC Intel® Core ™ i5-2467 M CPU 1.6 GHz.
Fifteen randomly generated datasets are tested to compare the performance of the
proposed methods. Each has a different number of cooperations, central markets,
border checkpoints, and end-markets. Table 5 details all test problems included in
the case study.
The stoppage criterion of Lingo is the computational time when it can find the
optimal solution within 24 h. If it cannot find the optimal solution within 48 h, the best
Solving Multi-Echelon Location Allocation Problem Using Modified … 157

Fig. 6 Pseudo code of GA used in this research

solution is instead recorded. The stoppage criterion of the heuristics is the computa-
tional time, which is set to five minutes for test instances 1–5, 10 min for instances
6¬–10, and 30 min for instances 11–15 (include case study). The computational
results are shown in Table 6 and the statistical results of the Wilcoxon signed-rank
test are shown in Table 7.
Tables 6 and 7 show that the MDE significantly improves the solution quality
versus other methods, including the best objective obtained from Lingo v.11, which
158 K. Pranet et al.

Table 5 Detail of the test instances

Test instance no Number of Number of Number of border Number of
cooperative center-market checkpoint end-market
1 15 6 4 11
2 20 6 4 11
3 25 6 4 11
4 30 6 4 11
5 35 6 4 11
6 40 6 4 11
7 45 6 4 11
8 50 6 4 11
9 55 6 4 11
10 60 6 4 11
11 65 6 4 11
12 70 6 4 11
13 75 6 4 11
14 80 6 4 11
Case study 86 6 4 11

Table 6 Computational result of all 14 test instances

Test Lingo v.11 GA DE MDE
instance profit ($) ComT Profit ($) ComT Profit ($) ComT Profit ($) ComT
no (min) (min) (min) (min)
1 199202.6 53.49 193052.0 5 199202.6 5 199202.6 5
2 251162.6 94.89 241034.7 5 251162.6 5 251162.6 5
3 318072.4 120.82 318072.4 5 314500.6 5 318072.4 5
4 356517.6 1440 356517.6 5 356517.6 5 356517.6 5
5 414876.4 1440 423794.8 5 430783.2 5 443033.8 5
6 508729.8 1440 514712.5 10 523468.0 10 533945.9 10
7 598124.5 1440 605844.8 10 614590.1 10 636206.7 10
8 610287.7 1440 640889.0 10 649937.2 10 661569.9 10
9 639812.4 1440 691052.0 10 705855.1 10 717526.8 10
10 710183.2 1440 733103.0 10 733725.4 10 756481.2 10
11 832718.3 1440 889632.4 30 899002.1 30 920912.0 30
12 901273.8 1440 941495.5 30 946175.4 30 971764.7 30
13 922371.7 1440 966128.3 30 969167.9 30 997934.3 30
14 987273.2 1440 1056063.3 30 1065787.4 30 1088201.6 30
Case 992741.7 2880 1145842.2 30 1155271.6 30 1167625.7 30
study
Solving Multi-Echelon Location Allocation Problem Using Modified … 159

Table 7 Statistical test of the

GA DE MDE
computational obtained in
Table 13 (p-value) Lingo v.11 0.001 0,001 0.001
GA 0.000 0.000
DE 0.000

9.00
8.00
Percent different from MDE

7.75
7.00
6.00
5.00
4.00
3.02
3.00
2.00 2.05

1.00
0.00 0.00
Profit 616223.19 647815.63 654343.12 668010.52
Meth-
Lingo
d v.11 GA DE MDE

Fig. 7 Percent different of the solution quality of various method compared with MDE

uses a computational time of less than 2880 h. Figure 7 shows that the MDE can
improve the solution of the original DE by 2.02%, which means that using the hybrid
reincarnation process is useful to improve the solution quality versus the original DE.
Moreover, the MDE produces 7.75% and 3.02% better solutions than those of the
best objective (obtained from Lingo v.11 using 2880 h computational time) and the
GA, respectively.
Figure 8 shows the numbers and percentages of methods that can find the
maximum profit of the proposed problem. We can see than the MDE can find 100%
of the maximum profit Figs. (15 cases), while Lingo V.11, GA, and the original DE
can find 26.67% (4 cases), 13.33% (2 cases), and 20% (3 cases), respectively.
Figure 9 shows that the MDE outperforms the original DE by finding a 100%
better solution than that of the DE, which mean that using the hybrid reincarnation
process contributes to improving the solution quality of the DE.

5 Conclusion and Future Outlook

This research aimed to design a transportation model for agricultural products’ inter-
national trade among Thailand, Laos, and Cambodia. The agricultural products are
transported from a farmers’ cooperative to a foreign end-market. During transporta-
tion, they pass through wholesaler markets and border checkpoints. The mathematical
160 K. Pranet et al.

100.00

100.00

Percent different from MDE

90.00
80.00
70.00
60.00
50.00
40.00 26.67
13.33 20.00
30.00
20.00
10.00
0.00 # of max. Profit
4 2 3 15
Methods
Lingo v.11 GA DE MDE

Fig. 8 Number of Maximum Profit Generated from Various Methods

2500000
Percent different from MDE

2000000

1500000

1000000

500000

Problem instances #
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

DE MDE

Fig. 9 Profit generated from DE and MDE

models GA, DE, and MDE have been developed to determine the optimal mode of
transportation. The hybrid reincarnation process has been added to the original DE
to improve its search capability.
The computational results show that the MDE gives better solutions than Lingo
V.11, GA, and the DE by 7.75%, 3.02%, and 2.05%, respectively. The MDE, which
is the DE with an added hybrid reincarnation process, can find 100% of the maximum
profit, while the DE can find 20% of the maximum profit, which means the MDE
can improve the solution quality of the original DE by 80%.
In future research, multiple fleets should be added into the model so the system
can select the right fleet type for the right amount of agricultural product to be
shipped. Moreover, planning over multiple periods and for multiple products could
Solving Multi-Echelon Location Allocation Problem Using Modified … 161

increase the performance of the model. These suggestions, when investigated in

future research, should increase the performance of the MDE model.

Acknowledgements Authors would like to gratitude the supports from (1) Research Unit on
System Modelling for Industry (Grant No. SMI.KKU 64/06), Faculty of Engineering, Khon Kaen
University, Thailand, (2) Meta-heuristics for Logistics Optimization Laboratory, Faculty of Engi-
neering, Ubon Ratchathani University, and (3) Department of Logistics Management, Faculty of
Industrial Technology, Ubon Ratchathani Rajabhat University.

References

Andersen J, Crainic T, Christiansen M (2009a) Service network design with asset management:
formulations and comparative analyses. Transp Res Part C 17(2):197–207
Andersen J, Crainic T, Christiansen M (2009b) Service network design with management and
coordination of multiple fleets. Eur J Oper Res 193(2):377–389
Andersen J, Christiansen M, Crainic T, Grønhaug R (2011) Branch and price for service network
design with asset management constraints. Transp Sci 45(1):33–49
Dechampai D, Tanwanichkul L, Sethanan K, Pitakaso P (2015) A differential evolution algorithm
for the capacitated VRP with flexibility of mixing pickup and delivery services and the maximum
duration of a route in poultry industry. J Intell Manuf 28(6). https://doi.org/10.1007/s10845-015-
1055-3
Derigs U, Gottlieb J, Kalkoff J, Piesche M, Rothlauf F, Vogel U (2011) Vehicle routing with
compartments: applications, modelling and heuristics. Or Spectr 33(4):885–914
Gu J, Liu W, Zhang K, Zhai L, Zhang Y, Chen F (2021) Reservoir production optimization based
on surrograte model and differential evolution algorithm. J Pet Sci Eng 205:108879. ISSN 0920-
4105.https://doi.org/10.1016/j.petrol.2021.108879
Hsu C, Hung S, Li H (2007) Vehicle routing problem with time-windows for perishable food
delivery. J Food Eng 80(2):465–475
Li X, Wei K, Aneja Y, Tian P (2016) Design-balanced capacitated multicommodity network design
with heterogeneous assets. Omega. Article in press
Luo G, Zou L, Wang Z, Lv C, Ou J, Huang Y (2021) A novel kinematic parameters calibration
method for industrial robot based on Levenberg-Marquardt and Differential Evolution hybrid
algorithm. Robot Comput Integr Manuf 71:102165. ISSN 0736-5845.https://doi.org/10.1016/j.
rcim.2021.102165
Mendoza J, Castanier B, Guéret C, Madaglia A, Velasco N (2011) Constructive heuristics for the
multicompartment vehicle routing problem with stochas- tic demands. Transp Sci 45(3):346–363
Mitchell M (1996) An introduction to genetic algorithms. Cambridge
Moccia L, Cordeau J, Laporte G, Ropke S, Valentini M (2011) Modeling and solving a multimodal
transportation problem with flexible time and scheduled services. Networks 57(1):53–68
Pedersen M, Crainic T, Madsen O (2009) Models and tabu search metaheuristics for service network
design with asset-balance requirements. Transp Sci 43(2):158–177
Pitakaso R, Sethanan K, Srijaroon N (2019) Modified differential evolution algorithms for multi-
vehicle allocation and route optimization for employee transportation. Eng Optim 1–19. https://
doi.org/10.1080/0305215x.2019.1640691
Pitakaso R, Sethanan S (2019) Adaptive large neighborhood search for scheduling sugarcane
inbound logistics equipment and machinery under a sharing infield resource system. Comput
Electron Agric 158:313–325. https://doi.org/10.1016/j.compag.2019.02.001
162 K. Pranet et al.

Qin K, Huang V, Suganthan P (2009) Differential evolution algorithm with strategy adaptation for
global numerical optimization. Evolutionary computation. IEEE Trans 13:398–417. https://doi.
org/10.1109/TEVC.2008.927706
Reis S, Leal J (2015) A deterministic mathematical model to support temporal and spatial decisions
of the soybean supply chain. J Transp Geogr 43:48–58
Sethanan K, Pitakaso R (2016) Differential evolution algorithms for scheduling raw milk
transportation. Comput Electron Agric 121:245–259. https://doi.org/10.1016/j.compag.2015.
12.021
Srichok T, Pitakaso R, Sethanan K, Sirirak W, Kwangmuang P (2020) Combined response surface
method and modified differential evolution for parameter optimization of friction stir welding.
Processes 8:1080
Srichok T, Pitakaso R, Sethanan K, Kumphon O, Pattanapairoj S, Worasan K (2021) Novel approach
to find the optimal parameters of a tractor disc plough used for land preparation. Eng Optim.
https://doi.org/10.1080/0305215X.2021.1901284
Storn R, Price K (1997) Differential evolution—a simple and efficient heuristic for global optimiza-
tion over continuous spaces. J Global Optim 11(4):341–359. https://doi.org/10.1023/A:100820
2821328.S2CID5297867
Su J, Wang H (2021) An improved adaptive differential evolution algorithm for single unmanned
aerial vehicle multitasking. Defence Technol. ISSN 2214-9147. https://doi.org/10.1016/j.dt.2021.
07.008
Tarantilis C, Kiranoudis C (2001) A meta-heuristic algorithm for the efficient distribution of
perishable food. J Food Eng 50(1):1–9
Tarantilis C, Kiranoudis C (2002) Distribution of fresh meat. J Food Eng 51(1):85–91
Thiongane B, Cordeau J, Gendron B (2015) Formulations for the nonbifurcated hop-constrained
multicommodity capacitated fixed-charge network design problem. Comput Oper Res 53:1–8
Mathematical Modelling for Optimizing
Tourist Trip Design with Considering
Scoring on Arc Visits

Ryo Geoffrey Widjaja, A. A. N. Pewira Redi, Parida Jewpanya,

Muhammad Asrol, and Nur Layli Rachmawati

1 Introduction

A number of application problems of logistics, tourism, and related fields have arisen
in the last decade (Vansteenwegen et al. 2011). These problems are modelled as “Ori-
enteering Problem”, which was first introduced by Golden et al. (Golden et al. 1987).
The generalized orienteering problem (GOP) is one extension of the orienteering
problem (Urrutia-Zambrana et al. 2021). The objective of the GOP is to find the
tour that maximizes the total score collected by visiting vertices without exceeding
a distance or time constraint. A set of weights of each vertex associated with the
vertices scores to calculate the objective functions.
The previous research has widely solved the GOP using a different algorithm for
many applications, especially for tourist trip design. (Geem et al. 2005) introduced
Harmony Search (HS), focusing on three significant behaviours (memory consider-
ation, pitch arrangement, and irregular choice) of music players that translated into
the context of the GOP. In the Harmony Matrix, the selection of any node invokes
memory consideration. Pitch arrangement is related to selecting the closest node
from the next node and irregular choice to choose one node from entire feasible
nodes. HS applied to search for the best tour and compared it with ANN from Wang

R. G. Widjaja · A. A. N. Pewira Redi (B) · M. Asrol

Industrial Engineering Department, BINUS Graduate Program – Master of Industrial
Engineering, Bina Nusantara University, Jakarta 11480, Indonesia
e-mail: [email protected]
P. Jewpanya
Department of Industrial Engineering, Rajamangala University of Technology Lanna, Tak 50300,
Thailand
N. L. Rachmawati
Department of Logistics Engineering, Universitas Pertamina, Jakarta 12220, Indonesia

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 163
P. Golinska-Dawson et al. (eds.), Smart and Sustainable Supply Chain
and Logistics—Challenges, Methods and Best Practices, EcoProduction,
https://doi.org/10.1007/978-3-031-15412-6_10
164 R. G. Widjaja et al.

et al. (Wang et al. 1995), which concluded that HS had performed better than Artifi-
cial Neural Network (ANN). Another further research from Wang et al. (Wang et al.
2008) proposes a Genetic Algorithm (GA) to solve a similar problem and outperform
the HS.
The GOP is not restricted only to the previous applications. The significant uses
of GOP are modelling the tour for tourist to plan their trip to visit different attrac-
tions. The tourist has limited time and requires having an effective trip during that
time. It means that the tourist does not necessarily visit all the attractions (Urrutia-
Zambrana et al. 2021). This would prepare the most pleasant trip according to the
tourist preferences.
One way to accommodate the tourist to their destination is using electric vehicles.
Before the COVID-19 occurred, the appearance of shared mobility options such as
bikes, electric scooters, or cars was increased (Mouratidis et al. 2021). This new
market trend then emerges in Indonesia, showing the high usage of electric vehi-
cles such as electric bikes or electric scooters for tourist accommodations. For that
purpose, the critical consideration is the score related to the attraction and the score
from the path they took since electric vehicles are necessary to have a compatible
road. Therefore, the generalized orienteering problem will be combined with the arc
orienteering problem.
Arc orienteering problem is also another extension orienteering problem. Only
a few research studies consider this new variant OP where the price is associated
with each arc instead of vertex (Aráoz et al. 2009a, b). The approach to consolidate
arc score as the routing decision is proposed by Gavalas et al. (2015). The problem
objective of AOP is to maximize the difference between profit and the total travel
distance. Several AOP applications are collecting recycling bins where every arc in
the area must be passed through, but only a few selected to have the highest profits
by Aráoz et al. (2009a, b), which also introduced the team arc orienteering problem.
Archetti et al. (2014) proposed the same problem using a different application where a
limited fleet of vehicles needs to serve potential customers. Meanwhile, the extension
orienteering problem is associated with node and arc score as the route decision has
not been investigated before.
The combination of GOP and AOP is later denoted as generalized arc orienteering
problem (GAOP) where the objective functions are calculated with the sum of vertices
and arc scores. The novelty of the research lies in the introduction of the GAOP that
indicates many potential applications in tourist trip design planning.
The major contributions of this paper are described as follows. (1) introduce the
mathematical model for the combination of GAOP. (2) a metaheuristic algorithm
will be used to solve the proposed problem categorized as an NP-hard problem. (3)
the algorithm will be implemented using several attractions in Kota Tua, Jakarta,
with the score attraction and arc based on user ratings and roads compatibility. Kota
Tua was chosen due to one of the heritage attractions and its renowned name having
a high historical value in Jakarta. Kota Tua is located in Indonesia’s capital city,
which gives an advantage in the strategic location compared to the other attraction
(Damanik et al. 2019).
Mathematical Modelling for Optimizing Tourist Trip Design … 165

The rest of this paper is organized as follows: the problem and mathematical model
are described in Sect. 2. Metaheuristic algorithms to solve the proposed problem and
the result are presented in Sect. 3. The discussion towards the result is shown in
Sect. 4. Lastly, the conclusion remarks and future research recommendations are
drawn in Sect. 5.

2 Problem Description and Mathematical Model

OP classified as NP-hard by Golden et al. (1987), means that no polynomial-time

algorithm is expected to get optimal solution by solving this problem. The exact solu-
tion method is very time consuming, due to the increase of the problem dimensions
associated with computation complexity. Therefore, metaheuristics application will
be necessary to discover a close optimal solution in a restrained schedule.
As explained before from Sect. 1, the main idea of the GAOP came up from
the tourist trip design and the emerging new market trend of electric bikes/scooters
in Indonesia. The GAOP consist of each vertex with a score vector and each arc
with a score arc representing different attributes, and the overall objective function is
nonlinear. The objective is to find the maximum tourist experience without exceeding
the distance/time constraint. The score vector is measured by the user ratings of
unique customer preferences from the google applications. The arc score is calculated
by the distance and road compatibility for the electric bikes/scooters.
The mathematical model resulted from the assessment of previous GOP (Urrutia-
Zambrana et al. 2021) and AOP (Yu et al. 2015) research with some combination
or additional constraints and variables. A Mathematical Programming Language
(AMPL) software was used to verify the combined mathematical model with Gurobi
9.0 solver. The problem illustration can be observed in Fig. 1. The proposed
mathematical optimization model is designed as follows.
Let G = (V, A), where V = {1, . . . , N } is the set of attractions, A =
{(i, j )/i, j  V } is a set of arcs between vertices. Vertex number 1 is the initial
vertex (depot), N is the path’s ending vertex. ti j is the distance associated with arc
(i, j)  A. Tmax is the total maximum allowed defined
 time constraint,
 S f (i ) is the
vertex score associated with each vertex S(i ) = Si (i ), ..., S f (i ) V , and Ai j is the
arc score associated with each arc. W = wi , ..., w f , w f is the weight based on the
traveller according to the f . The non-negative exponent for calculating the module
of the vector is k.
Sets:
V Set of vertices
A Set of arcs between vertices
N Set of all nodes
P Set of all path
F Set of traveler
166 R. G. Widjaja et al.

Fig. 1 Illustration of GAOP

depicted in Kota Tua

Parameters:
S f (i ) Score of vertex i
Ai j Arc score of vertex i and j
Oi Opening time window of vertex i
Ci Closing time window of vertex i
ti j Traveling time from vertex i to j
vi Visiting time at vertex i
Tmax Given time horizon
Ct i j Traveling cost from vertex i to j
M A large constant
Cvi Visiting price at vertex i
B Given budget for each path
wf Weight traveler
k Non negative vector
Decision variables:
xi j p If vertex i is followed by a visit to vertex j by path p, this binary variable equals
one; Otherwise, it equals zero
Mathematical Modelling for Optimizing Tourist Trip Design … 167

yi p If vertex i is visited by path p, this binary variable equals one; Otherwise, it

equals zero
si p A variable of starting service time at vertex i in path p
Objective
⎛ ⎞1/k

F 
P 
N
 k 
P 
N 
N
maxZ = wf⎝ S f (i) yi p ⎠ + Ai j ∗ x i j p (1)
f =1 p=1 i=2 p=1 i=1 j=1

Subject to


P 
N
x1 j p ≤ 1 (2)
p=1 j=2

 N −1
P 
xi N p ≤ 1 (3)
p=1 i=1


P 
N  N −1
P 
xN jp = xi1 p = 0 (4)
p=1 j=2 p=1 i=1


P 
N
∀ p = 1, . . . , P
xikp = xk j p = ykp (5)
∀ k = 2, . . . , N − 1
p=1 j=2

  ∀ i, j = 1, . . . , N
si p + ti j + vi − s j p ≤ M 1 − xi j p (6)
∀ p = 1, . . . , P


P
ykp ≤ 1 ∀k = 2, . . . , N − 1 (7)
p−1
⎛ ⎞
N −1
 
N
⎝vi ∗ yi p + ti j ∗ xi j p ⎠ ≤ Tmax ∀ p = 1, . . . , P (8)
i=1 j=2

(N
 −1) 
N 
N
Cti j ∗ xi j p + yi p ∗ Cvi ≤ B ∀ p = 1, . . . , P (9)
(i=1) ( j=2) (i=2)

∀i = 1, . . . , N
Oi ≤ si p (10)
∀ p = 1, . . . , P

∀i = 1, . . . , N
si p ≤ Ci (11)
∀ p = 1, . . . , P
168 R. G. Widjaja et al.

∀i, j = 1, . . . , N
xi j p , yi p ∈ {0, 1} (12)
∀ p = 1, . . . , P

The objective function (1) is to maximize the total user experience from the sum
of each traveler’s weight with the sum of each vertex score and arc score. Constraints
(2), (3), and (4) guarantee that each path start at vertex 1 (depot) and finish at vertex N
(depot). Constraints (5) ensure the connectivity of each path. Constraints (6) ensure
the feasibility of the schedule. Constraints (7) ensure that all vertex is visited at
most once. Constraints (8) and (9) ensure limited time horizon and travelling budget
in each path. Constraints (10) and (11) restrict the start of the service to the time
window.

3 Computational Study

The AMPL was used to carry out the proposed algorithm. The program was executed
on a laptop with specification: Intel® Core Processor™ i7-7700HQ 1.8 GHz, 64-bit
operating system with 8 Gb of RAM. The appropriate vertex score determination
could adopt from the tourist attraction evaluation proposed by Wang et al. (2020).
Unfortunately, the scope of this study only covers attraction determination by user
review based on google. The visiting cost is found from the entry price or the price
range of Kota Tua attractions/restaurants. The vertex score and visiting cost are
described in Table 1.
Distances covered by electric vehicles, especially e-scooters, are between 20 to
25 km, depending on battery capacity, driver weight, riding style, and the road’s
steepness (Martínez-Navarro et al. 2020). Although the power advantages, the appli-
cations highly correlate with user satisfaction. Therefore, the arc score was defined
through the range of shorter distances. The arc score is described in Table 2.
The limitation on budget and time may vary depending on the user needs. The
application budget and time limit have been defined as eighty thousand rupiah and
two hours. The route result is shown in Fig. 2.
The result: initial place from the depot was visited through nodes (2) (Museum
Mandiri), nodes (3) (Museum Bank Indonesia), nodes (4) (Djakarta Kedai Seni) and
nodes (10) (Museum Fatahillah) at last before going back to the depot. The objective
function reached 68.9 by visiting four attractions within two hours.

4 Discussion

AMPL application was used to obtain the optimal result by the exact method to
validate the mathematical model of GAOP. The parameter input for the model has
been given by observing the Kota Tua user review of the amount of attraction to
Mathematical Modelling for Optimizing Tourist Trip Design … 169

Table 1 Kota Tua attraction

Nodes Attractions Values
parameter
Vertex score Visiting cost (in
thou. Rupiah)
2 Museum 4.6 5
Mandiri
3 Museum Bank 4.7 5
Indonesia
4 Djakarta Kedai 4.5 50
Seni
5 Toko Merah 4.5 10
6 Kantin Mega 4.3 75
Rasa
7 Museum 4.5 5
Wayang
8 Kafe Batavia 4.4 100
9 Museum 4.4 5
Keramik
10 Museum 4.5 5
Fatahillah

Table 2 Arc score

Distance Arc score
interpretation
1–10 m 5
10–50 m 4
50–100 m 3
100–200 m 2
>200 m 1

generate the score of each node. The distance of the road determines the score of
the arc, while the electric bike/scooter is based on user satisfaction. The number of
attractions and the distance between attractions was also achieved from google maps
for Kota Tua in Jakarta. The data was executed, and the result showed that the optimal
solution within the constraint/limitation was only to visit four from ten attractions.

5 Conclusion

This research proposes a combination of generalized and arc orienteering prob-

lems. The problem is modelled corresponding as GOP with additional arc score and
constraints. The AMPL software performed the proposed model to get the optimal
solution by using the exact method. The result has shown a higher user experience
170 R. G. Widjaja et al.

Fig. 2 Tour result

represented as Z that is not necessary to visit all nodes. Meanwhile, the lower user
experience indicates having visited attractions with lower vertex and arc scores. Thus,
it is necessary to visit the best attractions based on the limited time and budget.
For further research, the arc score can be defined through the smoothness of the
road or conducting a survey to use AHP and illustrate more into the actual conditions.
The vertex score determination can also be specified using the proposed literature
from Section III and associated with AHP or survey to compose a solid argument. The
experiment can also be carried out by other heuristics or meta-heuristics approaches
that enable obtaining closer to optimal results.

Acknowledgements This work is supported by Research and Technology Transfer Office, Bina
Nusantara University, as a part of Bina Nusantara University’s International Research Grant,
entitled: An Intelligent Decision Support System Design for Improving Industry’s Sustainability
Performance, contract number: No.017/VR.RTT/Ill/2021 and contract date: 22 March 2021.
Mathematical Modelling for Optimizing Tourist Trip Design … 171

References

Aráoz J, Fernández E, Franquesa C (2009a) The clustered prize-collecting arc routing problem.
Transp Sci 43(3):287–300. https://doi.org/10.1287/trsc.1090.0270
Aráoz J, Fernández E, Meza O (2009b) Solving the prize-collecting rural postman problem. Eur J
Oper Res 196(3):886–896. https://doi.org/10.1016/j.ejor.2008.04.037
Archetti C, Speranza MG, Corberán Á, Sanchis JM, Plana I (2014) The team orienteering arc routing
problem. Transp Sci 48(3):442–457. https://doi.org/10.1287/trsc.2013.0484
Damanik D, Wachyuni SS, Wiweka K, Setiawan A (2019) The influence of social media on the
domestic tourist’s travel motivation case study: Kota Tua Jakarta, Indonesia. Curr J Appl Sci
Technol 1–14. https://doi.org/10.9734/cjast/2019/v36i630263
Gavalas D, Konstantopoulos C, Mastakas K, Pantziou G, Vathis N (2015) Approximation algorithms
for the arc orienteering problem. Inf Process Lett 115(2):313–315. https://doi.org/10.1016/j.ipl.
2014.10.003
Geem ZW, Tseng CL, Park Y (2005) Harmony search for generalized orienteering problem: best
touring in China. Lect Notes Comput Sci 3612(PART III):741–750. https://doi.org/10.1007/115
39902_91
Golden BL, Levy L, Vohra R (1987) The orienteering problem. Naval Res Logistics (NRL)
34(3):307–318. https://doi.org/10.1002/1520-6750(198706)34:3<307::AID-NAV3220340302>
3.0.CO;2-D
Martínez-Navarro A, Cloquell-Ballester VA, Segui-Chilet S (2020) Photovoltaic electric scooter
charger dock for the development of sustainable mobility in urban environments. IEEE Access
8:169486–169495. https://doi.org/10.1109/ACCESS.2020.3023881
Mouratidis K, Peters S, van Wee B (2021) Transportation technologies, sharing economy, and
teleactivities: implications for built environment and travel. Transp Res Part D Transp Environ
92(February):102716. https://doi.org/10.1016/j.trd.2021.102716
Urrutia-Zambrana A, Tirado G, Mateos A (2021) Variable neighborhood search to solve the gener-
alized orienteering problem. Int Trans Oper Res 28(1):142–167. https://doi.org/10.1111/itor.
12800
Vansteenwegen P, Souffriau W, Oudheusden DV (2011) The orienteering problem: a survey. Eur J
Oper Res 209(1):1–10. https://doi.org/10.1016/j.ejor.2010.03.045
Wang Q, Sun X, Golden BL, Jia J (1995) Using artificial neural networks to solve the orienteering
problem. Ann Oper Res 61(1):111–120. https://doi.org/10.1007/BF02098284
Wang X, Golden B, Wasil E (2008) Using a genetic algorithm to solve the generalized orienteering
problem. Oper Res/comput Sci Interfaces Ser 43. https://doi.org/10.1007/978-0-387-77778-8
Wang Y, Shen S, Sotiriadis M, Zhang L (2020) Suggesting a framework for performance evaluation
of tourist attractions: a balance score approach. Sustainability (switzerland) 12(15):1–22. https://
doi.org/10.3390/su12156220
Yu VF, Perwira Redi AAN, Jewpanya P (2015) The orienteering problem with time windows and
arc scores, pp 1–6
Two-Echelon Vehicle Routing Problem
for Agricultural Area Mapping
Operation with Drone

A. A. N. Pewira Redi, Muhammad Reza Chandra Kusuma,

Bertha Maya Sopha, Anna Maria Sri Asih, and Rahmad Inca Liperda

1 Introduction

A study conducted by the Indonesian National Land Agency also known as Badan
Pertanahan Nasional mention that the conversion rate of agricultural land to non-
agricultural lands in Indonesia has reached around 50,000 hectares or 500,000,000
m2 in 2011, which is categorized as high (Darmanto 2017). Given that it is now 2021,
the development of the rate of land-use change is substantial and very fast. Indonesia
as an agrarian country needs to ensure the provision of agricultural land in a sustain-
able manner by prioritizing the principles of efficiency, sustainability, independence
and environmental insight. On the other hand, the increasing population growth with
a fairly high growth rate of around 1.4–1.5% per year as well as economic and
industrial developments have resulted in degradation and conversion of agricultural
land so that it affects the national carrying capacity in maintaining self-reliance and
food security (Gunawan 2013). Based on this, the government formed a program that
specifically works to protect sustainability of agricultural land, such as by conducting
act of identifying and store information on sustainable agricultural land. The infor-
mation regarding agricultural land can be carried out by using current technology
such as remote sensing, photogrammetry, and GIS. Among many techniques for
measuring and mapping, photogrammetric mapping techniques with drones are one

A. A. N. Pewira Redi (B) · M. R. C. Kusuma

Department of Industrial Engineering Department, BINUS Graduate Program – Master of
Industrial Engineering, Jakarta, Indonesia
e-mail: [email protected]
B. M. Sopha · A. M. S. Asih
Department of Mechanical and Industrial Engineering, Universitas Gadjah Mada (UGM),
Yogyakarta, Indonesia
R. I. Liperda
Department of Logistics Engineering, Universitas Pertamina, Jakarta, Indonesia

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 173
P. Golinska-Dawson et al. (eds.), Smart and Sustainable Supply Chain
and Logistics—Challenges, Methods and Best Practices, EcoProduction,
https://doi.org/10.1007/978-3-031-15412-6_11
174 A. A. N. Pewira Redi et al.

of the favorable and cheapest techniques to carry out to map an agricultural land or
area. It is because drones are normally equipped with an autopilot system, equipped
with precision GPS, have high camera resolution, produce a high-scale map detail,
are not constrained by clouds, and the data retrieval can be done at any time (Utomo
2018).
Mapping activities with drones can have several problems. One in particular
including mapping operation that carried out in a wide area while the coverage
of a drone to operate is limited to its battery. This can be seen and led to a routing
problems where the access point from takeoff location or mapping area needs to
be planned accurately, taking into account the cost or capacity of land vehicles and
the type of drone used. This situation can be optimize by join operation between
drones combined with ground vehicle. This can be seen as an optimization problem
more specifically vehicle to solve routing problems (VRPD) which was developed by
combining ground vehicles that can carry drones to extend their reach to customers
with these drone vehicles (Wang and Sheu 2019). The application of cooperation
between land vehicles and drones is often found in delivery services. The mecha-
nism is modeled as a two-echelon vehicle routing problem with drones (2EVRPD)
which is applied for last-mile delivery of packages or parcels.
The use of ground vehicles and drones can increase the efficiency in mapping
operations. Moreover, it can increase the coverage area being mapped. Thus, this
study take the consideration on combination of ground vehicles and drones to reach
the entire mapping area of agricultural land. Despite that, the study on applying
collaboration of ground vehicles combined with drones for agricultural land mapping
operations is rare. One particular research use a combination of ground vehicles
and drones with the Single Depot Vehicle Routing Problem (SDVRP) model for
mapping forest area activities with land vehicles as filling drones (Maini and Sujit
2015). The closest research to this study perhaps is the application of 2EVRPD on the
disaster mapping operation which is later denoted as the two echelon vehicle routing
problem for mapping operation with drone (2EVRP-MOD) (Liperda et al. 2020).
The 2EVRP-MOD model is categorized as a complex problem (NP-Hard), so it can
be solved using a metaheuristic algorithm. One particular example of a metaheuristic
algorithm that perform good to solve VRP problems and its variant is the simulated
annealing algorithm (Liperda et al. 2020). Affi et al. (2013) reporting that SA finds
all known optimal solutions (20 out of 30) instances in a very short computation
time compared to other methods for solving a variant of VRP. Therefore, this study
adopt the problem of two-echelon vehicle routes for agricultural mapping operations
using drones (2EVRP-MOD) to be implemented for agricultural area mapping and
develop a simulated annealing algorithm to solve the problem.
The remainder of this paper is organized as follows: In Sect. 2, we shows the
methodology being used for this study including formulation of the two-echelon
vehicle routing with drone model and the procedure of simulated annealing algorithm.
We demonstrate the numerical experiments of this problem and discussion of the
result in Sect. 4. Finally, the conclusion and future research direction are discussed
in Sect. 5.
Two-Echelon Vehicle Routing Problem for Agricultural Area Mapping … 175

2 Problem Description and Mathematical Model

The 2EVRP model is widely used in delivering goods with a combination of ground
vehicles (trucks) and drones. The objective of this model also varies from cost mini-
mization or carry out relief mapping assessments on research. This study aims
to minimize the total operating time for mapping agricultural land inventories in
rice fields in Subang, West Java, Indonesia. The data needed for this research are
including: the type of drone used, the location, and the area included in the image.
One ground vehicle is assumed to carry one drone. The land inventory was obtained
from the sentinel-2A satellite imagery and the NDVI vegetation index processing for
the mapping area on May 18, 2021. The number of stopover point and mapping point
for this study can be seen in Table 1. Furthermore, an illustration of one instance can
be seen in Fig. 1. The following discussion describe the mathematical formulation
and the procedure of the solution algorithm.
This model is adopted from the proposed model by Liperda et al. (2020). The
mathematical formulation can be stated as follows. Let G = (N, E) be a graph where
N = C ∪ S is a set of nodes. S is a set of nodes in the first echelon in which S =
S0 ∪ {1} consist of S0 = {2 … |S|} represents the set of stopover nodes and [7]
represent the depot or the disaster relief centrecenter. While C = {|S| + 1 … |S|
+ |C|} represents the nodes in the second echelon which are the mapping points

Table 1 Number of stopover

Initial temp Final temp Alpha MaxIteration
points and mapping points at
each benchmark instance 200 0.0001 0.5 6000

Fig. 1 Illustration of the distribution of point map points in the mapping area
176 A. A. N. Pewira Redi et al.

or target points. Set E = E1 ∪ E2 is the set of edges (i,) such the edge set E1 =
{(i, j): i, j ∈ S} includes the set of edges connecting all nodes in the first echelon.
Meanwhile, set E2 = {(i,): i ∈ S0, j ∈ C} includes the set of edges connecting all
the satellites in the second echelon to the mapping points. Each mapping point has a
non-negative mapping time or surveillance time to gather information according to
the coverage area denoted as Di where i ∈ C. Each pair of nodes in the first echelon
(i,) is associated with traveling time Csij where edge(i, j) ∈ E1. While each pair of
nodes in the second echelon (i,) is associated with traveling time Ccij where edge(i, j)
∈ E2. In this model, each vehicle can carry one drone to serve each route v in a set of
routes V. The mapping operations of each route are limited by the drone flying time
limit, which is denoted as Qd v where v ∈ V. The formulation is stated as follows.
Sets:
S Set of all nodes the first echelon (including the depot)
So Set of all nodes of Stopover Point (without depot).
C Set of all nodes of mapping point
V Set of all vehicles in the first echelon
Parameters:
C Si j travel time of vehicle from gathering point or stopover point i to stopover point
j in the first echelon
C Ci j travel time of drone stopover point I to mapping point j in the second echelon
Di required mapping time for capturing the area of a mapping point i
Q dv maximum flight time of drone
L Sufficiently large number
Decision variables:
xi jv a binary variable that equals to one of vehicle v is traveled from node i to j
where edge (i, j) ∈ E1; otherwise it is zero
zi A binary variable equals one if the stopover point is visited where i ∈ S;
Otherwise, it is zero
yi j A binary variable equals one if drone is traveled from node i to j where edge
(i, j) ∈ E2; Jika tidak, nol
Q j The total amount of time needed for taking pictures in a stopover point j where
j∈S
T j The total amount of time needed for traveling in the second echelon, where j
∈S
aiv The total accumulated amount of information gathering time has been used up
to stopover point i using vehicle v where i ∈ S dan v ∈ V
biv The total accumulated amount of mapping time has been used up in a stopover
point i using vehicle v where i ∈ C dan v ∈ V
Objective
{{{ {{ {{
MinimizeZ = C SI J xi jv + CC I J yi j + Di yi j (1)
i∈S j∈S v∈V i∈C j∈S0 i∈So j∈C
Two-Echelon Vehicle Routing Problem for Agricultural Area Mapping … 177

Subject to
{{
xi jv = z j ∀ j ∈ So (2)
i∈S v∈V
{{
xi jv = z i ∀i ∈ So (3)
j∈S v∈V

{ { ∀l ∈ So
xilv − xl jv = 0 (4)
∀v ∈ V
i∈S j∈S
{
x1 jv ≤ 1 ∀v ∈ V (5)
j∈So
{
xi1v ≤ 1 ∀v ∈ V (6)
i∈So
{
Di yi j ≤ Q j ∀ j ∈ So (7)
i∈C

Q j ≤ L ∗ zJ ∀ j ∈ So (8)

{
yi j = 1 ∀i ∈ C (9)
i∈S0

( ) ∀i ∈ S
aiv + Q j − L 1 − xi jv ≤ a jv (10)
∀ j ∈ S0
{{ {{
xi jv Q j + CCi j yi j ≤ Qdv ∀v ∈ V (11)
i∈S j∈S i∈C j∈S0

The objective function (1) seeks to minimize the total mapping operation time
that consists of the total time traveled by ground vehicles, the total time traveled
by drone, and the total mapping or information gathering time at each target point.
Constraints (2) and (3) restrict that the stopover points are visited only once in the
routing decision of the first echelon. Constraint (4) ensures the consecutive movement
of vehicles in the first echelon. Constraints (5) and (6) ensure that a ground vehicle
starts and returns to the depot. Constraints (7) and (8) are utilized to denote the total
mapping time at each stopover point. Constraints (9) restrict a target point to be
accessed from only one stopover point. Constraints (10) ensure the flow continuity
of the ground vehicle. L is a sufficiently large number that should be greater than or
equal to the total mapping time for each target point. Lastly, constraints (11) ensure
the flying capacity of the drone at each route does not exceed its limit.
The algorithm used in this study is Simulated Annealing (SA). SA algorithm is
inspired by a slow cooling process on material, commonly called annealing. The
178 A. A. N. Pewira Redi et al.

purpose of this cooling process is to crystallize materials and reduce damage. A

material’s temperature needs to be reduced slowly in an annealing process to prevent
damage due to its irregular structure (Sylvester 2018). SA generates a new solution
that considers the previous solution within certain criteria. In each iteration, a new
solution can be accepted if it meets the Metropolis criteria. If the solution generated
by the iteration is better than the solution in the previous iteration, the new solution is
accepted and used as the current best solution in future iterations. The steps of the SA
algorithm can be seen in the pseudocode in Algorithm 1. SA algorithm also adopts
the thermodynamics method, namely the Boltzmann function shown in formula (13).
( )
E (m)
P (m) ≈ ex p − (13)
kT

Algorithm 1: Simulated Annealing Heuristics

Paramter initialization
S = generate the initial solution (); T0 = Initial temperature; N = umber of iteration
at each temperature;
T = T0
While T < T f inal ;
{
Until (N ≤ 1-Iter).
{
Generate Solution S’ in the neighborhood S.
if f (S' ) < (f (S).
S ← S' .
else
△ = f (S ' ) – f (S)
r = random();
if (r < exp(− △k ∗ T )).
S ← S' .
}
T = α x T;
}

3 Computational Study

Data processing is carried out using AMPL (mathematical programming tools) with
Gurobi solver and Microsoft Visual Studio 2019 with device specification are 8 GB
RAM, Intel® Core Processor™ i7-6500U, 64-bit operating system type and the
programming language used was C#.. The number of drones used is one vehicle with
maximum drone capacity flight time, which is 30 min. The time required to take a
Two-Echelon Vehicle Routing Problem for Agricultural Area Mapping … 179

Fig. 2 Illustration of the vehicle route from the depot to the mapping point

picture depends on the area. This value is adopted from one of the studies conducted
by Boccardo et al. (2015). Illustration of the maps can be seen in Fig. 2.Data prepa-
ration regarding to the initial map data and configuration of drone is predetermined.
The satellite image being used is Sentinel-2A satellite imagery to see the distribu-
tion of rice fields. The type of drone used is the DJI Phantom 4 Pro V2.0 with the
following specifications of flight time is 30 min and flying speedis 5.56 m/s. The
classification of rice fields using the NDVI vegetation index. While the process of
data processing is carried out after the post-stage. It is assumed that each vehicle can
carry one drone.
The final parameter being used for this SA algorithm are shown in Table 2 are
the values that generate the lowest average process time for the entire data. Each
area instance was experimented ten times with three neighborhoods with the same
probabilty: swap, insert, and reverse. Table 3 shown the results show that the average
objective value of SA for all instances is 171.079. Meanwhile the optimal objective
value reported by AMPL is 142.820. This result indicates that SA result is not yet
able to perform well. However, there are still a lot of potential procedure can be
implemented to be able to improve the SA result. Meanwhile in term of computational
time, the SA algorithm can generate solution with average computational time of
0.35 s.
180 A. A. N. Pewira Redi et al.

Table 2 Number of stopover points and mapping points at each benchmark instance
Instance AMPL Simulated annealing
1st echelon 2nd echelon Mapping time Total
Instance1.1 40.040 32.599 39.988 112.627 104.252
Instance1.2 34.328 47.221 37.593 119.142 122.907
Instance1.3 20.550 43.640 38.369 102.559 94.565
Instance1.4 6.386 40.725 24.331 71.442 72.057
Instance1.5 20.408 70.459 32.591 123.458 145.060
Instance1.6 44.559 27.130 19.337 91.026 220.155
Instance1.7 90.922 70.590 50.338 211.850 240.897
Instance2.1 45.312 37.659 38.274 121.245 107.773
Instance2.2 60.356 72.543 60.768 193.667 223.602
Instance2.3 26.126 59.647 43.656 129.429 125.737
Instance2.4 64.675 101.748 46.190 212.613 318.924
Instance2.5 83.899 79.508 61.375 224.782 277.016
Average 142.820 171.079

4 Discussion

The AMPL application is used to obtain optimal results with exact methods to validate
the Linear mathematical model. The parameters entered for the model were given
by observing the review of the characteristics and the drone flight of the number of
attractions to generate a score for each node. The route distance of the car vehicle
determines the arc score, while the distance from the flying drone is based on user
satisfaction. The number of attractions and the distance between attractions are also
obtained from google maps for the Mapping Area in Subang, West Java. The data
is executed, and the results show that the optimal solution is within the constraints
obtained by using the annealing simulation algorithm.

5 Conclusion

This study explore the application two echelon vehicle routing problem for drone
mapping operation of agricultural area. The problem is known as an NP-hard
problem. Therefore, a simulated annealing heuristics is proposed to find the solution
for the problem. The proposed algorithm is compared to the result obtained by the
exact solution solver that indicate the optimal solution. The result shows that in terms
of the solution quality, the proposed algorithm are not yet able to provide optimal
solution value same as provided by AMPL. In addition, the computational time of the
proposed algorithm are considerably fast. For further study, the difference in results
Two-Echelon Vehicle Routing Problem for Agricultural Area Mapping … 181

between AMPL and SA can be analyzed using statistical analysis. Other than that,
many improvement need to be performed on SA for example by introducing a better
initial solutions, better parameter tuning, and hybrid method so that it able to have a
performance that very near to optimal solution.

Acknowledgements This research is funded by the Indonesian Ministry of Education and Culture
(Kementrian Pendidikan dan Kebudayaan) under the World Class Research (WCR) Scheme.

References

Afifi S, Dang D-C, Moukrim A (2013) A simulated annealing algorithm for the vehicle routing
problem with time windows and synchronization constraints, pp 259–265. https://doi.org/10.
1007/978-3-642-44973-4_27
Boccardo P, Chiabrando F, Dutto F, Tonolo F, Lingua A (2015) UAV deployment exercise for
mapping purposes: evaluation of emergency response applications. Sensors 15(7):15717–15737.
https://doi.org/10.3390/s150715717
Darmanto (2017) Limitasi Laju Alih Fungsi Lahan Pertanian Ke Non Pertanian Melalui Pemben-
tukan Brigade LP2B Berintegritas Untuk Penguatan Dan Kedaulatan Pangan. Diklatpim Tingkat
III Angkatan XXIX. http://bpsdmd.jatengprov.go.id/eproper/cetakinovasi/?nourut=52
Gunawan B (2013) Pemetaan dan inventarisasi data komoditas pertanian dan informasi kondisi
lahan di kabupaten Kudus berbasis sistem informasi geografis
Liperda RI, Pewira Redi AAN, Sekaringtyas NN, Astiana HB, Sopha BM, Maria Sri Asih A
(2020) Simulated annealing algorithm performance on two-echelon vehicle routing problem-
mapping operation with drones. In: IEEE international conference on industrial engineering and
engineering management. https://doi.org/10.1109/IEEM45057.2020.9309923
Maini P, Sujit PB (2015) On cooperation between a fuel constrained UAV and a refueling UGV
for large scale mapping applications. In: 2015 international conference on unmanned aircraft
systems (ICUAS), pp 1370–1377.https://doi.org/10.1109/ICUAS.2015.7152432
Sylvester G (2018) E-agriculture in action: drones for agriculture. Food and agriculture organization
of the United Nations and International
Utomo B (2018) Drone Untuk Percepatan Pemetaan Bidang Tanah. Media Komunikasi Geografi
18(2):146. https://doi.org/10.23887/mkg.v18i2.12798
Wang Z, Sheu J-B (2019) Vehicle routing problem with drones. Transp Res Part B Method 122:350–
364. https://doi.org/10.1016/j.trb.2019.03.005
Sustainable, Social and Legal Challenges
in Supply Chain Management
and Logistics
Environmental and Economic Value
Prediction of Waste Electrical
and Electronic Equipment Recycle Using
Reverse Logistics: The Case of China’s
Waste Mobile Phones

Sung Woo Kang and Yu Quan

1 Introduction

As product lifecycles shorten due to continuous innovation, WEEE has gradually

increased to 44.7 (Mt) in 2016 (Kumar et al. 2017). Only 20% (8.9 Mt) of these
WEEEs are properly documented, collected, and recycled, and the remaining 80%
(35.8 Mt) are discarded as residual waste or are untraced and unreported (Baldé et al.
2017). Asia generated the largest amount of WEEE (18.2 Mt) in 2016, followed by
Europe (12.3 Mt), the Americas (11.3 Mt), Africa (2.2 Mt), and Oceania (0.7 Mt)
(Baldé et al. 2017). According to the Environmental Protection Agency, only 25%
of WEEE are recycled, of which computers account for 38%, televisions 17%, and
mobile phones only 8% (EPA 2014). The total global value of WEEE is 55 billion
euros. Most of the WEEEs that are unaccounted for become landfill, and the various
hazardous materials contained in WEEE can cause serious environmental pollution
(Baldé et al. 2017). Consequently, it is important to manage and reduce WEEE
(Cui and Zhang 2008). However, despite the importance, the volume of WEEE and
the amount that is disposed continue to grow globally (Niu and Li 2007). This is
explained by the fact that there is no specific system or network to control WEEE
properly. In addition to economic losses, disposal of WEEE, such as in landfills, can
cause environmental problems and human diseases (Olympio et al. 2017; Ongondo
et al. 2011; Widmer et al. 2005; Zhang and Xu 2016). Specifically, the landfill of

S. W. Kang (B)
Department of Industrial Engineering, Inha University, 100 Inha-Ro Michuhol-Gu,
Incheon 22212, South Korea
e-mail: [email protected]
Y. Quan
College of Economics and Management, Yanbian University, 977 Gongyuan Rd, Yanji, Jilin,
China
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 185
P. Golinska-Dawson et al. (eds.), Smart and Sustainable Supply Chain
and Logistics—Challenges, Methods and Best Practices, EcoProduction,
https://doi.org/10.1007/978-3-031-15412-6_12
186 S. W. Kang and Y. Quan

WEEE causes about 572 million euros worth of environmental damage (Baldé et al.
2017; Rabl et al. 2008).
In recent years, many global companies have used various SCM- (Supply Chain
Management) related strategies and techniques to solve these problems (Chen et al.
2012). For example, Apple Inc. recycles raw materials from old iPhones and re-uses
them to manufacture new iPhones (Apple 2017). In addition, many researchers have
also made efforts to increase the efficiency of various measures such as recycle and
reuse (Buekens and Yang 2014; Cucchiella et al. 2015; Gu et al. 2016). In particular,
reverse logistics (RL), a field of SCM, has attracted much attention as an eco-friendly
problem-solving method (Tsai and Hung 2009). RL, which helps sustainable manage-
ment strategy, is used in various fields such as aircraft, computer, and mobile phone
equipment industries. Unlike forward logistics (FL), which represents the process
from product manufacturing to consumer, RL is a concept that refers to the flow from
disposal or return of consumer. However, research to solve this problem is mostly
related to processing, construction, or theoretical model building (Ayvaz et al. 2015).
To provide a more realistic and quantitative indication, this paper predicts the envi-
ronmental and economic values of WEEE recycle using real industrial cost structure
and RL.

2 Literature Review

SCM is the concept that encompasses the entire lifecycle of a product from production
to consumer and FL and RL (Sari 2017). Figure 1 shows the overall supply chain
flows.

2.1 Forward Logistics

FL refers to the flow from the supplier through the manufacture and transportation of
products to delivery to consumer. Research related to FL has been primarily focused
on methodologies such as production methods (Lean manufacturing, Just In Time,
etc.), and transportation methods (milk-run logistics, cross docking systems, etc.)
from a supplier to a customer through a manufacturer and a retailer (Cua et al. 2001;
Kovács 2011; Shah and Ward 2003; Yu and Egbelu 2008). These studies were aimed
at cost reduction and service improvement. This paper will deviate from these goals
and discuss innovative value creation and environmental protection using RL, EOL
analysis, and WEEE recycling.
Environmental and Economic Value Prediction of Waste Electrical … 187

Fig. 1 Supply chain flow chart

2.2 Reverse Logistics

RL plays an important role in building sustainable systems such as recycling WEEE

(Kilic et al. 2015). In general, RL is defined as a series of activities such as recycling
products, collection, disassembly, cleaning, reassembly, distribution, and marketing
of reverse supply chain processes for the purpose of recouping value or proper
disposal (González-Torre et al. 2004; Rogers and Tibben-lembke 2001).

2.2.1 End-of-Life

There are two kinds of studies on RL, namely research on return logistics and research
on EOL (Ravi et al. 2005). In particular, EOL is a stage in which products are generally
reused, remanufactured, recycled, or disposed of through several stages of recovery
as shown in Fig. 2. (Kang et al. 2013).

Reuse

Reuse means using a product for longer than the existing lifecycle in terms of the
original function and purpose. For reuse, depending on the condition of the product,
a simple repair can be carried out if necessary (Lu et al. 2017).
188 S. W. Kang and Y. Quan

Fig. 2 General flow of EOL products

Remanufacture

Remanufacture is an important element in sustainable production and refers to

making a broken product look like new by replacing or repairing component parts
(Yang et al. 2016). WEEE remanufacturing, especially for EOL products, has difficul-
ties in system integration, data exchange, and resource management. Various studies
have been carried out to solve these difficulties such as introducing a new EOL
option known as resynthesis or a utilization of the cloud manufacturing paradigm in
the WEEE remanufacturing environment (Wang and Wang 2015; Kang et al. 2013).

Recycle

Recycle means that products that have reached the end of their life are decomposed
to the raw material level and used again as a resource to make new products. If
collection of WEEE increases, this can result in WEEE recycling (Geyer and Blass
2010).

Disposal

Disposal relates to the abandonment of the EOL product without using a relatively
productive method, such as the aforementioned three methods. The main disposal
methods of EOL products are landfill and incineration (Kulhawik 2016). Although
WEEE contains a lot of valuable resources, most are still disposed in landfills (Geyer
and Blass 2010).
Environmental and Economic Value Prediction of Waste Electrical … 189

3 Methodology

This section builds a basic model for forecasting the value of economic and
environmental benefits of WEEE recycling. Figure 3 shows the WEEE recycling
process.
First, to recycle WEEE, products and components should be classified as recy-
clable and non-recyclable. Then, non-recyclable products are disposed of and recy-
clable products are disassembled into smaller parts. Disassembled parts are also
classified as recyclable and non-recyclable parts, and the non-recyclable parts are
disposed of. Finally, raw materials are extracted from recyclable parts. Table 1 shows
the process, factors and their definitions for model building (Guo and Yan 2017; Rabl
et al. 2008).
Formula 1 is the basic model of value prediction using the factors above.

Formula 1. Value prediction model

Value = Raw material content · WEEE · Collection rate · Recycling rate

− Landfill cost · WEEE · (1 − Collection rate · Recycling rate)

Fig. 3 WEEE recycling process

Table 1 Process, factors and

Process Factor Definition
their definitions of EOL
Recycle WEEE Generation amount
Recycle Collection rate Current collection rate
Disassemble Disassembly cost Disassembly cost per
Gg
Disposal Landfill cost Landfill cost per Gg
Disposal Environmental Environmental costs
cost of WEEE Landfill
Disposal and raw Recycling rate Percentage of
materials recyclable materials
Raw materials Raw material Content of valuable
content raw materials per Gg
Raw materials Raw material Price per Gg of raw
price materials
190 S. W. Kang and Y. Quan

Fig. 4 Flow of EOL

products in case (A)

− Disassembly cost · WEEE · Collection rate

− Environmental cost · WEEE · (1 − Collection rate · Recycling rate) (1)

3.1 Case (A)

Case (A) is the case in which collection of WEEE is not performed at all (Fig. 4).
Formula 2 predicts the value of WEEE recycling in this case.

Formula 2. Prediction formula for case (A)

Value = − Landfill cost · WEEE − Environmental cost · WEEE (2)

3.2 Case (B)

Case (B) is the case in which some WEEE are collected (Fig. 5).
Formula 3 predicts the value of WEEE recycling in this case.

Formula 3. Prediction formula for case (B)

Value = Raw material content · WEEE · Collection rate · Recycling rate

− Landfill cost · WEEE · (1 − Collection rate · Recycling rate)
− Disassembly cost · WEEE · Collection rate
− Environmental cost · WEEE · (1 − Collection rate · Recycling rate)
(3)

Fig. 5 Flow of EOL products in case (B)

Environmental and Economic Value Prediction of Waste Electrical … 191

3.3 Case (C)

Case (C) is the case in which all WEEE are collected (Fig. 6).
Formula 4 predicts the value of WEEE recycling in this case.

Formula 4. Prediction formula for case (C)

Value = Raw material content · WEEE · Collection rate · Recycling rate

− Landfill cost · WEEE · (1 − Recycling rate)
− Disassembly cost · WEEE · Collection rate
− Environmental cost · WEEE · (1 − Recycling rate) (4)

4 Experiment

This section calculates economic and environmental values of recycling WMP, which
is a type of WEEE. WEEE has similar procedural characteristics in various processes
such as collection, decomposition, refurbishment, assembly, and so on (Kim et al.
2006). In this sense, WMP is the subject of this experiment. In addition, China was
selected as the target country because it has the widest territory and the largest number
of people in Asia that generated the largest amount of WEEE (Baldé et al. 2017).
To establish various aspects, the experiment consisted of three detailed cases, and
the estimated values were calculated at five-year intervals from 2000 to 2025. Table
2 shows the definition of the parameters and model parameters for predicting the
WMP value.
Table 3 refers to the price of each parameter, and in this paper CASEA, CASEB
and CASEC are calculated based on the price of Table 3.
In addition, Table 4 refers to the estimated amount of waste mobile phones every
five years based on the duration of mobile phone use and the amount of mobile phone
use of Chinese people, and the unit is ton.

Fig. 6 Flow of EOL products in case (C)

192 S. W. Kang and Y. Quan

Table 2 Parameters of formula for predicting each case

Process Factor Parameters Definition of parameter
Recycle WEEE OP The amount of WMP
Recycle Collection rate CR Current collection rate of
WEEE
Disassemble Disassembly cost DCCC Collection cost
DCLC Labor costs
DCPC Plant operating cost
DCUC Unrecyclable parts disposal
cost
Disposal Landfill cost LCLC Labor costs
LCFC Fuel cost
LCEF Excavator fee
LCBF Bulldozer fee
LCTF Tamping machine fee
LCLF Loader fee
LCCF Cargo truck fee
LCVF Vehicle fee
LCWF Tire Washing Machine fee
LCFF Street flusher truck fee
LCWB Water bill
LCEB Electricity bill
LCLB Landfill bill
LCSC Sewage treatment cost
LCOC Other maintenance costs
LCAE Administrative expenses
Disposal Environmental cost EC Environmental cost by
WEEE Landfill
Disposal and raw material Recyclable rate RR Percentage of recyclable
materials of mobile phones
Raw material Raw material content RCG Gold
RCS Silver
RCCP Copper
RCP Palladium
RCCB Cobalt
Raw material Raw material price RCGp Price of gold
RCSp Price of silver
RCCPp Price of copper
RCPp Price of palladium
Environmental and Economic Value Prediction of Waste Electrical … 193

Table 3 Price for each

Factor Parameters Cost
parameter
Landfill cost LCLC 3.18
LCFC 0.86
LCEF 0.08
LCBF 0.24
LCTF 0.24
LCLF 0.06
LCCF 0.01
LCVF 0.06
LCWF 0.11
LCFF 0.06
LCWB 0.06
LCEB 0.05
LCLB 0.42
LCSC 5.27
LCOC 2.42
LCAE 0.56
LCEC 1.49
Environmental cost EC 10.50
Recyclable rate RR 0.36
Raw material content RCG 280.00
RCS 2000.00
RCCP 190000.00
RCP 100.00
RCCB 168000.00
Raw material price RCGp 42.34
RCSp 0.52
RCCPp 0.01
RCPp 31.06
RCCBp 0.08

Table 4 Amount of waste

Year Reclamation amount (ton)
mobile phones for each year
2000 1300
2005 10,200
2010 30,000
2015 111,600
2020 130,000
2025 140,000
194 S. W. Kang and Y. Quan

4.1 Case (A)

Case (A) means that only landfill costs and environmental costs exist because WMP
is not collected or recycled. Formula 5 is used to predict the value in Case (A).

Formula 5. Formula for prediction of case (A)

Value Y(A) = − (LC LC + LC FC + LC E F + LC B F + LC T F + LC L F + LCC F

+ LC V F + LC W F + LC F F + LC W B + LC E B + LC L B + LC SC
+ LC OC + LC AE + LC EC) · O P − EC · O P (5)

4.2 Case (B)

Case (B) means that WMP is collected at the current level (Bian et al. 2016). There-
fore, there are landfill, environmental, and disassembly costs for recycling and bene-
fits from raw materials from WMP. Formula 6 is used to predict the value in Case
(B).

Formula 6. Formula for prediction of case (B)

Value Y(B) = (RC G · RC Gp + RC S · RC Sp + RCC P · RCC P p + RC P · RC P p

+ RCC B · RCC Bp) · O P · C R · R R − (LC LC + LC FC + LC E F
+ LC B F + LC T F + LC L F + LCC F + LC V F + LC W F + LC F F
+ LC W B + LC E B + LC L B + LC SC + LC OC + LC AE + LC EC)
· O P · (1 − C R · R R) − (DCCC + DC LC + DC PC + DCU C)
· O P · C R − EC · O P · (1 − C R · R R) (6)

4.3 Case (C)

Case (C) means that all WMP are collected. Therefore, there are disassembly costs for
recycling and a small amount of landfill and environmental costs for non-recyclable
parts of WMPs. However, there is a large amount of benefit from raw materials from
such phones. Formula 7 is used to predict the value in Case (C).

Formula 7. Formula for prediction of case (C)

Value Y(C) = (RC G · RC Gp + RC S · RC Sp + RCC P · RCC P p

+ RC P · RC P p + RCC B · RCC Bp) · O P · R R
Environmental and Economic Value Prediction of Waste Electrical … 195

− (LC LC + LC FC + LC E F + LC B F + LC T F + LC L F
+ LCC F + LC V F + LC W F + LC F F + LC W B + LC E B
+ LC L B + LC SC + LC OC + LC AE + LC EC) · O P
· (1 − R R) − (DCCC + DC LC + DC PC + DCU C) · O P
· C R − EC · O P · (1 − R R) (7)

5 Results

This section explores results of the experiment with each case. Calculation of Case
(A), Case (B), Case (C) using the COST of Table 3 is as follows in Table 5.

5.1 Case (A)

Figure 7 shows the predicted value of economic and environmental benefits of WMP
recycling in case (A). In 2000, the predicted value is around −$33,000. Since then,
it has dropped sharply, reaching around −$3.6 million in 2025. In particular, this
figure shows the rapid change from 2010 to 2015.

5.2 Case (B)

Figure 8 shows the predicted value of economic and environmental benefits of WMP
recycling in case (B). In 2000, the predicted value was around $960,000. Since then,
it has increased sharply, reaching around $104 million in 2025. In particular, this
figure shows the rapid change from 2010 to 2015.

Table 5 COST of Case (A),

Year Case (A) Case (B) Case (B)
Case (B), Case (C) for each
year 2000 −33,397 965,267 14,432,654
2005 −262,040 7,573,633 113,240,823
2010 −770,705 22,275,391 333,061,245
2015 −2,867,023 82,864,453 1,238,987,832
2020 −3,339,723 96,526,693 1,443,265,396
2025 −3,596,624 103,951,823 1,554,285,811
196 S. W. Kang and Y. Quan

Fig. 7 Predicted benefits of WMP recycling in case (A)

Fig. 8 Predicted benefits of WMP recycling in case (B)

5.3 Case (C)

As shown in Fig. 9, the predicted value was around $14 million in 2000. Since then,
it has increased sharply, reaching around $1.5 billion in 2025. The amount of WMP
increased sharply from 2010 to 2015 hereby incrementing benefits rapidly during
the same period.

5.4 Comparison of Case (A), (B), and (C)

As shown in Fig. 10, the difference between Case (B) and Case (C) is much larger than
that between Case (A) and Case (B). This means that the current WMP recovery and
recycling rate is very low compared to the ideal situation. In addition, if the recovery
and recycling rate of WMP increase, the model makes additional benefits as large as
the gap between Case (B) and Case (C).
Environmental and Economic Value Prediction of Waste Electrical … 197

Fig. 9 Predicted benefits of WMP recycling in case (C)

Fig. 10 Predicted benefits of WMP recycling in entire case

6 Conclusion and Future Work

This study analyzed the flow of RL and EOL products to create a value prediction
model of WEEE recycling from an economic and environmental viewpoint. The
literature review described forward logistics, RL, and four options regarding EOL
products (reuse, remanufacture, recycle, and disposal). This research proposed the
basic model for prediction through EOL analysis and created three models according
to the basic model of WEEE collection rate. In order to verify the models presented in
this research, the WMP recycling structure of the Chinese industry data set was used
for the experiment. The results from the case study confirmed that the economic
and environmental values significantly increase when the recovery rate of WEEE
198 S. W. Kang and Y. Quan

increases. Specifically, when all WMP are collected, companies can achieve profits
around 14 times higher than the current situation. This study confirms the value
of abstract WEEE recycling as a more quantitative figure. This figure can be an
indication of more intuitive assessment of the effectiveness of WEEE recycling.
However, this study focuses on recycling among various flows of EOL such as reuse
and remanufacture and as a result, there may be a few errors. In addition, the factors
and parameters of this study will need to be revised and supplemented in future
research. To overcome these limitations, various follow-up studies will be required
including a study that considers the total flow of EOL or the value of WEEE according
to the recycling rate.
In addition, as a future study, economic and environmental value analysis will
be conducted by considering various WEEE such as household appliances, office
supplies, and recycling other raw materials such as plastic.

Acknowledgements Editage proof read the entire paper.

This work has supported by the National Research Foundation of Korea(NRF) grant funded by
the Korea government(MSIT) (2021R1F1A1060753).

References

Apple (2017) Environmental report 2017. Apple 50

Ayvaz B, Bolat B, Aydın N (2015) Stochastic reverse logistics network design for waste of electrical
and electronic equipment. Resour Conserv Recycl 104:391–404. https://doi.org/10.1016/j.rescon
rec.2015.07.006
Baldé CP, Forti V, Gray V, Kuehr R, Stegmann P (2017) The global e-waste monitor
Bian J, Bai H, Li W, Yin J, Xu H (2016) Comparative environmental life cycle assessment of waste
mobile phone recycling in China. J Clean Prod 131:209–218. https://doi.org/10.1016/j.jclepro.
2016.05.047
Buekens A, Yang J (2014) Recycling of WEEE plastics: a review. J Mater Cycles Waste Manag
16:415–434. https://doi.org/10.1007/s10163-014-0241-2
Chen CC, Shih HS, Shyur HJ, Wu KS (2012) A business strategy selection of green supply chain
management via an analytic network process. Comput Math Appl 64:2544–2557. https://doi.org/
10.1016/j.camwa.2012.06.013
Cua KO, McKone KE, Schroeder RG (2001) Relationships between implementation of TQM, JIT,
and TPM and manufacturing performance. J Oper Manag 19:675–694. https://doi.org/10.1016/
S0272-6963(01)00066-3
Cucchiella F, D’Adamo I, Lenny Koh SC, Rosa P (2015) Recycling of WEEEs: an economic
assessment of present and future e-waste streams. Renew Sustain Energy Rev 51:263–272. https://
doi.org/10.1016/j.rser.2015.06.010
Cui J, Zhang L (2008) Metallurgical recovery of metals from electronic waste: areview. J Hazard
Mater 158:228–256. https://doi.org/10.1016/j.jhazmat.2008.02.001
EPA (2014) Statistics on the management of used and end-of- life electronics
Geyer R, Blass VD (2010) The economics of cell phone reuse and recycling. Int J Adv Manuf
Technol 47:515–525. https://doi.org/10.1007/s00170-009-2228-z
González-Torre PL, Adenso-Díaz B, Artiba H (2004) Environmental and reverse logistics policies
in European bottling and packaging firms. Int J Prod Econ 88:95–104. https://doi.org/10.1016/
S0925-5273(03)00181-6
Environmental and Economic Value Prediction of Waste Electrical … 199

Gu Y, Wu Y, Xu M, Wang H, Zuo T (2016) The stability and profitability of the informal WEEE
collector in developing countries: a case study of China. Resour Conserv Recycl 107:18–26.
https://doi.org/10.1016/j.resconrec.2015.12.004
Guo X, Yan K (2017) Estimation of obsolete cellular phones generation: a case study of China. Sci
Total Environ 575:321–329. https://doi.org/10.1016/j.scitotenv.2016.10.054
Kilic HS, Cebeci U, Ayhan MB (2015) Reverse logistics system design for the waste of electrical
and electronic equipment (WEEE) in Turkey. Resour Conserv Recycl 95:120–132. https://doi.
org/10.1016/j.resconrec.2014.12.010
Kim K, Song I, Kim J, Jeong B (2006) Supply planning model for remanufacturing system in reverse
logistics environment. Comput Ind Eng 51:279–287. https://doi.org/10.1016/j.cie.2006.02.008
Kovács A (2011) Optimizing the storage assignment in a warehouse served by Milkrun logistics.
Int J Prod Econ 133:312–318. https://doi.org/10.1016/j.ijpe.2009.10.028
Kulhawik K (2016) Direct landfill disposal versus Mechanical Biological Treatment (MBT)
Bezpośrednie deponowanie na składowisku a mechaniczno- biologiczne przetwarzanie odpadów
(MBP) 27:19–23
Kumar A, Holuszko M, Espinosa DCR (2017) E-waste: an overview on generation, collection,
legislation and recycling practices. Resour Conserv Recycl. https://doi.org/10.1016/j.resconrec.
2017.01.018
Lu B, Yang J, Ijomah W, Wu W, Zlamparet G (2017) Perspectives on reuse of WEEE in China:
lessons from the EU. Resour Conserv Recycl 0–1. https://doi.org/10.1016/j.resconrec.2017.
07.012
Niu X, Li Y (2007) Treatment of waste printed wire boards in electronic waste for safe disposal. J
Hazard Mater 145:410–416. https://doi.org/10.1016/j.jhazmat.2006.11.039
Olympio KPK, Gonçalves CG, Salles FJ, da Silva Ferreira APS, Soares AS, Buzalaf MAR, Cardoso
MRA, Bechara EJH (2017) What are the blood lead levels of children living in Latin America
and the Caribbean? Environ Int 101:46–58.https://doi.org/10.1016/j.envint.2016.12.022
Ongondo FO, Williams ID, Cherrett TJ (2011) How are WEEE doing? A global review of the
management of electrical and electronic wastes. Waste Manag 31:714–730. https://doi.org/10.
1016/j.wasman.2010.10.023
Rabl A, Spadaro JV, Zoughaib A (2008) Environmental impacts and costs of solid waste: a compar-
ison of landfill and incineration. Waste Manag Res 26:147–162. https://doi.org/10.1177/073424
2X07080755
Ravi V, Shankar R, Tiwari MK (2005) Analyzing alternatives in reverse logistics for end-of-life
computers: ANP and balanced scorecard approach. Comput Ind Eng 48:327–356. https://doi.org/
10.1016/j.cie.2005.01.017
Rogers DS, Tibben-lembke R (2001) An examination of reverse logistics practices 22:129–148
Sari K (2017) A novel multi-criteria decision framework for evaluating green supply chain
management practices. Comput Ind Eng 105:338–347. https://doi.org/10.1016/j.cie.2017.01.016
Shah R, Ward PT (2003) Lean manufacturing: context, practice bundles, and performance. J Oper
Manag 21:129–149. https://doi.org/10.1016/S0272-6963(02)00108-0
Tsai WH, Hung SJ (2009) Treatment and recycling system optimisation with activity-based costing
in WEEE reverse logistics management: an environmental supply chain perspective. Int J Prod
Res 47:5391–5420. https://doi.org/10.1080/00207540801927183
Wang XV, Wang L (2015) WR cloud: a novel WEEE remanufacturing cloud system. Proc CIRP
29:786–791. https://doi.org/10.1016/j.procir.2015.02.011
Widmer R, Oswald-Krapf H, Sinha-Khetriwal D, Schnellmann M, Böni H (2005) Global perspec-
tives on e-waste. Environ Impact Assess Rev 25:436–458. https://doi.org/10.1016/j.eiar.2005.
04.001
Sung Woo Kang S, Sane C, Vasudevan N, Tucker CS (2013). Product resynthesis: knowledge
discovery of the value of end-of-life assemblies and subassemblies. J Mech Des 136:011004.
https://doi.org/10.1115/1.4025526
200 S. W. Kang and Y. Quan

Yang SS, Nasr N, Ong SK, Nee AYC (2016) A holistic decision support tool for remanufacturing:
end-of-life (EOL) strategy planning. Adv Manuf 4:189–201. https://doi.org/10.1007/s40436-016-
0149-2
Yu W, Egbelu PJ (2008) Scheduling of inbound and outbound trucks in cross docking systems with
temporary storage. Eur J Oper Res 184:377–396. https://doi.org/10.1016/j.ejor.2006.10.047
Zhang L, Xu Z (2016) A review of current progress of recycling technologies for metals from waste
electrical and electronic equipment. J Clean Prod 127:19–36. https://doi.org/10.1016/j.jclepro.
2016.04.004
A Financial Feasibility Study of Zero
Waste Manufacturing: Opportunity
and Value Added of Agricultural
Products on the R9 Route

Supattraporn Saisomboon, Arunrat Sawettham, Sumalee Ngeoywijit ,

and Monika Kosacka-Olejnik

1 Introduction

The Food and Agriculture Organization (FAO) defines food waste as “wholesome
edible material intended for human consumption, arising at any point in the food
supply chain (FSC) that is instead discarded, lost or degraded Factors that influence
food losses and waste; that occur along a food supply chain (Machines, Materials,
Management, Methods, and People) come from the surroundings of a food supply
chain. (Kowalska 2017) Factors influencing food losses and waste; that occur along a
food supply chain (Machines, Materials, Management, Methods, and People) come
from the surroundings of a food supply chain (Kowalska 2017). Food losses or waste
are the quantities of edible material wasted or lost in the food supply chain at various
stages, including harvesting, post-harvest storage and material handling, processing,
distribution and consumption. A large amount of energy and other resources are
consumed for food production and distribution (Gustavsson et al. 2011). Food supply
will need to increase by around 70% from its current levels in order to meet the

S. Saisomboon
Department of Business Management, Faculty of Management Science, Ubon Ratchathani
University, Ubon Ratchathani 34190, Thailand
e-mail: [email protected]
A. Sawettham · S. Ngeoywijit (B)
Department of International Business Management, Faculty of Management Science, Ubon
Ratchathani University, Ubon Ratchathani 34190, Thailand
e-mail: [email protected]
A. Sawettham
e-mail: [email protected]
M. Kosacka-Olejnik
Faculty of Engineering Management, Poznan University of Technology, 2 Jacka Rychlewskiego
St., 60-965 Poznan, Poland
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 201
P. Golinska-Dawson et al. (eds.), Smart and Sustainable Supply Chain
and Logistics—Challenges, Methods and Best Practices, EcoProduction,
https://doi.org/10.1007/978-3-031-15412-6_13
202 S. Saisomboon et al.

world population growth of 9.6 billion by 2050. Food waste is the biggest challenge
in global food security, wherein approximately 20–30% of food waste occurs in the
post-harvest stage of the food supply chain in developing countries. This food waste
generates significant negative environmental effects in addition to the unnecessary
usage (and wastage) of resources consumed in producing the wasted food (Krishnan
et al. 2020).
Mango is a tropical fruit with a long history of production, cultivation and trade,
mainly cultivated in tropical Asian countries. The data statistics of The Food and
Agriculture Organization of the United Nations currently show that there are 103
countries in the world where mangoes are cultivated. The mango is considered a fruit
that has various benefits and properties such as helping to build immunity for the body,
good for liver and kidneys and it can also help reduce nausea or dizziness. But the fruit
is a difficult commodity to keep. Modern technology has been introduced into the food
processing industry which helps preserve the fruit. It also helps to elevate the value
of the fruit as well. Food processing industry plays an important role in enhancing
the economic growth of the country. It increases values for agricultural products in
accordance with close linkage with the agricultural sector, which is in the majority
in Thailand. The food processing industry creates more employment and boosts the
country’s exporting figures. In 2020, the value of Thailand’s mango exports was at
4,602,668,283 THB (Commerce 2021). Due to the COVID-19 pandemic exporting
the product becomes too difficult, resulting in domestic consumption only, this in turn
causes an overflow in the domestic market. C19 has raised food insecurity and safety
concerns, increased supply chain and logistics costs, and radically changed consumer
behavior. On the positive side, the pandemic has improved awareness of food waste
and the importance of self-grown foods (Rejeb et al. 2020). After consumption or
industrial processing of the mango, considerable amounts of seeds are discarded as
waste. We can analyze the amount of waste that may occur by predictive reference
markets, which are base on crowdsourcing, use collective crowd intelligence, and
support many business areas (Czwajda et al. 2019). The seed alone accounts for 20–
25% of the whole fruit. The kernels inside the seed are approximately 45–75% of the
seed and about 20% of the whole fruit. More than one million tons of mango seed
are produced annually and often discarded as waste. Approaches to reduce waste and
add value into new products for consumption (Babaria 2012). At the same time can
cure a serious environmental problem.
The east–west economic corridor (EWEC) is one of the most strategic cooper-
ation in the development of the Economic Corridor, with potential areas along the
three main communication lines, namely the north–south line. North south economic
corridor (NSEC), east–west economic corridor (EWEC) and southern economic
corridor (SEC) of the Economic cooperation development project in the 6 countries of
Greater Mekong Sub-region (Greater Mekong Subregion GMS) which was formed
by cooperation of Thailand, Myanmar, Lao PDR, Cambodia, Vietnam and South
China (Yunnan Province and Guangxi Province). The objective is to promote the
expansion of trade, investment, industry, agriculture and services that create employ-
ment and improve the livelihood of people in the area promoting the development
of cooperation in technology and education to use natural resources that promote
A Financial Feasibility Study of Zero Waste Manufacturing … 203

each other effectively and promoting capacity-building and economic opportunities

in the global trade arena. The initial development of the Economic Corridor is the
development of transportation infrastructure. EWEC or Route 9 (R9) is the most
developed route among the entire Economic Corridor, with a length of 1450 km in
Thailand. The route starts from the coastal port city of Da Nang in Vietnam, passing
through Hue City, Dong Ha City and the city of Laos (Lao Bao) which is a special
economic zone of Vietnam, then takes Highway no. 9 through Savan-Seno Special
Economic Zone of Lao PDR through Savannakhet and crossing over the Mekong
at the 2nd Thai-Laos Friendship Bridge (Mukdahan—Savannakhet) into Thailand,
at Mukdahan Province, Kalasin, Khon Kaen, Phitsanulok, Sukhothai through to the
end of the Mae Sot border checkpoint, Tak Province, crossing into Myanmar by the
Myawaddy passing through the bay of Matama in Mawlamyine. It is considered to
connect the area on the east side of the South China Sea. (Fig. 1) International logis-
tics corridors have many benefits. First, they significantly reduce transit time and
costs, thus expanding access to markets and opening new opportunities for regional
industries. Second, they eliminate infrastructural bottlenecks and increase the tech-
nical and organizational interoperability of national systems. However, it should note
that the development of these corridors requires efficiently functioning intermodal
logistics centers that act as load generators and play the crucial role of nodes in trans-
port systems and supply chains. Therefore, to make them feasible should build new
logistics centers, dry ports (Popp 2018), airports with ensured customs clearance and
storage services, and economic zones with tax incentives along the corridor should
be created (Joanicjusz Nazarko and Czerewacz-Filipowicz 2016).
The EWEC has a large and growing strategic advantage over other transportation
networks in terms of time for delivery between large commercial hubs. Since the
opening of the Second Friendship Bridge, time consuming from Bangkok to Da
Nang has been reduced by around three to three-and-a-half days. (Fig. 2) At present,
the export route to China is from the original route by ship and has to go through
many steps, such as by vessel from Laem Chabang, via Vietnam and Hong Kong,
before reaching Guangzhou or Shanghai of China, it takes almost 2 weeks (average
12–13 days). There is an additional option, which is road transportation. Thailand can
deliver fruit to China by four routes one of them is the R9 route from Mukdahan via
Da Nang, transit Hanoi. and then to Guangxi—Beijing. Cinfady market is the biggest
distribution center of Thai fruit in Beijing. It takes only 36 h by truck from Thailand
(Bangkokbanksme 2015). Nevertheless, as of late June 2020, despite the lack of
scientific evidence, China is demanding that all foreign food shipments guarantee
that the food and food containers have tested for the virus before shipping to China
(Patton 2020).
The province located on the R9 of Thailand is an important mango growing area in
Thailand. Therefore, there is interest in studying the possibility of setting up a mango
processing plant and using this strategic point for transportation for export to China
to increase economic value according with national economic development plan
no.12 aimed at strengthening the economy by sector goal 1: sectoral economy grows
strongly and serves as a base for building economic growth of the country. Including
sustainable development goals (SDGs) goal 9 builds an infrastructure that is resilient
204 S. Saisomboon et al.

Fig. 1 EWEC complementary corridors and routes (Lord 2009)

Fig. 2 EWEC road network (Lord 2009)

to change, promotes industrial development comprehensively and sustainably and

innovation. Sub-goal 9.2 promotes inclusive and sustainable industrial development
and by 2030 increase the industrial sector share in employment and gross domestic
product in accordance with the national context and to double the share in the least
developed countries (Council 2017).
Route number 9 was expected to be the location for setting up a mango oil produc-
tion plant. This paper attempts to fill this research gap within the three suitable
locations by using loading distances method, factors rating method. In the model is
A Financial Feasibility Study of Zero Waste Manufacturing … 205

included criteria, weights and classification. AHP gives the opportunity for under-
standing indicators to help make decision. The export route to China is from the
original route one steps for to reduce production costs in the transportation sector.

2 Methodology

2.1 Collect Mango Data

1.1 Import and export data of mango

1.2 Export value of dried mangos
1.3 Information on mango production in the provincial area on route R9

2.2 Analyze the Optimal Location of the Mango Processing

Plant by Center of Gravity Method
{ {
Xi Qi Yi Qi
X = { Y = {
Qi Qi

Qi = Quantity of goods
Xi = X—coordinate of location i
Yi = Y—coordinate of location i.

2.3 Verify the Results of the Data Analysis to the Analytic

Hierarchical Process (AHP)

The use of the AHP is to help analyze problems in decision making in plant loca-
tion selection or decision making in various fields is popular because AHP is an
effective method. There is a simple and easy to understand operation procedure.
Therefore, the analytical hierarchical process was applied in selecting the factory
location. Therefore, it is something that should be considered highly.
The development process consists of three main steps for both AHP. Step one using
analytical process to define criteria. In this study, the criteria have been identified
based on the raw materials, labors, markets, land prices, transportation, infrastruc-
ture, environmental, society and community, laws and taxes, opportunities in future,
industrial estates, community attitude, political freedom and free trade zones. Next,
206 S. Saisomboon et al.

the 12 experts from a mango manufacturing company located area on route R9.
Tak, Sukhothai, Phitsanulok, Khon kaen, Kalasin and Mukdahan provinces and and
from inquiry 10 samples of experts from 5 groups, namely logistics entrepreneurs,
industrial factory, agricultural product collector, related government agencies and
academic groups. The comparative importance of attributes is provided by the deci-
sion maker using a rating a 1–9 scale. Finally, the AHP method obtains the priority
weights of attributes of each category by computing the vector of matrix.
⎡ ⎤
a11 a12 . . . a1j
⎢ a21 a22 . . . a2j ⎥
⎢ ⎥
⎢ a32 . . . a ⎥
A = ⎢ a31 3j ⎥
⎢ ⎥
⎣... ... ... ⎦
ai1 ai2 . . . aij

aij are the judgments or the relative importance of alternative i over alternative j and
aij = 1 for i − j and aij = l/ aij for i /= j.

2.4 Economic Analysis Data (The Processed Products are

Mango Crisps and Mango Seed Oil)

2.4.1 The Break-Even Point (BEP)

Fixed cost
BEP =
Selling price − V ariable cost

2.4.2 Net Present Value (NPV)

It is the difference between the sum of present value of benefits and sum of present
value of costs for a given discount rate. If a positive value of NPV is obtained when
discounted at the opportunity cost of capital, then the investment is considered viable.

Rt
NPV
(1 + i)t

t = time of the cash flow

i = discount rate
Rt = net cash flow
A Financial Feasibility Study of Zero Waste Manufacturing … 207

2.4.3 Interest Rate of Return (IRR)

The IRR is the discount rate that makes the net present value of cash flow equal to
zero. The investment is considered viable if the calculated IRR is greater than the
opportunity cost of capital.

{
t
Ct
IRR = − C0
t−1
(1 + r)t

Ct = Net Cash Inflow during the period t

r = Discount Rate
t = Number of time periods
C0 = Total initial investment cost.

3 Results

3.1 Import and Export Data of Mango

From the Table 1 of quantity and value of mango imports and exports from 2017–
2020, the export and import grew every year, except the year 2019 when the coro-
navirus outbreak occurred. Thailand imported mangoes as raw material for mango
processingInventory practically accounts for over 70% of the food processing firm’s
overall capital assets in developed and developing economies. Its proper management
is key to promoting the performance, growth, and competitiveness of food processing
firms and their supply chains (Opoku et al. 2021).

Table 1 Volume and value of mango imports and exports during the year 2017–2020 (Commerce
2021)
Quantity (KG)
Year 2017 2018 2019 2020
Export 59,138,566 94,102,385 85,182,868 117,656,521
Import 24,838,355 55,352,592 29,081,368 53,731,619
Value (Bath)
Year 2017 2018 2019 2020
Export 3,326,799,100 4,383,712,861 4,067,105,304 4,602,668,283
Import 219,718,778 289,928,938 331,697,885 803,593,125
208 S. Saisomboon et al.

The significant increases of mango consumption in domestic activity lead to the

accumulation of waste. Mango seed kernels are usually wasted when it is processed.
After consumption or industrial processing of mangoes, approximately 40–60%
waste is generated; 12–15% consists of peels and 15–20% of kernels. According
to mango varieties, the seed represents from about 10 to 25% of the whole fruit
weight. The kernel inside the seed represents from 45 to 75% of the seed and about
20% of the whole fruit. However, more than one million tons of mango seeds are
being treated as waste. Mango kernel extracts are hidden treasures. Because the
kernel, when cold pressed, renders mango kernel butter, not oil. This butter itself is
highly prized and could pose as a good substitute for cocoa or shea butter. Being an
excellent source of essential fatty acids, rich in minerals and vitamins, mango kernel
butter exhibits beneficial moisturizing properties for skin lotions and lubricants. It is
solid at room temperature, smells sweet and nutty in its pure form and has a smooth
creamy color. It has a mild aroma, similar to olive oil.
Semi-solid mango kernel oil is obtained during the refining of mango kernel
butter. This soft yellow coloured oil is said to have a stronger odor than the butter,
and has a melting point of around 23–27 °C, meaning that it effectively melts when
it comes in contact with skin. An average mango kernel contains about 8% to 15%
extract potential (butter and oil). This seed, which is usually discarded, can be used
in cosmetics and beauty products. It’s shelf life is 3–4 years if stored under cool
conditions. Most mango oils are refined and during that process the therapeutic
quality is altered (Karunanithi et al. 2015).
From Table 2, Quantity and value of dried mango exports from 2017 to 2020,
there is an upward trend. But the price per kilogram has a downward trend. The
popularity of mango crisps (freeze dried mango) and mango seed oil processing is
trending these days. Lean production is a well-established managerial concept that
helps companies provide customer value and reduce cost (Golińska 2014).
Euromonitor forecasts that the fruit snack market is expected to grow in value
at 15–20% annually, bringing the market size to 2740 million THB in 2023
compared to the current market size. The growth rate of the market size is increasing
(Suwannachote 2018).
Laokasemsukwong (2013) A study on the potential of the dried fruit industry in
Thailand shows that Thailand has an advantage to export dried fruit products to the
world. The factors that contribute to competitive advantage are abundance in fruits,
skilled workers, increasing demand for dried fruit from Thailand and support from
the government. However, entrepreneurs have to face tough competition from China
and Vietnam and deal with the volatility of agricultural prices.

Table 2 Export value of dried mangoes during the year 2017–2020 (Commerce 2021)
Year 2017 2018 2019 2020
Quantity (KG) 1,619,615 2,594,805 3,135,936 4,959,761
Value (Bath) 438,027,122 619,352,914 708,886,363 991,674,347
Price (Bath): 1 kg 270.4513863 238.6895794 226.0525607 199.9439786
A Financial Feasibility Study of Zero Waste Manufacturing … 209

Table 3 Information of mango production in the provincial located on route R9 during the
year 2017–2020 (Sources Information Technology and Communication center: Department of
Agricultural Extension (Center 2021))
Mango harvested produce (KG)
Year 2017 2018 2019 2020 Total
Tak 1,888,510 1,812,500 1,115,500 298,250 5,114,760
Sukhothai 27,723,012 43,021,490 28,559,076 50,825,583 150,129,161
Phitsanulok 23,700 16,506,130 11,906,300 41,897,450 70,333,580
Khon kaen 13,556,087 7,811,028 7,557,352 5,867,379 34,791,846
Kalasin 281,213 231,654 218,357 621,868 1,353,092
Mukdahan 209,960 57,093 133,593 351,000 751,646
Total 43,682,482 69,439,895 49,490,178 99,861,530 262,474,085
Remark Mangoes including Khiew Sawoey, Raet, Golden, okrong, Falan, Chotanan, Mahachanok
and Kaew mangoes

From Table 3 of the provinces located on route R9, the volume of mango between
the year 2017– 2020 in total was 262,474,085 kg. Sukhothai has the highest volume
of mangoes at 50,825,583 kg. Followed by Pitsanulok at 41,897,450 kg, and Khon
Kaen at 5,867,379 kg.
.

3.2 Location Analysis Data

Location of an international logistics center in Poland as a part of the One Belt One
Road initiative. The method used is the analysis of literature sources. As the choice
and location of an international logistics center in Poland on the New Silk Road is
the new one. (Kauf and Laskowska-Rutkowska 2019).
From a detailed study and information collected by a mode of road transport,
which is, a 10-wheel truck carries 1.45 tons/trip of mango for mango processing. If
we consider the inevitable imbalance in intercity freight demand, the efficient trans-
port of empty trucks becomes necessary to a certain extent (Hirata and Fukaya 2020).
The location was analyzed using the center of gravity technique (Center of Gravity
Method). Azeem and Hussain (2012) applied the Gravity Model to analyzing the
geographical pattern of international trade. It was found that the mango consolida-
tion place should be set up at the geographic coordinates X = 16.93751677, Y =
100.295235, which is at Wat Bot district in Phitsanulok province as shown in Table 4.
210 S. Saisomboon et al.

Table 4 Location analysis data by center of gravity method

Latitude (X) Longitude (Y) Mango volume in Wi X Wi Y
2020 (Kg)
Tak 16.679717 98.572691 298,250 4974725.595 29399305.09
Sukhothai 17.035915 99.891168 50,825,583 865860311.8 5077026850
Phitsanulok 16.910321 100.356777 41,897,450 708499328.6 4204693047
Khon kaen 16.349338 102.806863 5,867,379 95927762.45 603206829
Kalasin 16.578963 103.967506 621,868 10309926.56 64654065.02
Mukdahan 16.621893 104.431703 351,000 5834284.443 36655527.75
Total 99,861,530 1,691,406,339 10,015,635,624 16.93751677 100.295235

3.3 Verify the Analysis of the Optimum Location

of the Mango Factory by Means of AHP

Thomas L. Saaty developed the Analytical Hierarchy Process (AHP) theory in 1970.
The principle is to divide the problem structure into individual layers. The first layer is
Goal, Criteria, Criteria. Sub-criteria and Alternatives, respectively (Saaty and Vargas
2001), analyze the best alternatives. Analyzing and comparing (Trade-off) the selec-
tion criteria for each pair individually (Pairwise) makes it easier to decide. They have
graded the comparative significance of each criterion according to importance. After
scoring to prioritize the criteria, consider analyzing the options one by one according
to the specified criteria until they meet all the criteria. If the rating of importance or
preference is reasonable, Consistency will be able to prioritize alternatives to find the
best option. Supporting investment choices in oil tank farm (Meteethunyarat 2020),
biomass plant using AHP (Phankong et al. 2020) and Multi-criteria comparative
analysis method can create a hierarchy of optimal locations of logistics centers. In
this method, the final set of criteria is not strictly defined and can be supplemented
by criteria relevant to a particular investor and the investment. When deciding on
financial investments, it is advisable to consider several criteria (including all kinds
of indicators, simple and discounted stimulators, and development), chosen based on
knowledge and literature and tested on the example of the Lubuskie Province. This
approach allows the use of the proposed method regardless of legal conditions and
geographical location (Witkowski et al. 2018).
The review of related research found that AHP location determination was used
to determine local and national locations. Therefore, it is why the researcher has
adopted the AHP method. So, come check out the location of the mango processing
plant. The researcher has designed the decision-making with AHP as shown in Fig. 3.
The factors considered for applying in AHP are raw material sources, labor
sources, markets, land prices, transportation, infrastructure, environmental, society
and community, laws and taxes, opportunities in future, industrial estates, commu-
nity attitude, political freedom and free trade zones and investment promotion zones
which was used to select the location of the factory from 3 alternative locations,
A Financial Feasibility Study of Zero Waste Manufacturing … 211

Fig. 3 The process of decision-making with AHP

namely Muang District, Sukhothai Province. Wat Bot District Phitsanulok Province
and Puey Noi District Khon Kaen Province as shown in the following Fig. 4 and
Table 5.
Then the experts in each field consider and scrutinize the factors. The results of the
consideration revealed that the top 5 most important factors are raw material sources,
transportation, labor sources, public infrastructure and investment promotion zones
which can create a decision chart as shown in the following Fig. 5.
From a study of the importance of all 5 factors, it was found that the most important
factor influencing the selection of the agricultural processing plant location was
the raw material sources, transportation, labor sources, public infrastructure and
investment promotion zones respectively as shown in the Table 6.
The CR calculating value 0.10, comparisons by pairs are consistent within accept-
able criteria. The weight of each factor that influences the location selection. It is
consistent and acceptable. Then, the total ranking of the alternatives was calculated
by weighting the importance of the evaluation criteria for each factor. It was found
that the most appropriate location for the processing mango plant is Wat Bot District,
Phitsanulok Province as shown in the following Table 7.
This conforms to the industrial area development project to support investment
of the Department of Industrial Works. Furthermore, stressing the importance of
212 S. Saisomboon et al.

Fig. 4 Map of 3 alternative locations (google map)

Table 5 Latitude and

Location Latitude (X) Longitude (Y)
longitude of 3 alternative
locations Muang district, Sukhothai 17.035915 99.891168
Province
Wat Bot district, Phitsanulok 16.93751677 100.295235
Province
Puey Noi district, Khon Kaen 16.349338 102.806863
Province

Fig. 5 A hierarchical chart or a model of decision making

A Financial Feasibility Study of Zero Waste Manufacturing … 213

Table 6 The evaluation criteria weight value for each factor

Criteria Raw Transportation Labor Infrastructure Investment Mean
materials sources Promotion (X)
sources zones 100%
(%)
Raw materials 0.43 0.60 0.39 0.29 0.23 38.80
sources
Transportation 0.14 0.20 0.39 0.29 0.23 25.00
Labor sources 0.14 0.07 0.04 0.03 0.08 7.20
Infrastructure 0.14 0.07 0.13 0.29 0.23 17.20
Investment 0.14 0.07 0.04 0.10 0.23 11.58
Promotion
Zones
Total 1.00 1.00 1.00 1.00 1.00 100

Table 7 The overall weight criteria of each option

Criteria Raw Transportation Labor Infrastructure Investment Total
materials sources promotion Ranking
sources zones
0.39 0.25 0.17 0.12 0.07
Muang 0.22 0.04 0.05 0.02 0.01 0.33
District,
Sukhothai
Province
Wat Bot 0.05 0.14 0.10 0.07 0.04 0.40
District
Phitsanulok
Province
Puey Noi 0.11 0.07 0.02 0.03 0.02 0.27
District Khon
Kaen Province

investment in the potential industries of Phitsanulok Province, ranking number one

is the Agro-processing industry which has a strong point of being the transportation
hub of the North. Phitsanulok is one of the provinces located along the EWEC and
NSEC Economic Corridors and is also a center for agricultural products, vegetables,
and fruits (Works 2016).

3.4 Economic Analysis Data

Farmers in Phitsanulok province must steam treat mangoes to avoid attracting golden
flies before export to prevent outbreaks in the destination countries. The farmers have
214 S. Saisomboon et al.

to carry mangoes to steam plants in Chiang Mai, Chon Buri, Nakhon Pathom, farmers
must also bear costs of approximately 7 THB per kilogram. During the Covid-19
crisis, over 100 tons of mango products of the Phitsanulok farmers were damaged,
due to the steam room at Chanthaburi province not being available. Another reason is
that the mangoes are not delivered to the steam plant on time. If the government could
offer support by building a steam plant located in Phitsanulok, the mango farmers
would be relieved (News 2020).
A feasibility study of investing in the construction of a mango processing plant in
Phitsanulok Province by using the investment structure as follows: 30% own capital
and 70% from financial institutions based on the cost calculation from Lorenzo
Stratta: Economic Analysis of a Freeze-Drying Cycle (Stratta et al. 2020). The
economic feasibility study found that there was a high investment in the construction
of a factory and buying machines. But mango crisps and mango seed oil value are
also high and it tends to increase consumption in the future. Therefore, it is suitable
for investment in a buildup of mangoes processed factories for processing crispy
mangoes and mango seed oil. The researchers used economic tools to make it easier
to analyze the results as follows:
1. Break-even point is 130,097,639 THB
2. Net present value NPV is 72,970,316 THB, which is greater than 0 means it is
worth it to invest
3. The internal rate of return for the IRR project is 27%.

4 Conclusion and Directions for Future Investigations

Mangoes are grown in Thailand in large numbers most of which are for consump-
tion followed by processing after being consumed and processed. It produces waste
both in the household and the mango processing plant has to pay for the mango
seed removal as well. Food waste management is a fundamental challenge in the
global economy and contemporary business in the light of sustainable development
requirements. Generally, a spectrum of international, national, and local initiatives,
including legal acts and voluntary programs, has been developed to reduce food
waste. Similarly, different supply chain practices have advanced so far to decrease
food waste and enhance sustainability (Ocicka and Razniewska 2018). Therefore,
this research has selected mangoes to find value-added opportunities by processing
mangoes into crispy mangoes, which are becoming popular in the food snack market
among health and mango seed oil extraction which has many benefits and can be
extended to other industries as well. This paper has developed a financial feasibility
study of zero waste manufacturing: Opportunity and value added of agricultural prod-
ucts on the R9 route in accordance with analyzing the optimal location of the mango
processing plant by Center of Gravity Method and then verifying the results of the
data analysis to the analytic hierarchical process (AHP) and analysis of economic
data. The optimal location for the manufacturing plant is Phitsanulok Province and
it is feasible from the values of BEP, NPV, and IRR. Definitely suitable for investing
A Financial Feasibility Study of Zero Waste Manufacturing … 215

to build a mango processing plant. The new transport route promotes investments
into production sites for export at locations in the Northern provinces. It opens new
opportunities for exports of industrial goods and FMCG for the growing middle
class in China (Wagener et al. 2020). Making the sectoral economy grow strongly
and serves as a base for building economic growth of the country. In the National
Economic Development Plan No. 12 (Council 2017). Including Sustainable Devel-
opment Goal 9 build an infrastructure that is resilient to change, promote industrial
development comprehensively and sustainably and promote innovation. Sub-goal
9.2 promotes inclusive and sustainable industrial development and by 2030 increase
the industrial sector share in employment and gross domestic product in accordance
with the national context and to double the shares in the least developed countries.
With further research, the feasibility study could consider other target markets
apart from China and the cost of preferred stock to calculate the WACC and the other
methods like financial value added (FVA) and economic value added (EVA).

Acknowledgements The study is based on the National research and innovation information
system, Thailand. The authors appreciate Professor Dr.Rapeepan Pitakaso and this work is supported
by the program of the Bachelor of Business Administration in Business Management, FY 2021 under
Ubon Ratchathani University.

References

Azeem SW, Hussain H (2012) The determinants of foreign investment in Pakistan: a gravity model
analysis. Logforum 8(2):81–97
Babaria PM, Vashney AK (2012) Extraction of oil from mango kernel by hydraulic pressing. LAP
Lambert Academic Publishing
Bangkokbanksme (2015) 4 + 2 routes to send Thai fruits to China
Center ITC (2021) Information of mango production in the provincial located on route R9
Commerce MO (2021) tradereport. http://tradereport.moc.go.th/TradeThai.aspx
Council OOTNEASD (2017) 12th national development plan (2017–2021). Retrieved from https://
www.nesdc.go.th/ewt_dl_link.php?nid=6422
Czwajda L, Kosacka-Olejnik M, Kudelska I, Kostrzewski M, Sethanan K, Pitakaso R (2019) Appli-
cation of prediction markets phenomenon as decision support instrument in vehicle recycling
sector. Logforum 15(2):265–278. https://doi.org/10.17270/J.Log.2019.329
Golińska P (2014). The lean approach for improvement of the sustainability of a remanufacturing
process. Logforum, 10(3):235–293
Gustavsson J, Cederberg C, Sonesson U, Van Otterdijk R, Meybeck A (2011) Global food losses
and food waste. In: FAO Rome
Hirata T, Fukaya T (2020) Potential of truck platooning for transporting empty trucks considering
intercity freight demand imbalances. Logforum 16(3):373–383. https://doi.org/10.17270/J.Log.
2020.450
Joanicjusz Nazarko KAK, Czerewacz-Filipowicz K (2016) The new silk road—analysis of the
potential of new Eurasian transport corridors. In: Paper presented at the 9th international scientific
conference, business and management
Karunanithi B, Bogeshwaran M, Tripuraneni S, Reddy (2015) Extraction of mango seed oil from
mango kernel 32–41
216 S. Saisomboon et al.

Kauf S, Laskowska-Rutkowska A (2019) The location of an international logistics center in Poland

as a part of the one belt one road initiative. Logforum 15(1):71–83. https://doi.org/10.17270/J.
Log.2019.311
Kowalska A (2017) The issue of food losses and waste and its determinants. Logforum 13(1):7–18.
https://doi.org/10.17270/J.Log.2017.1.1
Krishnan R, Agarwal R, Bajada C, Arshinder K (2020) Redesigning a food supply chain for envi-
ronmental sustainability—an analysis of resource use and recovery. J Clean Prod 242:118374.
https://doi.org/10.1016/j.jclepro.2019.118374
Laokasemsukwong K (2013) The potential of dried fruit industry in thailand. Thummasat
Lord DMJ (2009) East-West Economic Corridor (EWEC) strategy and action plan. In: Munich
Personal RePEc Archive
Meteethunyarat T (2020) Oil tank farm site selection in Thailand by using the analytic hierarchy
process. Ladkrabang Eng J 32(3):37–42
News MO (2020) Mango farmers submit to the Administrative Court again pointing out that the
state-private groups Northern provinces to move budget to set up a steam plant affecting exports.
In: Manager Online News
Ocicka B, Razniewska M (2018) Food waste reduction as a challenge in supply chains management.
Logforum 14(4):549–561. https://doi.org/10.17270/J.Log.2018.303
Opoku RK, Abboah CKB, Owusu RT (2021) Inventory management strategies of food manufac-
turing industries in a developing economy. Logforum 17(1):37–48. https://doi.org/10.17270/J.
Log.2021.557
Patton D (2020) Food exporters to China asked to declare produce is coronavirus-free. Reuters
Phankong W, Manesri T, Manesri C (2020) Applicate analytic hierarchy process for evaluation of
biomass plant location. Ladkrabang Eng J 33(2):70–76
Popp J, Oláh J, Nestler S, Nobel T, Rakos M (2018). Development of dry ports in Europe. Int J
Appl Manag Sci 10(4):269–289
Rejeb A, Rejeb K, Keogh JG (2020) Covid-19 and the food chain? Impacts and future research
trends. Logforum 16(4):475–485. https://doi.org/10.17270/J.Log.2020.502
Saaty T, Vargas L (2001) Models, methods, concepts & applications of the analytic hierarchy process
Stratta L, Capozzi L, Franzino S, Pisano R (2020) Economic analysis of a freeze-drying cycle.
Processes 8:1399. https://doi.org/10.3390/pr8111399
Suwannachote T (2018) Businees plan for dnacher freeze dried fruit, Master Degree. Thammasat
university, Retrieved from http://ethesisarchive.library.tu.ac.th/thesis/2018/TU_2018_6002030
614_9851_9903.pdf
Wagener N, Aritua B, Zhu T (2020) The new silk road: opportunities for global supply chains and
challenges for further development. Logforum 16(2):193–207. https://doi.org/10.17270/J.Log.
2020.403
Works DOI (2016) Project to develop industrial areas with potential to support investment. Retrieved
from http://gisweb.diw.go.th/diwzonning/diwzoning62/images/content/5download/pdf2/8zonin
g59_Phitsanulok.pdf
The Impact of Tighter Regulations
of Export Controls in Global Logistics

Kuninori Suzuki , Yoji Murayama, and Yi Xing

1 Introduction

As the globalization of international logistics progresses, tariff barriers have lowered.

The introduction of information technology has advanced and export and import
procedures have simplified. Opportunities to export high-performance industrial
products, materials, and parts to developing countries, emerging nations, and “rogue
states” has increased, as these products, materials, and parts were not easily exported
in the past.
At the same time, various risks associated with exports have attracted attention.
The terror attacks against the United States and its allies, such as September 11,
2001, international disputes and civil wars have also occurred frequently, and Japan
has been constantly exposed to the risk of missile strike from North Korea. In this
international situation, the direct export of conventional weapons and weapons of
mass destruction has been restricted. Moreover, a sense of crisis has arisen about the
situation in which generally used in daily living are being converted into weapons in
rogue states.
From Japan’s viewpoint, this study attempts to confirm that the shipping processes
of international logistics bears a heavy burden from the series of export processes
under tightened export regulations.

K. Suzuki (B)
Nihon University, Tokyo, Japan
e-mail: [email protected]; [email protected]
Y. Murayama
Saitama Institute of Technology, Fukaya, Japan
e-mail: [email protected]
Y. Xing
Shibaura Mechatronics, Yokohama, Japan

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 217
P. Golinska-Dawson et al. (eds.), Smart and Sustainable Supply Chain
and Logistics—Challenges, Methods and Best Practices, EcoProduction,
https://doi.org/10.1007/978-3-031-15412-6_14
218 K. Suzuki et al.

2 Tightening Regulations on Export Controls

The terrorist attacks on September 11, 2001, led to the introduction of export controls.
Since the attacks, successive events have occurred in which the United States’ science
and technology is converted into military technology in Iraq, Iran, Libya, and North
Korea. Therefore, the movement for ensuring strict measures for export controls has
accelerated in the United States. In Japan, illegal exports have also occurred repeat-
edly. For example, a freeze dryer and a pulverizer that can have military applications
were exported without permission in November 2006, as was an unmanned helicopter
in 2007.
Moreover, from the end of 1980 until recently, textile manufacturer A and its
subsidiary have exported carbon fiber made by textile manufacturer A to Taiwan
for the purpose of “manufacturing sporting goods, such as tennis rackets.” However,
carbon fiber can have military uses such as in the manufacture of missiles. Therefore,
Japan’s Ministry of Economy, Trade and Industry (METI) issued a warning to textile
manufacturer A and its subsidiary for “false applications.”
The advancement of economic globalization and integration was followed by an
increase in the export of advanced technology in developed countries to developing
countries and emerging nations. Further, advanced technologies have frequently been
put to military use in rogue states. Consequently, strict restrictions have been imposed
on exports. To avoid the risk of nonmilitary products being put to military use,
catch-all controls were introduced in developed countries.
The Foreign Exchange and Foreign Trade Act provides that a person seeking to
export specific kinds of goods (hereinafter referred to as specific goods) that have been
found to compromise world peace and international security must first get permission
from METI to do so (hereinafter referred to as export permission) (Article 48). Before
a resident or a non-resident conducts a transaction with the objective of providing
technologies that have been found to compromise world peace and international
security, they must get the permission from METI (Article 25).

3 Previous Studies

Research on export control has increased in recent years and is not limited to the ban
on arms exports from the perspective of military or international politics such as the
United States. General-purpose products produced on the basis of US technology are
also regulated. Scholars examine Americans’ concern over the risk of military diver-
sion outside the United States. For example, Black and Holle describe and discuss
the export control policy perspectives shared by most of the US satellite industry.
Also, Nayan examines the salient features of the principal statutory authority. In
addition, Rudney and Anthony help us to understand the strengths and weaknesses
of the export control systems of the major supplier countries. A paper by Gill, Ebata,
and Stephenson pioneered Japan’s export control.
The Impact of Tighter Regulations of Export Controls in Global Logistics 219

However, Seyoum points out that export restrictions could be a response to inter-
national threats, with supply chain security in mind. However, export restrictions
have not kept pace with rapid changes in the global economy and technological
innovation, and thus concerns arose that export restrictions may adversely affect the
economic performance of global companies.

4 The List and Catch-All Controls

4.1 Outline

The list control regulates the export of weapons or technologically sophisticated

general-purpose products that can have military uses by listing the product names. To
export these goods, exporters must obtain permission from the Minister of Economy,
Trade, and Industry regardless of the stated purpose or who the end users are. The
catch-all controls regulate the export of all goods except foodstuffs and wood. There-
fore, when exporting a product to a country not on the “whitelist” and that product
corresponds to either “informed” or “objective” conditions then permission must first
be obtained from METI.

4.2 Classification

The classifications assessment process begins with an examination slip to determine

whether the engineer providing the technology has followed the list control. The
assessment often adopts a double-check system in which the engineering section
assesses technical aspects such as the specifications, while the export control section
confirms the technical aspects from a legal perspective. Using this double-check
system, the contents of invoices and parameter sheets are confirmed to ensure that
they are correct.
The classification is performed not only for items subject to the list control (related
to weapons of mass destruction and conventional weapons) but also for items that
may have military uses and may be used to develop weapons of mass destruction
and conventional weapons (items subject to catch-all controls). Items that fall under
the objective and informed conditions cannot be exported without permission from
METI. The objective conditions assess the end-use and end-user conditions based
on the exporter’s judgment. The informed conditions are applied when an exporter
obtains permission from METI.
220 K. Suzuki et al.

4.3 Transaction Screening Using the ST Control Form

The export control section assesses the end use and end user of a product using the
ST control form before deciding whether to grant approval. The process is explained
in greater detail below.
General transactions include the export of goods and technologies subject to
the list control, which do not correspond to products related to weapons of mass
destruction, restricted purchasers, and suspicious transactions.
Specially designated countries, such as Iran, Iraq, North Korea, Cuba, Syria,
and Sudan, are subject to the transactions of products related to weapons of mass
destruction. Inquiries from foreign militaries, national defense-related organizations,
or persons who have received orders of products from the above-mentioned militaries
or organizations or other items that have military uses are subject to the transactions
of products for military or military-related end use.
For restricted purchasers (in Japan and other countries), a master file of restricted
purchasers was created. Companies and persons who placed on the master file are
subject to the conditions of restricted purchasers.
Suspicious transactions are defined as: (1) the objective condition, which consists
of the objective condition for weapons of mass destruction (end-user and end-use
conditions) and the objective condition for conventional weapons (end-use condi-
tions), (2) the informed condition (a notice from METI), (3) the administrative guid-
ance condition (including cases in which a product is proven to be used for the
development of nuclear weapons), and (4) other suspicions can be cited.
Attachments for the ST control form include (1) transaction documents (inquiry
documents, order sheets, specifications, contracts, and minutes), (2) customer infor-
mation (company brochures, company yearbooks, and documents related to a
company’s homepage), (3) classification documents (classification sheets, the master
files of products and parts, the master file of instruments, and judgment lists), (4)
HRP check sheets, and (5) end-use confirmation letters are prepared.

4.4 Export Permission from METI

There are two export permission licenses, the individual validated license (IVL) and
the general bulk license (GBL).
The IVL grants permission to export goods and technologies based on the Foreign
Exchange and Foreign Trade Act. To export goods, export permits must be issued,
and to provide technologies, service transaction permits must be issued.
Regarding GBL, the export of goods and the provision of technologies in the
specified range to the specified regions are collectively permitted for up to three years.
When permission is obtained, no individual application for a license is required for
the export of goods and the provision of technologies within the permitted range.
The Impact of Tighter Regulations of Export Controls in Global Logistics 221

4.5 Catch-All Check Sheets

Regarding catch-all controls, the following are confirmed by creating the catch-all
check sheets: (1) whether the purchaser is restricted, (2) whether the products are
related to weapons of mass destruction, (3) whether end-user conditions apply, (4)
whether end-use conditions apply, (5) whether informed conditions apply, and (6)
whether the transactions are suspicious.
Regarding the classification of specific goods and technologies, transaction
screening is performed after confirming the transaction classification (general trans-
actions, suspicious transactions, transactions for specially designated countries,
transactions for military or military-related end use, and transactions for restricted
purchasers).
The roles of transaction-related sections are explained below. The sales section
confirms whether the transaction corresponds to the list of restricted purchasers,
products for military or military-related end-use, or specially designated countries. If
it does, an ST control form is issued. The sales section also confirms the classification
of inquired products.
The technology section classifies goods and technologies and controls for
providing technologies. For technologies subject to the list control, a management
ledger for technology provision histories is issued.
The export control section performs transaction screening, approves EMC trans-
actions, checks the ST control form, approves catch-all check sheets, checks export
licenses (export permission from the Minister of Economy, Trade and Industry), and
obtains export permission from the United States Secretary of Commerce.

5 Case: Exports of Personal Computers

Whether a product corresponds to specific goods that require export permission is

confirmed in both the list control shown in rows 1 to 15 of the appended Table 1 of
the Export Trade Control Order (hereinafter referred to as the Export Order) and the
catch-all controls shown in row 16 of the same table.

5.1 Classification of Goods in the List Control

Personal computers belong to the category of “electronic computers, electronic

assemblies or components therefor whose specifications comply with the Ordinance
of the Ministry of Economy, Trade and Industry” in row 8 of appended Table 1 of the
Export Order. The specifications are provided in Article 7 of the Ordinance of the
Ministry Specifying Goods and Technologies Pursuant to Provisions of appended
Table 1 of the Export Control Order and appended table of the Foreign Exchange
222 K. Suzuki et al.

Table 1 Minimum total time/Maximum total time

Assistant work Minimum total time (h) Maximum total time (h)
➀ Liaison and related work with 48 48
export control departments and
the ministry of economy, trade
and industry
➁ Liaison with list control/catch-all 48 48
control classification department
➂ Liaison with classification 32 32
division for specific technology
➃ Confirmation of the catch-all 48 48
check sheet
➄ Liaison with procurement/Import 20 20
department
➅ Liaison and related work with the 20 20
sales department
➆ Other miscellaneous affairs, 24 24
meetings, etc.

Order (hereinafter referred to as the Ministry Ordinance). When commercially avail-

able personal computers are collated with the specifications in Article 7 of the
Ministry Ordinance, those computers do not correspond to goods which require
export permission. Moreover, commercially available personal computers do not
correspond to the classifications of specified technologies.
However, wireless local area network (LAN) cryptographic equipment embedded
in personal computers may correspond to goods which require export permission.
Cryptographic equipment corresponds to “cryptographic equipment or electronic
assemblies, modules or integrated circuits having cryptographic functionality” in
row 9 of appended Table 1 of the Export Order. Cryptographic equipment embedded
in ordinary personal computers corresponds to the regulation of “those which use a
symmetric algorithm employing a key-length in excess of 56 bits.”
However, the cryptographic functionality of cryptographic equipment embedded
in a commercially available personal computer cannot be changed by its user. In
this case, export permission is not required because its use does not need technical
support by the supplier or sales agency.
Based on the above, the classification of personal computers is performed after
collating the specifications of personal computers with the specifications provided
in the Ministry Ordinance.
When a person temporarily leaves Japan with a personal computer equipped with
cryptographic equipment for their own use, export permission is not required. In
many cases, a certificate of non-applicability is not created for a company exporting
personal computers, but the certificate is acquired from the manufacturer. Therefore,
personal computers can be exported problem free. However, companies exporting
personal computers bear export responsibilities due to the Compliance Standards for
The Impact of Tighter Regulations of Export Controls in Global Logistics 223

Exporters and Persons Conducting Similar Transactions and are obliged to appoint
a person in charge of confirming the company’s non-applicability.

5.2 Applications for Export Permission and Catch-All

Controls

Personal computers are subject to regulations on “weapons of mass destruction

for catch-all controls” and “conventional weapons for catch-all controls.” When
the export of personal computers to countries and regions other than those on the
“whitelist” corresponds to the informed, objective, end-use, and end-user conditions,
applying to the Minister of Economy, Trade, and Industry for export permission is
required.
Personal computers often correspond to goods subject to US re-export controls
because they often contain software and embedded parts made in the United States.
Exporters of personal computers to countries restricted by the federal government of
the United States (Cuba, North Korea, Iran, Sudan, and Syria) must first get export
permission from the federal government of the United States.
Taking specified technologies that are regulated by the Foreign Exchange and
Foreign Trade Act out of Japan is subject to the Foreign Exchange and Foreign
Trade Act. Removing specified technologies stored in personal computers or USB
memories out of Japan for service transactions requires the permission from the
Minister of Economy, Trade, and Industry.

5.3 Regulations in the Destination Countries of Exports

The export of personal computers may be restricted due to the regulations of

importing countries. For example, the European Union has the following regulations:
the Electromagnetic Compatibility Directive, Low Voltage Directive, Restriction of
Hazardous Substances Directive, Waste Electrical and Electronic Equipment Direc-
tive, and Directive on Eco-Design of Energy-using Products. Exporters of personal
computers must address each and every regulation.

6 Systems to Assist Shipping and Export Controls

As mentioned above, shipping operations may become a heavy burden for the
export section of a manufacturing company due to the strengthened list and catch-all
controls. When exports increase due to the relaxation of regulations at the macro-
economic level, such as the expansion of regional economic integration, shipping
224 K. Suzuki et al.

operations become more complicated. This study proposes a cross-sectional organi-

zational system to supervise the entire international operation instead of a vertical
sectioning system. To this end, an international logistics management supervisory
group or section (a tentative name) should be established to assist the highly special-
ized export control section and to provide a bridge connecting export control, sales,
import, and procurement sections, because these sections need to strengthen in-house
collaboration.

6.1 Conceptual Organizational Model

According to Chester Barnard, organizational goals, willingness to serve, and infor-

mation sharing must exist in an organization in a well-balanced manner. An interna-
tional logistics management supervisory group (hereinafter referred to as the super-
visory group), which is cross-sectionally constructed, will sufficiently satisfy the
above-mentioned three elements of an organization proposed by Barnard.
Organizational goal
Due to the expansion of regional economic zones, customs clearance has been simpli-
fied. The supervisory group appropriately copes with complicated shipping matters,
the amount of which has rapidly increased, and applications for export permissions.
The supervisory group closely collaborates with the sales, procurement, and import
sections, and it also has a monitoring function.
Willingness to serve
The supervisory group will have to acquire sufficient practical knowledge to cope
with the list control, the catch-all controls, and the Act against Delay in Payment
of Subcontract Proceeds, and so forth, to subcontractors. Therefore, the supervisory
group’s contribution to corporate activities is high. The supervisory group’s will-
ingness to serve is highly stimulated by always being involved in various important
procedures that are related to export restrictions.
Information sharing
By providing a bridge connecting export control, procurement, import, and sales
sections, the supervisory group plays a key role in information sharing.

6.2 Organizational Reform Related to Foreign Trade

Company B has been established for more than 100 years. With the introduction of a
new information system, this company undertook significant organizational reform
related to export controls, including the list and catch-all controls. Major problems
encountered during its organizational reform were: (1) complicated examination
The Impact of Tighter Regulations of Export Controls in Global Logistics 225

routes, (2) wasted time due to the distribution of jobs, and (3) operational errors and
difficulties in understanding and sharing information in a timely manner.
Company B has the three divisions: medical services, science, and visual content.
These divisions handle products subject to the list and catch-all controls. The route
to approving a project was long, many business activities were redundant, and the
database was not shared among divisions. To solve these three problems, the organiza-
tion was reformed, and a new information system was introduced. As a result, exam-
ination times were shortened by 25–33%, personnel assignments were efficiently
performed, the business affairs of each division were visualized, and information
was properly collected.
The introduction of a new information system is essential for organizational
reform. Many of company C’s products are subject to the list control due to their
characteristics. Therefore, company C constructed a system that can automatically
apply the results of past transaction screenings to a new transaction and connected
the function of summarizing the classification results of similar products to the SCM
support system. Through this system, company C performs process management
from order acceptance to shipping.
However, many manufacturing companies cannot introduce expensive systems. It
may therefore be more realistic to improve business efficiency through organizational
reform before introducing new information systems.

6.3 Establishment of a Cross-Sectional Organization

The person in charge of business related to the list and catch-all controls may cover
the export control section, the classification section of the list and catch-all controls,
the classification section of specified technologies, the confirmation section of catch-
all check sheets, the sales section, the procurement section, and the import section.
Because it is difficult for that person to concentrate on the businesses of existing
sections, the establishment of a cross-sectional organization is required.
The amount of export-related business depends on the handling volume of goods
to be exported. In the case of companies listed on the first section of the Tokyo Stock
Exchange whose annual turnover is between 50 and 100 billion yen, the following
organization is proposed as a standard prototype:
(1) Number of staff: six.
(2) Shifts: Staff members assist the section that is responsible for the particular day,
in the form of half-day units, such as morning and afternoon shifts.
(3) Assistive duties consist of the following (a)–(f):
(a) Communication with the export control section and METI, and
communication-related duties
(b) Communication with the classification section of the list and catch-all
controls
(c) Communication with the classification section of specified technologies
226 K. Suzuki et al.

(d) Confirmation of catch-all check sheets

(e) Communication with the procurement and import sections
(f) Communication with the sales section, and communication-related duties.
In addition to the above six duties, staff members would perform miscellaneous
duties and preliminaries. If necessary, work shifts would be arranged every week
based on the requests of each staff member responsible.

6.4 Construction of a Scheduling Management System

Staff members of the above-mentioned organization would schedule work shifts

every week and take charge of all the duties described in (a)–(f). To reduce the burden
on each staff member, scheduling management would be performed. To create a work
shift table, the following conditions are considered:
(1) Constraint conditions for shifts.
Conditions related to the composition of staff members (satisfaction with
working status).
(a) Keep the minimum and maximum total working hours of each work shift
(assistive duties) within the target period.
(b) Keep the minimum and maximum number of staff members for each day,
time zone, and work shift (assistive duties).
(2) Constraint conditions for the staff members responsible.
Conditions related to the workload and work restrictions of each staff member.
(c) Keep the number of duties of each staff member in each work shift (assistive
duties) in the target period within the range of the minimum and maximum
numbers of duties.
(d) Consider each staff member’s desired duties in charge at the day and time
zone levels.

6.5 Formulation

On the premise of the constraint conditions in the previous section, the formulation
is performed as follows:
Formula (1) is an objective function, which minimizes the sum of penalties against
the maximum total working hours of each work shift, with the constraint conditions
of (a) above.
{{{
Minimize ys (1)
d∈D n∈N s∈S
The Impact of Tighter Regulations of Export Controls in Global Logistics 227

where D represents the aggregate of days in the target period, N represents the
aggregate of staff members responsible, S represents the aggregate of work shifts
(assistive duties), variable ys represents penalties against the maximum total working
hours of each work shift with the constraint conditions of (a) above. In other words,
a penalty is imposed on the number of hours that exceed the maximum total working
hours in work shift s.Inaddition, ys is defined using Formula (11).
Formula (2) expresses that each staff member is assigned to shift s when the
member works in each time zone h on each day d.
{
xdhns = 1, d ∈ D, h ∈ H, n ∈ N (2)
s∈S

where H represents the aggregate of time zones, and variable xdhns is a 0/1 variable,
which is 1 when a duty assigned to each responsible staff member n in each time
zone h on each day d is s and that is 0 in other cases. In addition, xdhns is defined
using Formula (10).
Formulae (3) and (4) express constraint conditions of (a) above and indicate that
the total working hours of each shift s are within the range between the minimum
total working hours c1 and the maximum total working hours c2. Formula (5) allows
a breach (penalty), and variable ys represents the deviation amount (the number of
hours exceeding the maximum total working hours).
{{ {
xdhns ≥ c1s , s ∈ S (3)
d∈D h∈H n∈N
{{ {
xdhns ≤ c2s + ys , s ∈ S (4)
d∈D h∈H n∈N

Regarding the constraint conditions of (b) above, Formulae (5) and (6) restrict the
number of staff members required for shift s in each time zone h on each day d by
the minimum number of people c3s and the maximum number of people c4s .
{
xdhns ≥ c3s , d ∈ D, h ∈ H, s ∈ S (5)
n∈N
{
xdhns ≤ c4s , d ∈ D, h ∈ H, s ∈ S (6)
n∈N

Regarding the constraint condition of (c) above, Formulae (7) and (8) restrict the
number of duties of each staff member in shift s in the target period by the minimum
and maximum numbers of duties (c5ns and c6ns , respectively).
{{
xdhns ≥ c5ns , n ∈ N , s ∈ S (7)
d∈D h∈H
228 K. Suzuki et al.
{{
xdhns ≤ c6ns , n ∈ N , s ∈ S (8)
d∈D h∈H

Formula (9) expresses the constraint condition of (d) above and restricts shift s of
each staff member n in the time zone h on the specific day d using the desired duty
L + dhns .

L + dhns − xdhns ≤ 0, d ∈ D, h ∈ H, n ∈ N , s ∈ S (9)

Variable xdhnb is a 0/1 variable, which is 1 when a duty assigned to each staff
member responsible n in each time zone h on each day d is s and which is 0 in other
cases. This value is defined using Formula (10).

xdhns ∈ {0, 1}, d ∈ D, h ∈ H, n ∈ N , s ∈ S (10)

Variable ys expresses the deviation amount (the number of hours exceeding the
maximum total working hours of each shift s), and is defined using Formula (11).

ys s ∈ S (11)

In these formulae, all of c1–c6 are constants.

c1: Minimum total working hours of each shift in the target period.
c2: Maximum total working hours of each shift in the target period.
c3: Minimum number of people in each shift in each time zone on each day.
c4: Maximum number of people in each shift in each time zone on each day.
c5: Minimum number of duties of each staff member in each shift in the target
period.
c6: Maximum number of duties of each staff member in each shift in the target
period.

6.6 Numerical Experiment

The target period was set to five days and the time zone was divided into two. These
were between 9 a.m. and 1 p.m. (four hours including a break) and between 2 p.m.
and 6 p.m. (four hours). The minimum and maximum total working hours in each
shift were set, as shown in Table 1. The required number of people were set, as
shown in Table 2. The number of duties of each staff member in each shift in the
target period were set, as shown in Table 3, and values shown in this table were
applied to all staff members. Each staff member’s desired duty was set, as shown in
Table 4. Table 5 shows the results obtained in the numerical experiments. As shown
in Table 5, a solution that satisfied all the constraint conditions could be obtained.
The sum of penalties was 0.
The Impact of Tighter Regulations of Export Controls in Global Logistics 229

Table 2 Required number of people

Assistant work Minimum number of people Maximum number of people
➀ Liaison and related work with 0 2
export control departments
and the ministry of economy,
trade and industry
➁ Liaison with list 0 2
control/catch-all control
classification department
➂ Liaison with classification 0 2
division for specific
technology
➃ Confirmation of the catch-all 0 2
check sheet
➄ Liaison with 0 1
procurement/Import
department
➅ Liaison and related work with 0 1
the sales department
➆ Other miscellaneous affairs, 0 1
meetings, etc.

Table 3 Minimum times and Maximum times

Assistant work Minimum times Maximum times
➀ Liaison and related work with export control 2 2
departments and the ministry of economy, trade and
industry
➁ Liaison with list control/catch-all control 2 2
classification department
➂ Liaison with classification division for specific 1 2
technology
➃ Confirmation of the catch-all check sheet 2 2
➄ Liaison with procurement/Import department 0 1
➅ Liaison and related work with the sales department 0 1
➆ Other miscellaneous affairs, meetings, etc. 1 1

Table 4 Desired working

Staff no. Desired work
1 Day 1 AM assistant ➁
4 Day 4 assistant PM ➄
5 Day 3 AM assistant ➃
230 K. Suzuki et al.

Table 5 Work shift

1 2 3 4 5
Sun Mon Tue Wed Thu Fri Sat
Staff AM PM AM PM AM PM AM PM AM PM AM PM AM PM
1 2 5 1 2 1 3 6 4 4 7
2 7 1 4 2 2 3 1 4 6 5
3 6 2 1 1 3 2 7 3 4 4
4 1 4 6 7 4 1 3 5 2 2
5 4 1 7 3 4 2 1 2 5 6
6 2 2 4 1 5 4 3 1 7 3

*Gray shading indicates desired work

6.7 Discussion

To improve the efficiency of shipping and export operations, a cross-sectional super-

visory organization must be established, and its operations must be assisted so that
no heavy workload is placed on individual staff members. To this end, the scheduling
management of the staff members responsible for any activities is indispensable. It
is desirable to link a scheduling management system with a search system, which
can automatically apply the results of transaction screening in the past to new trans-
actions and the SCM support system. With this system, process management from
order acceptance to shipping can be performed.

7 Conclusion

In this paper, we have considered, in detail, a series of export control processes. It is

undeniable that the time and burden required to establish whether to determine the
specific cargo/technology and to confirm it with the catch-all check sheet is enormous,
and this burden is a new trade barrier that replaces customs procedures such as
tariffs. The strengthening of global export control has been promoted, while each
country is greatly influenced by the United States; however, ASEAN countries lack a
sufficient legal system to comply with list regulations, catch-all regulations, and the
like. Therefore, many countries do not have a legal system for exports management.
It has been pointed out that although legislation is being developed, for example in
China and India, judicial issues remain due to security restrictions. Although reducing
the number of procedures is difficult, the smoothness of shipping operations will be
hindered by the effects of new barriers created by the globalization of the economy.
Prior to 1993, Japan invited export managers from ASEAN countries to hold the
“Asia Export Control Seminar” every year to deepen the understanding of ASEAN
The Impact of Tighter Regulations of Export Controls in Global Logistics 231

countries about the importance of export control. The seminar hosted by the Secu-
rity and Trade Center (CISTEC) is attended by Japan’s METI, Ministry of Foreign
Affairs, leading companies, among others. For practitioners of shipping and export
control, discussions with government officials tend toward institutional theory and
ideal theory but still reflect the voices of businesspeople, including knowledge about
business.
As described above, it has become clear that while customs operations are
being reduced due to the expansion of the economic zone, export-related shipping
operations are becoming more complicated due to regulations.
In addition, to promote the efficiency of shipping and export operations by showing
a model in this paper and simulation, a cross-sectional control department was
newly established to assist operations without imposing a heavy labor load on the
newly established organization. For that purpose, schedule management of the person
in charge is indispensable. Furthermore, linking the schedule management system
with a search system and SCM support system that can automatically apply past
transaction examination results from ordering to shipping is critical.

Bibliography

Beck M (2000) Reforming the multilateral export control regimes. Nonprolif Rev 7(2):91–103
Black S, Holle M (2015) Satellite industry perspectives on export control regulations. Astropolitics
Int J Space Politics Policy 13(2–3)
CISTEC (2018) Export control compliance and American academia. Yushutukanri No Johoshi
173:94–99
Gill B, Ebata K, Stephenson M (1996) Japan’s export control initiatives: meeting new nonprolifer-
ation challenges. Nonprolif Rev 4(1)
Holm K (2006) Europeanising export controls: the impact of the European union code of conduct
on arms exports in Belgium, Germany and Italy. Eur Secur 15(2):213–234
Mathias C (1982) A new U.S. export control policy. Wash Q
Nayan R (2005) Emerging frontiers of US dual-use export control laws. Strateg Anal 29(1)
Noble MJ (2008) Export controls and United States space power. Int J Space Politics Policy 6(3)
Peterson D, Goorevich RS, Hooper R, Scheinman L, Tape JW (2008) Export controls and
international safeguards. Nonprolif Rev 15(3):515–527
Rajeswari PR (1998) US export control policy and Wassenaar arrangement. Strateg Anal 22(3)
Rajeswari PR (1999) Economics of export controls: a study of US export control mechanism. Strateg
Anal 23(7):1199–1213
Roberts B (1998) Export controls and biological weapons: new roles. New Chall Crit Rev Microbiol
24(3):235–254
Rudney R, Anthony TJ (1996) Beyond CoCom: a comparative study of five national export control
systems and their implications for a multilateral nonproliferation regime. Comp Strateg Future
Russia 15(1):41–57
Sato H (2019) Export control and international system. J World Aff 67(4):53–67
Sato T (2020) Issues based on the historical backgrounds and developments in export law and
policies before introducing the U.S.export control reform act (ECRA) about check and balance
by separation of powers and dual uses etc. Aoyama Law Rev 61(4):403–419
Seyoum B (2017) Export controls and international business: a study with special emphasis on
dual-use export controls and their impact on firms in the US. J Econ Issues 51(2017):45–72
232 K. Suzuki et al.

Suzuki K, Murayama Y (2021) A study on hybrid solving of scheduling problems by swarm

intelligence based on business trends, report A of the research institute of industrial technology,
Nihon University Number 54, vol 1, pp 1–7
Suzuki K, Murayama Y, Wakabayashi K, Watanabe A (2019) Efficiency of crew assignment in
truck freight operation from the view of logistics, environmental sustainability in Asian logistics
and supply chains, international federation of logistics and SCM systems. Springer, pp 219–231
Xing Y, Suzuki K, Murayama Y, Wakabayashi K, Watanabe A (2021) A study on a series of processes
in export control. J Jpn Soc Direct 19(1):44–51
Two Echelon Vehicle Routing Problem
for Disaster Mapping Process Using
Cooperated Vehicles Assignment

A. A. N. Perwira Redi , Nandini N. Sekaringtyas, Handina B. Astiana,

Rahmat Inca Liperda , Anna Maria Sri Asih , and Bertha Maya Sopha

1 Introduction

A natural disaster is a phenomenon that is caused by natural forces and potentially

leads to significant damages and loss of life (WHO 2002). According to Djalante
et al. (2017), Indonesia is one of the countries with the highest risk of natural disas-
ters globally, and one of the most common disasters that occur in this country is
floods (Djalante et al. 2017). This disaster has a significant impact on Indonesia,
especially on Bekasi City at the beginning of 2020. The damages caused by floods
affected some infrastructures such as residences, roads, bridges, medical facilities,
and others, resulting in potential obstacles in distributing disaster relief. Therefore,
the availability of accurate maps to provide up-to-date information on the affected
area condition is needed so that the disaster aid can be distributed quickly and
precisely (Sarma et al. 2020).
Mapping disaster-affected areas are one of the disaster management activities
(Das et al. 2013). Disaster management is a series of activities that aim to prevent,

A. A. N. Perwira Redi (B)

Department of Industrial Engineering Department, BINUS Graduate Program – Master of
Industrial Engineering, Jakarta, Indonesia
e-mail: [email protected]
N. N. Sekaringtyas · H. B. Astiana · R. I. Liperda
Department of Logistics Engineering, Universitas Pertamina, Jakarta, Indonesia
e-mail: [email protected]
A. M. S. Asih · B. M. Sopha
Department of Mechanical and Industrial Engineering, Universitas Gadjah Mada (UGM),
Yogyakarta, Indonesia
e-mail: [email protected]
B. M. Sopha
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 233
P. Golinska-Dawson et al. (eds.), Smart and Sustainable Supply Chain
and Logistics—Challenges, Methods and Best Practices, EcoProduction,
https://doi.org/10.1007/978-3-031-15412-6_15
234 A. A. N. Perwira Redi et al.

mitigate, prepare, emergency response to disasters and recover the affected areas.
However, practically mapping or recording disaster-affected areas in Indonesia is
relatively time-consuming. This is due to the limited number of mapping methods
providing accessible routes to reach affected areas. In practice, mapping can be done
by using satellite imagery and drone. Mapping using satellite imagery enables one to
record large areas because of its ability to record from thousands of kilometres from
the ground. However, using this tool costs a lot and requires a long time for image
processing, so this method is rarely used for disaster management purposes (Kerle
and Oppenheimer 2002). This method is also inflexible since the image resolution
is strongly affected by weather conditions, and it is difficult to record in the exact
location (Zulkipli and Tahar 2018). Meanwhile, mapping with drones has advantages
such as having high speed and accessing areas without being affected by road infras-
tructure conditions (Kitjacharoenchai et al. 2020). Furthermore, drones can generate
actual data using sensors such as the Global Positioning System (GPS), Inertial
Navigation System (INS), and infrared (Zulkipli and Tahar 2018). Disaster mapping
using drones also increases logistics activities efficiency and reduces disaster risk,
early warning, and flood monitoring (for the Coordination of Humanitarian Affairs
(OCHA 2014).
Despite its advantages in disaster mapping operation, drones have a limited range
in capturing affected areas. Therefore, a suitable mechanism and planning are needed
to execute mapping with drones to reach all affected areas. Based on the previous
research, using a combination between ground vehicles and drones could extend
the mapping range (Marinelli et al. 2018). However, practically using this combi-
nation for disaster-affected areas mapping is rarely found, especially in Indonesia.
Therefore, this topic becomes the focus to discuss.
This research will focus on using a combination of ground vehicles and drones
to record the flood disaster-affected area in Bekasi, January 2020. The problem
will be modelled into 2E-VRP (Two-Echelon Vehicle Routing Problem), which is a
simplification of Multi-Echelon VRP (Crainic et al. 2010). Since it is still categorized
as a VRP model, this model will determine the route of vehicles with the most
minimum possible time. However, this model is able to cover a problem with two
different types of vehicles in different echelons. The first echelon determines the
route for ground vehicles carrying drones, while the second echelon determines the
route for drones to record the areas. A study has discussed the application of the 2E-
VRP model with the drone that moves in a round trip called 2E-VRP-MOD (Liperda
et al. 2020). This study modifies the drone movement into cooperative movement
with the same objective: record the entire affected area with the minimum time. This
is based on previous research that says that a combination of ground vehicles and
drones that work cooperatively finish the tasks more efficiently and at lower costs.
Based on the description mentioned, this study aims to build a mathematical model
for mapping the affected areas with a combination of ground vehicles and drones
that move cooperatively.
Two Echelon Vehicle Routing Problem for Disaster Mapping … 235

2 Problem Description and Mathematical Model

This research problem is formulated in a mathematical model, namely 2ECOVRP-

MOD (Two Echelon Cooperated Vehicle Routing Problem—Mapping Operation
with Drone), which describes disaster mapping operation with two types of routes
(echelons) and these routes are accessed by ground vehicles and drones. The ground
vehicle used in this research motorcycle due to its flexibility to reach the affected
areas. A motorcycle carries one drone. The operation starts in the first echelon, from
the depot point where these motorcycles take the drones to each stopover point.
These stopover points are also where the drones will be launched to the affected
areas to map the second echelon. The drones will return to stopover points, and they
will be carried back to the depot by motorcycles after capturing the images of all
affected areas. This model is modified from prior research with a similar case and
is called 2E-VRP-MOD (Liperda et al. 2020). This model allows a drone to return
to the stopover point once it has captured an image of a mapping point. However,
in this research, the drones move cooperatively where they are able to visit more
than one affected area as long as the flight time does not exceed the battery capacity.
Cooperative movements between ground vehicles and drones are more efficient with
lower costs compared to when these vehicles do not move cooperatively (Murray
and Chu 2015). The illustration of the model can be seen in Fig. 1.

Fig. 1 Illustration of 2ECOVRP-MOD

236 A. A. N. Perwira Redi et al.

Model 2ECOVRP-MOD has an objective to minimize the total time required to

record all the affected areas. The details of the 2ECOVRP-MOD model is define as
follows.
Set:
S Set of all nodes in the first echelon (includes depot)
S0 Set of all stopover points where S0 = {1 . . . |S|}
C Set of all mapping points
V Set of all vehicles in the first echelon
K Amount of the drone routes in the second echelon
Parameter:
Cs i j Travel time of vehicle from gathering point or stopover point i to stopover
point j includes depot) in the first echelon
Cci j Travel time of drone from stopover point i to mapping point j in the second
echelon
Di Required mapping time for capturing the area of a mapping point i
Qd v Maximum flight time of a drone
L A sufficiently large number
Decision variables:
xi jv A binary variable equals one if vehicle v is travelled from node i to j where
edge (i, j ) ∈ E1 Otherwise zero.
zi A binary variable equals one if the stopover point i is visited where i ∈ S
Otherwise zero.
yi jk A binary variable that equals to one if drone routes k is travelled from node
i to j where edge (i, j ) ∈ E2 Otherwise zero.
Qj The total amount of time needed for taking pictures in a stopover point i
where i ∈ S
Q K jk The total amount of time needed for taking pictures in a stopover point i at
each route k where i ∈ S
T jk The total amount of time needed for travelling in a stopover point i where
i∈S
Uj The total amount of time needed for drone travelling in E2 where j ∈ S0
aiv The total accumulated amount of information gathering time has been used
up to stopover point i using vehicle v where i ∈ S and v ∈ V
biv The total accumulated amount of transportation time has been used up to
mapping point i using vehicle v where i ∈ C and v ∈ V
c jk The total accumulated amount of mapping time has been used up in a stopover
point j at routes k where j ∈ S0 ∪ C and k ∈ K
Objective
   
Min Z = Cs i j xi jv + Cci j yi jk
i∈S j∈S v∈V i∈S0 ∪C j∈S0 ∪C k∈K
Two Echelon Vehicle Routing Problem for Disaster Mapping … 237
 
+ Di yi jk (1)
j∈S0 ∪C i∈C k∈K

Subject to

xi jv = z j ∀ j ∈ S0 (2)
i∈S v∈V

xi jv = z i ∀i ∈ S0 (3)
j∈S v∈V

  ∀l ∈ S0
xilv − xl jv = 0 (4)
∀v ∈ V
i∈S j∈S

x1 jv ≤ 1 ∀v ∈ V (5)
j∈S0

xi1v ≤ 1 ∀v ∈ V (6)
i∈S0
 
yi jk = 1 ∀j ∈ C (7)
j∈S0 ∪C k∈K

 ∀i ∈ S0 ∪ C
yi jk ≤ 1 (8)
∀k ∈ K
j∈S0 ∪C

  ∀i ∈ S0 ∪ C
y jik = yi jk (9)
∀k ∈ K
j∈S0 ∪C j∈S0 ∪C

  ∀l ∈ C
yilk − yl jk = 0 (10)
∀k ∈ K
i∈S0 ∪C j∈S0 ∪C

∀ j ∈ S0
Q K jk ≥ cik · yi jk ∀i ∈ C (11)
∀k ∈ K

Qj ≥ Q K jk ∀ j ∈ S0 (12)
k∈K
 
Qj = Di (13)
j∈S0 i∈C

Qj ≤ L · zj ∀ j ∈ S0 (14)
238 A. A. N. Perwira Redi et al.

∀i ∈ S0
bik = 0 (15)
∀k ∈ K

∀ j ∈ S0
c jk = 0 (16)
∀k ∈ K

∀j ∈ C
c jk ≥ 0 (17)
∀k ∈ K

∀ j ∈ S0
T jk ≥ bik + Cci j ∀i ∈ C (18)
∀k ∈ K

Uj ≥ T jk z j ∀ j ∈ S0 (19)
k∈K

  ∀i ∈ S
aiv + Q j − L 1 − xi jv ≤ a jv ∀ j ∈ S0 (20)
∀v ∈ V

  ∀i ∈ S0 ∪ C
bik + Cci j − L 1 − yi jk ≤ b jk ∀j ∈ C (21)
∀k ∈ K

  ∀i ∈ S0 ∪ C
cik + D j − L 1 − yi jk ≤ c jk ∀j ∈ C (22)
∀k ∈ K
  
xi jv Q j + U j ≤ Qdv ∀v ∈ V (23)
i∈S j∈S0

The objective function (1) is the sum of total travel time in the first echelon,
which is a motorcycle, total travel time in the second echelon, and the total duration
time of recording Constraints (2) and (3) restrict the ground vehicles to visit the
stopover points more than once on the first echelon. Constraint (4) aims to ensure
the consecutive movement of vehicles on the first echelon. Constraints (5) and (6)
ensure the ground vehicle to start and return the depot. Constraint (7) ensures that
all mapping points are visited, while constraint (8) limits any mapping point to be
visited only once. Constraint (9) validates the vehicle routes on the second echelon.
Constraint (10) ensures the consecutive movement of the vehicles on the second
echelon. Constraint (11) denotes the duration on the route k if the mapping point
is visited. Constraint (12) denotes the accumulated mapping time of each mapping
point at the route k. Constraint (13) signifies that the accumulated total recording
time on each stopover point equals the accumulated total required to record for that
area. Constraint (14) calculates each mapping point is accumulated recording time
at the stopover point and indicates that the stopover point has unlimited capacity.
Two Echelon Vehicle Routing Problem for Disaster Mapping … 239

Constraint (15) ensures that the calculation of drones’ total travelling time starts
from 0. Constraint (16) ensures that the mapping duration of drones from 0 and
constraint (17) shows that the operation time of mapping points of each route k
can be equal to more than 0. Constraint (18) denotes the accumulation of drones’
travelling time and constraint (19) calculates the total flight time of all drones from
the stopover points. Constraints (20) and (21) are the sub-tour elimination of the
first and second echelon, respectively. Constraint (22) signifies the validation of the
mapping process. Constraint (23) limits the drones’ mapping operations and flight
time so it do not exceed the battery capacity.

3 Computational Study

The areas of Bekasi City which was affected by floods in January 2020 are the
object research, and these are illustrated using Google Maps by marking the points
of a depot (black icon), stopover points (blue icons), and the affected areas (red and
orange icons). The coordinates of these places on the Google Maps are used in QGIS
as inputs to measure the distances between points and then converted into travelling
time with the speed of motorcycles equals 20 km/h. The illustration of the affected
areas can be seen in Fig. 2.
This also applies to travelling time calculation in the second echelon. The drone
used in this research is DJI Phantom 4 Pro V2.0, with a constant speed of 5.56 m/s
and a maximum flight capacity of 30 min. Two batteries are used for a drone so that
the maximum flight time is 60 min without considering the time needed to charge
the battery. Secondary data of the affected locations were obtained from the regional
disaster management agency of Bekasi City Regional Disaster Management Agency
(BPBD). It is also needed to know the duration of mapping on each affected point,
and this duration depends on the area. The larger the area, the more time is needed
for the drone to do the mapping process. The areas of these affected places were
then used to calculate the mapping duration. This duration calculation was also done
by referring to research conducted by Boccardo et al. (2015) in mapping a 400 m
× 400 m area, and this was done within 13 min at the altitude of 70 m. In this
study, the mapping duration is assumed to be 8125 × 10–5 min/m2 when the drone
is 70 m above the ground. After calculating the travelling time on the first echelon,
the drones’ flight times on the second echelon and the mapping durations, these data
then were categorized into eight to ten small instances. The model is verified using
AMPL with the computer specifications as an Intel® Core ™ i7-7500U processor
with 8 GB RAM and a 64-bit operating system.
240 A. A. N. Perwira Redi et al.

Fig. 2 Flood affected areas in Bekasi City, January 2020

4 Discussion

The computational results can be seen in Table 1. The results shown a feasible result
with all drones utilize to visits all mapping points. The results show that the problem
can be solved using 2ECOVRP-MOD with the objective value of mapping time from
19 to 48 min. The solution can be obtained with a computational time from 0.25 to
1.2 s. This result is considerably efficient in terms of obtaining the optimal solution.
Furthermore, the detailed route can be seen in Table 2. However, due to the small
size of the instance being used, there is a need to further investigate in a larger size
instance.

5 Conclusion

This study proposes an integer linear programming model for ground vehicles and
drones to conduct mapping operations in the post-disaster situation denoted as
2ECOVRP-MOD (Two Echelon Cooperated Vehicle Routing Problem—Mapping
Two Echelon Vehicle Routing Problem for Disaster Mapping … 241

Table 1 The computational results

Data Stopover point Mapping point Objective value (min) Solve time (s)
Small 1 2 7 19,851 0.20313
Small 2 3 8 35,167 0.29688
Small 3 3 8 30,32 0.25
Small 4 3 8 39,074 1.01562
Small 5 3 10 29,371 1.1875
Small 6 2 7 25,782 0.25
Small 7 3 8 31,136 0.35938
Small 8 2 9 36,457 0.34375
Small 9 2 10 38,632 0.29688
Small 10 2 6 23,23 0.15625
Small 11 2 7 22,797 0.328125
Small 12 2 7 33,606 0.140625
Small 13 3 10 46,964 0.703125
Small 14 2 10 40,637 0.546875
Small 15 3 10 48,22 0.5625
Small 16 2 8 19,872 0.34375
Small 17 3 8 25,336 0.25

Table 2 The detailed routes

Data 1st echelon routes 2nd echelon routes Total routes
Small 1 1–3, 3–1 3–6-7–5-4–3 1–3-6–7-5–4-3–1
Small 2 1–4, 4–1 4–7-8–5-6–4 1–4-7–8-5–6-4–1
Small 3 1–4, 4–1 4–6-5–8-7–4 1–4-6–5-8–7-4–1
Small 4 1–4, 4–1 4–8-6–7-5–4 1–4-8–6-7–5-4–1
Small 5 1–2, 2–1 2–10-8–9-5–7-6–2 1–2-10–8-9–5-7–6-2–1
Small 6 1–3, 3–1 3–6-5–4-7–3 1–3-6–5-4–7-3–1
Small 7 1–3, 3–1 3–6-8–4-5–7-3 1–3-6–8-4–5-7–3-1
Small 8 1–3, 3–1 3–4-9–7-8–6-5–3 1–3-4–9-7–8-6–5-3–1
Small 9 1–2, 2–1 2–7-8–9-10–4-6–5-2 1–2-7–8-9–10-4–6-5–2-1
Small 10 1–2, 2–1 2–5-6–4-2 1–2-5–6-4–2-1
Small 11 1–2, 2–1 2–4-5–6-7–2 1–2-4–5-6–7-2–1
Small 12 1–2, 2–1 2–7-5–6-4–2 1–2-7–5-6–4-2–1
Small 13 1–2, 2–3, 3–1 2–6-7–8-9–5-2, 3–10-3 1–2-6–7-8–9-5–2-3–10-3–1
Small 14 1–2, 2–1 2–8-9–7-10–6-5–4-2 1–2-8–9-7–10-6–5-4–2-1
Small 15 1–3, 3–1 3–5-10–9-8–7-6–3 1–3-5–10-9–8-7–6-3–1
Small 16 1–3, 3–1 3–8-7–6-4–3 1–3-8–7-6–4-3–1
Small 17 1–2, 2–1 2–7-8–5-4–6-2 1–2-7–8-5–4-6–2-1
242 A. A. N. Perwira Redi et al.

Operation with Drone). The model is applied to solving a mapping operation in a real
case of flood disaster mapping in Bekasi, Indonesia. It is shown that the model can be
effectively used to provide an optimal solution for small size data. However, further
investigation is needed to verify the model’s ability to solve a larger data size. Further
research needs to investigate the possibility of using cruise speed optimization to get
the best result among different cruise speeds. Moreover, the application of heuristics
or meta-heuristics approaches enables finding a solution for more significant size
instances of as 2ECOVRP-MOD is needed.

Acknowledgements This research is funded by the Indonesian Ministry of Education and Culture
(Kementrian Pendidikan dan Kebudayaan) under the World Class Research (WCR) Scheme.

References

Boccardo P, Chiabrando F, Dutto F, Tonolo F, Lingua A (2015) UAV deployment exercise for
mapping purposes: evaluation of emergency response applications. Sensors 15(7):15717–15737.
https://doi.org/10.3390/s150715717
Crainic TG, Perboli G, Mancini S, Tadei R (2010) Two-echelon vehicle routing problem: a satellite
location analysis. Proc Soc Behav Sci 2(3):5944–5955
Das K, Lashkari RS, Biswas N (2013) Disaster assessment and mitigation planning: a humanitarian
logistics based approach. Ind Eng Manag Syst 12(4):336–350
Djalante R, Garschagen M, Thomalla F, Shaw R (2017) Introduction: disaster risk reduction in
Indonesia: progress, challenges, and issues. In: Disaster risk reduction in Indonesia. Springer, pp
1–17
Kerle N, Oppenheimer C (2002) Satellite remote sensing as a tool in Lahar disaster management.
Disasters 26(2):140–160. https://doi.org/10.1111/1467-7717.00197
Kitjacharoenchai P, Min B-C, Lee S (2020) Two echelon vehicle routing problem with drones in
last mile delivery. Int J Prod Econ 225:107598. https://doi.org/10.1016/j.ijpe.2019.107598
Liperda RI, Pewira Redi AAN, Sekaringtyas NN, Astiana HB, Sopha BM, Maria Sri Asih A
(2020) Simulated annealing algorithm performance on two-echelon vehicle routing problem-
mapping operation with drones. In: IEEE international conference on industrial engineering and
engineering management. https://doi.org/10.1109/IEEM45057.2020.9309923
Marinelli M, Caggiani L, Ottomanelli M, Dell’Orco M (2018) En route truck–drone parcel delivery
for optimal vehicle routing strategies. IET Intel Transport Syst 12(4):253–261
Murray CC, Chu AG (2015) The flying sidekick traveling salesman problem: optimization of drone-
assisted parcel delivery. Transp Res Part C Emerg Technol 54:86–109
Office for the Coordination of Humanitarian Affairs (OCHA) (2014) Unmanned aerial vehicles in
humanitarian response
Sarma D, Das A, Bera UK (2020) Uncertain demand estimation with optimization of time and cost
using Facebook disaster map in emergency relief operation. Appl Soft Comput 87:105992
WHO (2002) Disasters & emergencies; definitions training package. In: WHO/EHA training
package (Issue March). http://apps.who.int/disasters/repo/7656.pdf
Zulkipli MA, Tahar KN (2018) Multirotor UAV-based photogrammetric mapping for road design.
Int J Opt 2018:1–7. https://doi.org/10.1155/2018/1871058
A Relief Aids Distribution Model
for the Pre-cooked Meals: The Case
of the Probable Disasater During Fasting
Periods in Padang City, Indonesia

Rahmad Inca Liperda , Nikorn Sirivongpaisal , Reinny Patrisina,

and Sirirat Suwatcharachaitiwong

1 Introduction

Disaster phenomena are increasingly recognized as a serious considerable concern

due to their tremendous effects on societies and environments. Among several types
of disasters, the tsunami is known as the sudden-onset natural disaster characterized
by its massive impacts. From 1992 to 2016, there are at least 290 tsunamis occurrences
registered (Gusiakov et al. 2019). During this 25-year period, the 2004 Indian Ocean
tsunami was acknowledged to be one of the deadliest catastrophes in history. In
Indonesia alone, this trans-oceanic mega-tsunami has led to 221,000 people death or
missing, half of million people being affected, as well as damages to infrastructures
and surroundings (Ali et al. 2019). The hazards with a similar potential magnitude
that may take place in the future must be recognized as an essential issue in order to
minimize the possible consequences of the upcoming calamities.
The severities and complexities carried out by a disaster substantially require a
thorough deliberation in enhancing the capability of the disaster countermeasure. In
order to come up with the effects of this casualty, disaster management is widely

R. I. Liperda
Department of Logistics Engineering, Universitas Pertamina, South Jakarta, Indonesia
e-mail: [email protected]
N. Sirivongpaisal (B) · S. Suwatcharachaitiwong
Department of Industrial Engineering, Prince of Songkla University, Hat Yai, Thailand
e-mail: [email protected]
S. Suwatcharachaitiwong
e-mail: [email protected]
R. Patrisina
Department of Industrial Engineering, Universitas Andalas, Padang City, Indonesia
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 243
P. Golinska-Dawson et al. (eds.), Smart and Sustainable Supply Chain
and Logistics—Challenges, Methods and Best Practices, EcoProduction,
https://doi.org/10.1007/978-3-031-15412-6_16
244 R. I. Liperda et al.

known as the fundamental strategy for providing decent assistance to the benefi-
ciaries. Disaster management mainly consists of four stages including mitigation,
preparation, response and recovery (Altay and Green III 2006). The mitigation stage
involves the efforts of either preventing or reducing the impacts of the disaster onset.
Meanwhile in the preparedness stage, the community is prepared to respond when
a disaster occurs. In the response stage, the activities of resources’ employment as
well as the implementation of the emergency procedures are performed to preserve
lives, properties, the environments, and the socio-economic structures of the commu-
nity. Subsequently, the recovery stage implicates the long-term activities aiming to
facilitate the community to recover from a disaster.
In humanitarian operations, the level of preparedness significantly affects to the
effectiveness of relief response (Van Wassenhove 2006). The preparedness strategies
must be comprehensively managed to provide a successful response since the lack
of preparation may result with delays and losses in the response stage. During the
preparedness stage, the process of organizing effective humanitarian logistics deci-
sions to be applied in the response stage is crucial. With regards to its vital role in
assisting about 80% of the disaster relief activities (Van Wassenhove 2006), human-
itarian logistics has been thought of as a key factor in determining the successful or
failed operations.
The central issue arising at the relief logistics operations is to ensure that the
beneficiaries’ basic needs are sufficiently fulfilled during the initial response, partic-
ularly in the first 72 h after the disaster occurrence. Within these periods, an adequate
amount of relief aids must be distributed to the demand points in a timely moment
(Ahmadi et al. 2015). The decisions on the relief delivery must take into account
the trade-offs between the increment of the required demands over the planning
periods (Inca and Nikorn 2019) and the scarce availability of resources in the first
72 h of the emergency periods. The other problems related to network failures are
also critical to be addressed in defining the appropriate relief distribution decisions
(Ahmadi et al. 2015). Another constraint regarding the foods that are being delivered
during these crucial periods must be taken into deliberation since it is very diffi-
cult, if not impossible, to establish a kitchen facility that can serve the meals to the
recipients. Delivering the pre-cooked meals becomes a feasible alternative so that the
recipients can consume them directly after their arrival. Furthermore, particularly in
the majority of Muslim-populated countries, the disaster that occurs in the fasting
periods may lead to another distinctive challenge in the relief delivery efforts. In the
fasting periods, Muslims are obligated to eat or drink only between sunset and dawn
(Azizi 2002). The emerging challenge regarding the relief aids delivery during these
periods is to send the relief foods within the specified duration in order to keep the
freshness of the foods as well as to ensure that the beneficiaries receive the basic
needs at the right time with the right amount of aids.
This research intends to develop a multi-period multi-trip relief delivery for a two-
tier relief distribution network in order to distribute the pre-cooked meals by taking
into account the possibility of disaster occurrence during the fasting periods. This
research addresses several substantial considerations: (1) the distinctive demands
throughout the first 72 h of emergency periods, (2) the scarcity of relief vehicles, (3)
A Relief Aids Distribution Model for the Pre-cooked Meals: The Case … 245

the expected network failures and (4) the specified delivery duration. In this work,
the expected network failures due to the effects of the disaster impacts are modelled
in the Geographic Information Systems (GIS) environment to obtain the expected
passable roads to be further utilized in the optimization process. A Mixed Integer
Linear Programming (MILP) model is then proposed to minimize the transporta-
tion costs regarding the routing decisions as well as the lateness and the shortage
penalty costs due to the late-satisfied and unsatisfied fulfilment. A case study of the
predicted upcoming hazard Mentawai Megathrust in Padang City, Indonesia is used
to demonstrate the numerical experiments of our model.
The remainder of this paper is organized as follows. Section 2 reviews the relevant
literatures related to the intended problem. Section 3 presents the problem descrip-
tion and model development. Section 4 provides the observed case study. Section 5
presents the computational results and analyses. Section 6 presents the conclusions
obtained from this research.

2 Literature Review

In recent years, a number of studies have sought to formulate the distribution models
related to the relief aids fulfilment in the post-disaster aftermath. Tzeng et al.,
conducted a fuzzy multi-objective programming model for minimizing the total
costs and maximizing the minimal satisfaction level in relief distribution (Tzeng
et al. 2007). Balcik et al., formulated the last mile distribution model for distributing
the emergency relief supplies in order to minimize the transportation costs and maxi-
mizing benefits to the beneficiaries (Balcik et al. 2008). A two-stage vehicle routing
model for both the planning stage and operational stage is proposed by Shen et al.
(2009). In their research, the mathematical formulation for both stages is used to solve
the large-scale bioterrorism case. Liberatore et al., concerned on the damaged distri-
bution infrastructures to generate the delivery planning of the emergency goods in
the aftermath of a disaster (Liberatore et al. 2014). Ahmadi et al., developed a multi-
depot location-routing problem by taking into account the probability of network
failures, the use of multiple vehicles and the consideration to standard relief time in
the last mile distribution (Ahmadi et al. 2015). Patrisina et al., considered a three-
tier relief distribution network including warehouses, local distribution centres, and
refugee camps to propose a two-stage programming model applied in the human-
itarian operations (Patrisina et al. 2018). Maghfiroh and Hanaoka presented multi-
modal transportation with multi-trip delivery for the three-tier relief chain in the
disaster response (Maghfiroh and Hanaoka 2020). By considering the condition of
network infrastructures and accessibility of supplies and modes of transport, their
work determined the location of the logistics operational areas as well as the assigned
modes of transport along with its allocated loaded amount over the planning horizon.
Mahtab et al., proposed a decision model for pre-disaster and post-disaster events
which is multi-objective stochastic programming applied to a case study of flood
in Bangladesh. The results of the model can support optimal resource allocation in
246 R. I. Liperda et al.

humanitarian logistics. Sources of uncertainty in the model were considered such as

demand, node reachability under a specific mode of transport, etc. The model solu-
tions provide the location of temporary facilities, the amount of commodity to be
prepared, the distribution schedules with the dispatched vehicles (Mahtab et al. 2021).
Nezhadroshan et al., designed a humanitarian logistics network integrating multiple
central warehouses with local distribution centers based on a probabilistic-stochastic
programming approach. Their model was intended to be robust and resilient. It guar-
antees the delivery of the crucial supplies to the beneficiaries under both operational
and disruptive risks. The proposed model was focused on the transportation network’s
routes after an earthquake which took a real case study of Mazandaran province in
the north of Iran to validate the model (Nezhadroshan et al. 2021). Boostani et al.,
proposed a sustainable relief logistics network model under the strategic and tactical
levels. The model describes a three-tier chain network in the pre-disaster and post-
disaster phases, focusing on facility location, procurement, and resource allocation. It
was designed as a multi-objective mixed-integer stochastic programming model with
minimizing the total costs of the relief supply chain, maximizing the social welfare,
and minimizing the environmental impacts. Two solution methods, which are the
Compromise Programming (CP) technique and Lexicographic Optimization Method
(LOM), were compared under two numerical examples based on surveying data of
probable Tehran earthquakes (Boostani et al. 2021). Shiripour et al., conducted an
empirical study of the earthquakes in the city of Tehran, Iran. They devised the inte-
grated mathematical model of the casualties’ transportation to uncover the temporary
locations of aid stations and their capacities, the percentage of the casualties under
various severities allocated to each station, as well, to discover the different routes
and the number of vehicles in order to minimize the total relief time. An approach,
called the circle-based approach, was developed and solved by an evolutionary algo-
rithm based on simulated annealing and a discrete version of the imperialist compet-
itive algorithm (Shiripour and Mahdavi-Amiri 2021). Garcia-Alviz et al., purposely
studied the performance of post-disaster relief operations under the transportation
networks disruption. They aimed to propose the strategic plans which can aid the
victims promptly. The occurrence of heterogeneous types of roads disruptions was
incorporated into the mathematical model with coordinating road network restoration
and relief activities. The objective of the developed model was to acquire a restoration
plan supporting the relief operations under the set of scheduling, routing restoration
machines, and relief vehicles. The heuristic algorithm was built and applied with a
realistic case study of flooding in the Mojana region of northern Colombia in 2010–
2011 (García-Alviz et al. 2021). Moddassir et al., proposed a robust optimization
model to enclose the uncertainty situations in the relief distribution centers (RDCs),
mainly are RDCs’ capacity and relief demand. The key idea in the study was focused
on the occurrence of the second disaster which is relatively in time and location with
the first disaster. It can result in more misery to the victims and more difficulties
to the relief distribution. The proposed model to minimize the total transportation
cost, additional RDC cost, and shortage of commodities were formulated and solved,
effectively. Also, the sensitivity analysis was performed to support an insightful deci-
sion (Nayeem and Lee 2021). Zhang et al., constructed a distribution relief model
A Relief Aids Distribution Model for the Pre-cooked Meals: The Case … 247

by proposing a sustainable last mile relief network concentrating on three perspec-

tives which are social factors, disaster relief efficiency, and the economic cost. A
multi-objective model was formulated to maximize the fair supplies distribution and
to minimize the transportation time and operating cost. Uncertainties in the model
are including the disaster situation, transportation time, freight, road capacity, and
demand. These uncertainty variables were considered as an ambiguity set incorpo-
rating the bounds, means, and mean absolute deviations. With the solution method
of a revised multi-choice goal programming, the answer were attained and verified
with a case study of the Banten tsunami (Zhang et al. 2022).

3 Problem Description and Model Development

This section firstly describes the problem description of the observed system. After-
ward, we present the modelling approach as well as the mathematical formulation.
This research addresses a two-tier relief distribution network consisting of a
permanent warehouse and its associated demand points whereas a warehouse is
allowed to deliver the relief items to a number of demand points. We consider the
variation of demand requirements over the first 72 h of the initial response phase to
obtain the appropriate relief delivery decisions. In particular, this work deliberates the
essential conditions related to the delivery duration in the fasting periods regarding
the pre-cooked meals distribution, as shown in Fig. 1. During the fasting periods,
the recipients are allowed to have a meal after sunset, which is further defined as
the delivery targeted time. The process of delivery will be started before the targeted
time. However, late-delivery that exceeds the targeted time is permitted by adding
an additional lateness penalty reflecting the recipients’ suffering time to wait for the
late-arrival delivery. In order to keep the freshness of the foods on its arrival, the late
delivery is only allowed to be performed within a specific duration. The shortage
penalty will be charged if the available vehicles are not able to fulfil the demand
needs during the assigned duration. With regards to the possibility of road disrup-
tions, the GIS environment is employed to generate safe candidate routes that avoid
the expected damaged roads. Furthermore, this study takes into account the scarcity
of the available vehicles in the post-disaster occurrence by allowing the vehicles to
perform multi-trip delivery according to their allocated capacities during the assigned
duration in each day of emergency periods. In summary, the operational decisions
related to the intended problem involve: (1) the number of pre-cooked meals to be
delivered to each demand point, (2) the vehicle routing and (3) the number of the
assigned vehicles along with their allocated trips.
The modelling approach encompassed in this research consists of three phases,
which is basically modified from the work of (Balcik et al. 2008), as shown in
Fig. 2. Phase 1 is intended to generate the travel time and distance matrix of each
location by considering the avoidance of the undesired road networks by utilizing
GIS environment. Next, the determination of the possible delivery routes including
the cost and duration for each vehicle along with the demand points visited by each
248 R. I. Liperda et al.

Fig. 1 Illustration of delivery duration

route is enumerated in Phase 2. Finally, the optimization process is employed in

Phase 3 to obtain the daily delivery routing including the assigned vehicles along
with their associated trips and the daily delivered amount to be sent to each demand
points.
The proposed model regarding the relief aids distribution is presented in the
following notations:
Sets
T set of days in the planning horizon
K set of vehicles
R set of route candidates
H set of trips
N set of demand locations
N set of demand locations visited on route r ∈ R

iments of the case study.

Set of warehouse Set of demand points Set of road network Set of network barriers

Phase 1: GIS environment

Travel time and distance matrix of each location with consideration to the avoidance of the expected
damaged roads

Phase 2: Numerical Calculation Set of vehicles

Set of candidates routes:

Cost and duration for each vehicle and its associated routes
Set of demand points visited by each route Daily delivery routes and
associated trips of each
assigned vehicle
Phase 3: Optimization
Daily delivered amount
Set of days in Daily demand Set of penalty Capacity (in Maximum Set of delivery sent to each location by
emergency needs at each costs factor: weight) of allowable durations: each vehicle
periods location Penalty cost each vehicle trips for Prompt
factor for late each vehicle delivery
delivery duration
Penalty cost Late delivery
factor for duration
shortage
delivery

Fig. 2 Modelling approach

A Relief Aids Distribution Model for the Pre-cooked Meals: The Case … 249

Routing Parameters
crk cost of route r ∈ R for vehicle k ∈ K
qk capacity of vehicle k ∈ K (weight)
Erk travel time (in hour) required by vehicle k ∈ K to use route r ∈ R
F duration allowance (in hour) to perform prompt delivery
G delivery duration span (in hour)
Demand parameters
dit demand required (in weight) by location i ∈ N on day t ∈ T
P penalty cost factor for late delivery.
U penalty cost factor for unsatisfied fulfilment
Routing decision variables
(
t = 1 if router ∈ R is visited by vehicle k ∈ K using trip h ∈ H on period t ∈ T
Xrhk
0 otherwise

Lateness decision variables

( { { t
1 if r∈R h∈H Xrhk Erk ≤ F
Atk =
0 otherwise

Delivery decision variables

t
Yirhk amount of demand delivered to location i ∈ N by vehicle k ∈ K using route
r ∈ R on trip h ∈ H on day t ∈ T
Ltk penalty cost associated with the late delivery using vehicle k ∈ K on day t ∈ T
Sit penalty cost associated with the unsatisfied fulfilment at location i ∈ N on day
t∈T
Bit fraction of unsatisfied demand at location i ∈ N on day t ∈ T
Objective function
{{{{ {{ {{
Minimize Z = t
crk Xrhk + Ltk + Sit (1)
r∈R h∈H t∈T k∈K t∈T k∈K i∈I t∈T

Constraints
(( ) )
{{
Ltk = PAtk t
Xrhk Erk −F ∀t ∈ T , k ∈ K (2)
r∈R h∈H
{{
t
Xrhk Erk ≤ G ∀t ∈ T , k ∈ K (3)
r∈R h∈H

Sit = UBit ∀i ∈ N , t ∈ T (4)

250 R. I. Liperda et al.

⎛ ⎞
{ {{ ( )
Bit = ⎝dit − t ⎠
Yirhk / dit ∀i ∈ N , t ∈ T (5)
r:i∈N (r) h∈H k∈k
{ {{
t
Yirhk ≤ dit ∀i ∈ N , t ∈ T (6)
r:i∈N (r) h∈H k∈K
{
t
Yirhk ≤ qk Xrhk
t
∀r ∈ R, h ∈ H , t ∈ T , k ∈ K (7)
i∈N (r)

0 ≤ Bit ≤ 1 ∀i ∈ N , t ∈ T (8)

t
Yirhk ≥0 ∀i ∈ N , r ∈ R, h ∈ H , t ∈ T , k ∈ K (9)

t
xrhk ∈ {0, 1} ∀r ∈ R, h ∈ H , t ∈ T , k ∈ K (10)

Atk ∈ {0, 1} ∀t ∈ T , k ∈ K (11)

The objective function (1) aims to minimize total routing cost, total lateness
penalty cost and total shortage cost over the planning periods. Constraint (2) defines
the lateness penalty cost due to the late delivery performed by each vehicle on
each day. Constraint (3) guarantees that the total travel time of each assigned
vehicle on each day does not exceed the delivery duration. Constraint (4) deter-
mines shortage penalty cost regarding the unsatisfied fulfilment at each location on
each day. Constraint (5) deliberates the fraction of the unsatisfied fulfilment for each
location on each day. Constraint (6) ensures that the total amount of items delivered
is less or equal to the required demand in each location on each day. The vehicle
capacity constraint is defined in Constraint (7). Constraint (8) guarantees that the
fraction of unsatisfied fulfilment is between zero and one. Constraint (9) is the non-
negativity constraint. Constraint (10) and (11) define the binary value of the routing
variable and the lateness delivery variable, respectively.
The proposed model however consists of non-linear formula particularly in the
lateness decision variables. A linearization technique suggested in the optimization
tool is employed by converting the non-linear function to a series of linear, math-
ematically equivalent expression (Lindo, 2020). The resulted Mixed Integer Linear
Programming (MILP) model is then used to solve the computational experiments of
the case study.
In order to solve the proposed MILP model, the Branch and Boundary Algorithm
is employed. This algorithm is widely known for its efficiency to generate the exact
solutions of the non-convex and combinatorial problems that cannot be solved in
polynomial time (Huang et al. 2021). Prior to commencing the investigation, the
problems are iteratively split into smaller sub-problems by creating a search tree. The
A Relief Aids Distribution Model for the Pre-cooked Meals: The Case … 251

divide-and-conquer algorithm is implemented by removing the integral conditions

of a problem to compute the linear programming (LP) relaxation.

4 The Case Study

This research takes into account the case study of the probable hazard Mentawai
Megathrust in Indonesia to demonstrate the effectiveness of our modelling approach.
The predicted ruptures that may suddenly displace along the Megathrust faults are
expected to generate great earthquakes triggering a large volume of tsunami waves
(Sieh 2006). Padang is becoming the most threatened area due to its flat topography
and the high population living in this city (Liperda et al. 2021). In this study, a
two-tier relief distribution network is deliberated by allowing direct delivery from a
warehouse to its demand points. Table 1 represents the location of each node as well
as the required demands over the three days of emergency periods. The demand needs
are calculated based on the expected arrival of the refugees in each day referring to the
recent contingency planning data developed by the West Sumatra’s local authority
for disaster countermeasure (BPBD Kota Padang 2017). This research assumes that
a refugee will require a package of pre-cooked meals with an associated weight of
0.59 kg.
This work considers four developed scenarios according to the substantive condi-
tions in the real-world case study, as depicted in Table 2. With regards to the scarce
availability of the vehicles that is common in the post-disaster occurrence, this
study takes into account the use of the limited number of vehicles with multiple
trips allowance to perform an efficient delivery. The set of homogenous vehicles is
represented in scenarios 1 and 3 while the set of heterogeneous vehicles is defined
in scenarios 2 and 4, respectively. In addition, this work presumes that the worth
consuming span of the pre-cooked meals is five hours. This assumption is adopted to
restrict the delivery duration allowance in order to ensure that the beneficiaries receive

Table 1 Location of depot and demand points

ID Locations Latitude Longitude Demand Requirements (ton)
Day 1 Day 2 Day 3
Depot
0 UPT BNPB Pusdalops PB −0.953111 100.428493 – – –
Demand points
1 Sungai Sapih −0.902755 100.39697 9.43 14.34 18.38
2 Kubu Marapalam −0.950678 100.383768 25.22 43.91 59.32
3 Ship Loading Point −0.961905 100.39454 1.92 5.78 8.96
4 Banuaran nan XX −0.968938 100.384775 5.32 14.81 22.64
Total 41.89 78.84 109.3
252 R. I. Liperda et al.

Table 2 Developed scenarios

Scenario Vehicle, Max Delivery Duration Duration Penalty Penalty
capacity trips duration of prompt of late Cost cost factor
6 4 (hr) delivery delivery factor for for
Wheel Wheel (hr) (hr) late unsatisfied
delivery fulfilment
($/hr) ($/% of
unsatisfied
1 3 (6 0 4 5 2 3 10 1000
ton)
2 3 (6 1 (4 4 5 2 3 10 1000
ton) ton)
3 3 (6 0 4 5 4 1 10 1000
ton)
4 3 (6 1 (4 4 5 4 1 10 1000
ton) ton)

decent-consumable meals. Moreover, different sets of the prompt delivery allowance

are investigated in this work by setting up the different values for the prompt delivery
span. Ultimately, the criticality between the late-satisfied and unsatisfied fulfilment
is interpreted by the discrepancy of the penalty amount charged.

5 Computational Results and Analyses

This section demonstrates the computational results and analyses of the observed
case study. The route candidates are firstly determined by considering the expected
network failures. Next, the optimization tool is employed to generate the routing
assignments based on the developed scenarios. Subsequently, the trade-offs between
the obtained solutions and the late-satisfied as well as unsatisfied fulfilment are
described to show how the components of the problem affect the routing decisions.
In this work, the computer specifications utilized to obtain both GIS-based analysis
and optimization solutions are CPU core i5 with 4 GB of RAM.
The depiction of the Origin–Destination Matrix by considering the expected
network failures is mapped in Fig. 4. This work utilizes ArcMap 10.2 to obtain the
straight-line representation of the OD Matrix in Padang City. The Network Analyst
Extension in ArcMap 10.2 is employed to conduct the least-distance measurements
between multiple nodes using Dijkstra’s Algorithm. As illustrated in Fig. 3, OD
matrix analyst allows users to specify the number of origins and destinations to
be measured by including several intended constraints and parameters such as the
restricted road to be avoided during delivery activities. The route candidates are
then generated by taking into account the travel time for each candidate as well
as the distribution cost for each vehicle type, as shown in Table 3. As indicated in
Table 3, we restrict the number of route candidates to nine candidates in order to
A Relief Aids Distribution Model for the Pre-cooked Meals: The Case … 253

decrease the computational time in the optimization phase by limiting the total travel
time. However, further investigation may be carried out by considering all of the
feasible candidates.
The optimization phase of the proposed model is performed by utilizing LINGO
18.0 that employs the branch-and-bound algorithm to generate the optimization
results. The proposed MILP model of each developed scenario results with the global
optimal solutions. Table 4 presents the results of the routing decisions as well as the
number of delivered items for each test problem over the three days of emergency
periods. In each scenario, the model primarily chooses the full truckload shipment
to maximize the assigned capacity constraints. From Table 4, it is observed that the
daily routing assignments are immensely affected by the increment of the demand
requirements over the three days of emergency periods. Further analysis shows that
the multi-trip delivery assignments are essential to be applied in the emergency
periods to come up with the scarce availability of vehicles. Consequently, the report
on the total costs of the associated test problems is presented in Table 5. With regards
to the model objective for minimizing the sum of total costs, Table 5 reveals that

Fig. 3 Illustration of Network Analyst Interface

254 R. I. Liperda et al.

Fig. 4 OD Matrix Representation by Considering Expected Damaged Roads

Table 3 Route candidates

Route candidates Tours Travel time (hr) Cost ($)
6 Wheel 4 Wheel
R1 0, 1, 0 0.95 9 8
R2 0, 2, 0 0.78 8 8
R3 0, 3, 0 0.82 8 8
R4 0, 4, 0 0.71 8 7
R5 0, 1, 2, 0 1.29 13 12
R6 0, 1, 3, 0 1.37 13 12
R7 0, 1, 4, 0 1.26 12 12
R8 0, 2, 3, 0 1.16 12 11
R9 0, 2, 4, 0 1.05 11 11

the optimal solutions are achieved by balancing the late-delivery penalty costs, the
shortage penalty costs and routing costs. These relative values changes accordingly
with the changes of the input parameters. As presented in Table 5, it is observed
that scenario 4 result with the least total costs. Moreover, this scenario gives us the
least penalty costs regarding the late-delivery and unsatisfied fulfilment. Due to the
dynamic and uncertain circumstances that might occur in the disaster aftermath, the
solutions may turn into different possible outcomes.
A Relief Aids Distribution Model for the Pre-cooked Meals: The Case … 255

Table 4 Routing assignments and amount of delivered items

Vehicle Assigned route Trip Amount delivered in day 1 (ton)
Node 1 Node 2 Node 3 Node 4
Scenario 1
6 Wheel A R1 1 6
R9 1 3.14 2.86
6 Wheel B R8 1 4.08 1.92
R2 2 6
6 Wheel C R7 1 3.43 2.46
R2 1 6
Scenario 2
6 Wheel A R2 1 6
R4 1 5.32
6 Wheel B R1 1 6
R2 1 6
6 Wheel C R8 1 4.08 1.92
R2 1 6
4 Wheel A R1 1 3.43
R2 1 3.14
Scenario 3
6 Wheel A R1 1 6
R8 1 4.08 1.92
R2 1 6
6 Wheel B R2 1 5.89
R9 1 3.25 2.75
R7 1 3.43 2.57
6 Wheel C R2 1 6
Scenario 4
6 Wheel A R8 1 4.08 1.92
R7 1 3.43 2.57
6 Wheel B R1 1 6
R9 1 3.25 2.75
6 Wheel C R2 2 6
(continued)

The detailed information on the lateness and shortage percentage values as well as
the amount of relief aids delivered for the developed scenarios are presented in Table
6 and Table 7, respectively. The lateness time is acknowledged when the total routing
time exceeds the prompt delivery limit in order to delineate the refugees’ suffering
time waiting for the late-arrival delivery. Meanwhile, the shortage percentage values
256 R. I. Liperda et al.

Table 4 (continued)
Vehicle Assigned route Trip Amount delivered in day 1 (ton)
Node 1 Node 2 Node 3 Node 4
R2 1 5.89
Vehicle Assigned route Trip Amount delivered in day 2 (ton)
Node 1 Node 2 Node 3 Node 4
Scenario 1
6 Wheel A R2 3 6
R3 1 5.78
6 Wheel B R1 1 6
R2 2 6
R4 2 6
6 Wheel C R1 1 2.34
R1 1 6
R2 1 4.72
R2 1 6
R9 1 3.19 2.81
Scenario 2
6 Wheel A R1 2 6
R2 1 6
6 Wheel B R2 3 6
R3 1 5.78
R4 1 6
6 Wheel C R2 3 6
R4 1 6
4 Wheel A R5 1 2.34 1.66
R4 1 2.81
Scenario 3
6 Wheel A R2 4 6
R3 1 5.78
6 Wheel B R2 2 6
R2 1 5.29
R4 2 6
6 Wheel C R1 2 6
R1 1 2.34
R9 1 2.62 2.81
(continued)
A Relief Aids Distribution Model for the Pre-cooked Meals: The Case … 257

Table 4 (continued)
Vehicle Assigned route Trip Amount delivered in day 2 (ton)
Node 1 Node 2 Node 3 Node 4
Scenario 4
6 Wheel A R1 1 6
R2 1 6
R9 1 3.19 2.81
R3 1 5.78
6 Wheel B R1 1 6
R2 2 6
R4 2 6
6 Wheel C R2 3 6
R2 1 4.72
4 Wheel A R1 1 2.34
Vehicle Assigned route Trip Amount delivered in day 3 (ton)
Node 1 Node 2 Node 3 Node 4
Scenario 1
6 Wheel A R1 1 6
R2 4 6
R3 1 6
6 Wheel B R8 1 3.04 2.96
R2 2 6
R4 1 4.64
R4 2 6
6 Wheel C R1 2 6
R2 3 6
R4 1 6
Scenario 2
6 Wheel A R2 4 6
R9 1 5.32 0.68
R3 1 6
6 Wheel B R1 1 5.34
R1 1 6
R2 2 6
R4 1 6
6 Wheel C R1 1 6
R2 3 6
R4 2 6
4 Wheel A R4 1 3.96
R6 1 1.04 2.96
(continued)
258 R. I. Liperda et al.

Table 4 (continued)
Vehicle Assigned route Trip Amount delivered in day 3 (ton)
Node 1 Node 2 Node 3 Node 4
Scenario 3
6 Wheel A R8 1 3.04 2.96
R2 2 6
R4 3 6
6 Wheel B R1 2 6
R2 3 6
R4 1 4.64
6 Wheel C R1 1 6
R2 4 6
R3 1 6
Scenario 4
6 Wheel A R1 1 6
R2 3 6
R4 1 6
6 Wheel B R2 4 6
R4 2 6
6 Wheel C R1 1 6
R2 2 6
R2 1 5.32
R4 1 4.64
4 Wheel A R1 1 4
R3 1 3.34
R3 1 4
R6 1 2.38 1.62

Table 5 Total costs

Scenario Total cost ($) Transportation cost Penalty cost of late Penalty cost of
($) delivery ($) unsatisfied fulfilment
($)
Scenario 1 546.61 337 150.50 59.11
Scenario 2 468.19 347 115.50 5.69
Scenario 3 423.01 337 26.90 59.11
Scenario 4 356.40 351 5.40 0
A Relief Aids Distribution Model for the Pre-cooked Meals: The Case … 259

are yielded as a result of the insufficient fulfilment regarding the lack of capability in
delivering the required amount of the relief items within the given duration limit. The
relative values of the penalty cost factor charged for both late satisfied and unsatisfied
fulfilment emphasize the criticality between these decisions. As a consequence, the
model will prioritize the late-delivery rather than the shortage fulfilment. However,
we should notify that the late-satisfied and unsatisfied degrees regarding the victims’
suffering are difficult to be quantified into the real costs. A further study with more
focus on investigating the vulnerabilities of the disaster victims particularly in the
immediate aftermath of a disaster is therefore suggested.
Table 8 shows the solution statistics for each scenario using LINGO 18.0. As
shown in Table 8, scenario 4 results with the longest run time to generate the global
optimal solutions. It can be implied from these results that the scale of the model
as well as the components of the problems obviously affect the computational time.
In addition, the computer specifications used in this research may also lead to the
length of the required running time. With regards to its classification as an NP-
hard problem, this is becoming an important issue for future research to provide
a practical methodology in order to effectively solve the larger instances within
reasonable computational time.

6 Conclusions

This paper was aimed to develop the analytical approach for a multi-period multi-
trip relief delivery for optimizing routing decisions for distributing the pre-cooked
meals during fasting periods. This study has shown that the proposed modelling
approach has the capability to be applied in the humanitarian logistics operations
by enabling the relief stakeholders to address several substantial issues of the real-
world problems. Several noteworthy contributions of this paper are considering the
distinct demand requirements over the planning horizon, utilization of the scarce
resources, avoidance of expected network failures and limited delivery duration. In
this study, GIS environment was employed to the generate the expected passable
roads between each node by considering the possible network failures. Furthermore,
the MILP model was developed to minimize the total transportation cost as well as
the late-satisfied and unsatisfied fulfilment. The computational results of the case
study Mentawai Megathrust hazard in Padang City, Indonesia was used to highlight
the relationship between various decisions with the changes in the model parameters.
The findings in this report are subject to at least three limitations. First, the solutions
may change according to dynamic and uncertain condition occurring in the real-case.
Second, the degrees of the victims’ vulnerabilities in the post-disaster aftermath must
be further investigated. Third, the problems with larger instances may increase the
solution times which require practical methodology approach to effectively solve
this NP-hard problem within reasonable computational time. Particularly for the last
260 R. I. Liperda et al.

Table 6 Routing time and lateness time

Vehicle Day
Day 1 Day 2 Day 3
Scenario 1
6 Wheel A Routing time (hr): 2 3.16 4.89
Lateness (hr): 0 1.16 2.89
6 Wheel B Routing time (hr): 3.10 3.93 4.85
Lateness (hr): 1.10 1.93 2.85
6 Wheel C Routing time (hr): 2.04 4.51 4.95
Lateness (hr): 0.04 2.51 2.95
Scenario 2
6 Wheel A Routing time (hr): 1.49 2.68 4.99
Lateness (hr): 0 0.68 2.99
6 Wheel B Routing time (hr): 1.73 3.87 4.17
Lateness (hr): 0 1.87 2.17
6 Wheel C Routing time (hr): 1.94 3.05 4.71
Lateness (hr): 0 1.05 2.71
4 Wheel A Routing time (hr): 1.73 2 2.08
Lateness (hr): 0 0 0.08
Scenario 3
6 Wheel A Routing time (hr): 2.89 3.94 4.85
Lateness (hr): 0 0 0.85
6 Wheel B Routing time (hr): 3.09 3.76 4.95
Lateness (hr): 0 0 0.95
6 Wheel C Routing time (hr): 0.78 3.90 4.89
Lateness (hr): 0 0 0.89
Scenario 4
6 Wheel A Routing time (hr): 2.42 3.60 4
Lateness (hr): 0 0 0
6 Wheel B Routing time (hr): 2 3.93 4.54
Lateness (hr): 0 0 0.54
6 Wheel C Routing time (hr): 2.34 3.12 4
Lateness (hr): 0 0 0
4 Wheel A Routing time (hr): 0 0.95 3.96
Lateness (hr): 0 0 0
A Relief Aids Distribution Model for the Pre-cooked Meals: The Case … 261

Table 7 Amount of delivered relief aids

Node (s) Day
Day 1 Day 2 Day 3
Scenario 1
Node 1 Delivery amount (ton): 9.43 14.34 18
Shortage percentage (%): 0% 0% 2.07%
Node 2 Delivery amount (ton): 25.22 43.91 57.04
Shortage percentage (%): 0% 0% 3.84%
Node 3 Delivery amount (ton): 1.92 5.78 8.96
Shortage percentage (%): 0% 0% 0%
Node 4 Delivery amount (ton): 5.32 14.81 22.64
Shortage percentage (%): 0% 0% 0%
Scenario 2
Node 1 Delivery amount (ton): 9.43 14.34 18.38
Shortage percentage (%): 0% 0% 0%
Node 2 Delivery amount (ton): 25.22 43.66 59.32
Shortage percentage (%): 0% 0.57% 0%
Node 3 Delivery amount (ton): 1.92 5.78 8.96
Shortage percentage (%): 0% 0% 0%
Node 4 Delivery amount (ton): 5.32 14.81 22.64
Shortage percentage (%): 0% 0% 0%
Scenario 3
Node 1 Delivery amount (ton): 9.43 14.34 18
Shortage percentage (%): 0% 0% 2.07%
Node 2 Delivery amount (ton): 25.22 43.91 57.04
Shortage percentage (%): 0% 0% 3.84%
Node 3 Delivery amount (ton): 1.92 5.78 8.96
Shortage percentage (%): 0% 0% 0%
Node 4 Delivery amount (ton): 5.32 14.81 22.64
Shortage percentage (%): 0% 0% 0%
Scenario 4
Node 1 Delivery amount (ton): 9.43 14.34 18.38
Shortage percentage (%): 0% 0% 0%
Node 2 Delivery amount (ton): 25.22 43.91 59.32
Shortage percentage (%): 0% 0% 0%
Node 3 Delivery amount (ton): 1.92 5.78 8.96
Shortage percentage (%): 0% 0% 0%
Node 4 Delivery amount (ton): 5.32 14.81 22.64
Shortage percentage (%): 0% 0% 0%
262 R. I. Liperda et al.

Table 8 Solution statistics

Scenario Extended Total solver Computational
solver steps iterations time (hr)
Scenario 1 619,950 8,924,534 1.29
Scenario 2 1,501,469 26,743,764 4.50
Scenario 3 1,044,432 12,291,150 1.91
Scenario 4 5,751,805 84,720,971 27.03

limitation addressed in this research, we are developing a metaheuristic algorithm to

provide good solutions with reasonable computational time in our current research.

References

Ahmadi M, Seifi A, Tootooni B (2015) A humanitarian logistics model for disaster relief operation
considering network failure and standard relief time: a case study on San Francisco district. Transp
Res Part E Logist Transp Rev 75:145–163. https://doi.org/10.1016/j.tre.2015.01.008
Ali MSS, Arsyad M, Kamaluddin A, Busthanul N, Dirpan A (2019) Community based disaster
management: Indonesian experience. IOP Conf Ser Earth Environ Sci 235(1). https://doi.org/10.
1088/1755-1315/235/1/012012
Altay N, Green III WG (2006) Interfaces with other disciplines OR/MS research in disaster opera-
tions management 175(December 2004):475–493. https://doi.org/10.1016/j.ejor.2005.05.016
Azizi F (2002) Review article. Res Islamic Fasting Health 22:186–191
Balcik B, Beamon BM, Smilowitz K (2008) Last mile distribution in humanitarian relief. J Intell
Transp Syst Technol Plan Oper 12(2):51–63. https://doi.org/10.1080/15472450802023329
Boostani A, Jolai F, Bozorgi-Amiri A (2021) Designing a sustainable humanitarian relief logistics
model in pre- and postdisaster management. Int J Sustain Transp 15(8):604–620. https://doi.org/
10.1080/15568318.2020.1773975
BPBD Kota Padang (2017) Dokumen Rencana Kontijensi Tsunami Kota Padang. BPBD Kota
Padang, Padang
García-Alviz J, Galindo G, Arellana J, Yie-Pinedo R (2021) Planning road network restoration
and relief distribution under heterogeneous road disruptions. OR Spect 43. Springer, Berlin,
Heidelberg. https://doi.org/10.1007/s00291-021-00644-x
Gusiakov VK, Dunbar PK, Arcos N (2019) Twenty-five years (1992–2016) of global tsunamis:
statistical and analytical overview. Pure Appl Geophys 176(7):2795–2807. https://doi.org/10.
1007/s00024-019-02113-7
Huang L, Chen X, Huo W, Wang J, Zhang F, Bai B, Shi L (2021) Branch and bound in mixed
integer linear programming problems: a survey of techniques and trends (November), pp 1–29.
Retrieved from http://arxiv.org/abs/2111.06257
Inca LR, Nikorn S (2019) The location-allocation decision under the dynamic increment of demand
for selecting the local distribution centers to face sumatra megathrust : study case of Padang. City
9(1):293–299
Liberatore F, Ortuño MT, Tirado G, Vitoriano B, Scaparra MP (2014) A hierarchical compromise
model for the joint optimization of recovery operations and distribution of emergency goods in
humanitarian logistics. Comput Oper Res 42:3–13. https://doi.org/10.1016/j.cor.2012.03.019
Lindo. (2020). LINGO: the modeling language and optimizer. 1415 North Dayton Street Chicago,
Illinois 60642: LINDO Systems, Inc. Retrieved from https://www.lindo.com/downloads/PDF/
LINGO.pdf
A Relief Aids Distribution Model for the Pre-cooked Meals: The Case … 263

Liperda RI, Aulya AR, Sukarno I, Lusiani M, Sirivongpaisal N (2021) Geological behavior (GBR)
a GIS—based optimization for the fuel location-allocation to face the probable tsunami: case of
Padang city 5(1):13–16. https://doi.org/10.26480/gbr.01.2021.13.16
Maghfiroh MFN, Hanaoka S (2020) Multi-modal relief distribution model for disaster response
operations progress in disaster science multi-modal relief distribution model for disaster response
operations. Progr Disaster Sci 6(April):100095. https://doi.org/10.1016/j.pdisas.2020.100095
Mahtab Z, Azeem A, Ali SM, Paul SK, Fathollahi-Fard AM (2021) Multi-objective robust-stochastic
optimisation of relief goods distribution under uncertainty: a real-life case study. Int J Syst Sci
Oper Logist. https://doi.org/10.1080/23302674.2021.1879305
Nayeem MK, Lee GM (2021) Robust design of relief distribution networks considering uncertainty.
Sustainability (switzerland) 13(16):1–23. https://doi.org/10.3390/su13169281
Nezhadroshan AM, Fathollahi-Fard AM, Hajiaghaei-Keshteli M (2021) A scenario-based
possibilistic-stochastic programming approach to address resilient humanitarian logistics consid-
ering travel time and resilience levels of facilities. Int J Syst Sci Oper Logist 8(4):321–347. https://
doi.org/10.1080/23302674.2020.1769766
Patrisina R, Sirivongpaisal N, Suthommanon S (2018) A logistical relief distribution preparedness
model : responses to a probable tsunami case study in West Sumatra, Indonesia. Ind Eng Manag
Syst 17(4):850–863
Shen Z, Dessouky M, Ordónez F (2009) A two-stage vehicle routing model for large-scale
bioterrorism emergencies. Networks 54(4):255–269
Shiripour S, Mahdavi-Amiri N (2021) Disaster relief on destructive transportation networks using a
circle-based approach. Transp Lett 13(8):568–590. https://doi.org/10.1080/19427867.2020.174
2417
Sieh K (2006) Sumatran megathrust earthquakes : from science to saving lives (June), pp 1947–1963.
https://doi.org/10.1098/rsta.2006.1807
Tzeng GH, Cheng HJ, Huang TD (2007) Multi-objective optimal planning for designing relief
delivery systems. Transp Res Part E Logist Transp Rev 43(6):673–686. https://doi.org/10.1016/
j.tre.2006.10.012
Van Wassenhove LN (2006) Blackett memorial lecture humanitarian aid logistics: supply chain
management in high gear. J Oper Res Soc 57(5):475–489. https://doi.org/10.1057/palgrave.jors.
2602125
Zhang P, Liu Y, Yang G, Zhang G (2022) A multi-objective distributionally robust model for
sustainable last mile relief network design problem. In: Annals of operations research, vol 309.
Springer US. https://doi.org/10.1007/s10479-020-03813-3
The Impact of the Use of Intelligent
Supply Chain Tools on the Transport,
Forwarding and Logistics Industry

Sylwia Konecka and Anna Łupicka

1 Introduction

The analysis of the benefits and costs of using intelligent solutions for the implemen-
tation of transport processes is of great importance not only due to the innovative
nature of these tools—although they require ordering and closer examination, but also
due to the great importance of the activities of Polish carriers in the European Union
(EU). The TFL industry (transport-forwarding-logistics) significantly contributes to
the development of the Polish economy, carriers support the operation of the industry
and the production sector, and also generate a considerable part of GDP (6%). 40%
of cabotage and 17% of all transports in the EU are handled by Polish transport
companies, which have been consistently in the first place in terms of the number of
tonnes of cargo transported in international transport since 2012 (Zysińska 2019).
Nevertheless, there are some problems in the industry. Experts recognize, first of all,
the dangers of the regulations currently being introduced at the EU level, which may
result in:
. an increase in the costs of transport networks (due to a reduction in their
efficiency), fuel costs and wages,
. reduction of revenues related to the restriction of market access—in the field of
cross-trade and cabotage services, and consequently also handling the import and
export of goods,
. increased risk and barriers to doing business, especially in the case of small
carriers, due to the increased complexity of the rules,

S. Konecka (B) · A. Łupicka

Department of Logistics, Poznań University of Economics and Business, Al. Niepodległości 10,
61-875 Poznań, Poland
e-mail: [email protected]
A. Łupicka
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 265
P. Golinska-Dawson et al. (eds.), Smart and Sustainable Supply Chain
and Logistics—Challenges, Methods and Best Practices, EcoProduction,
https://doi.org/10.1007/978-3-031-15412-6_17
266 S. Konecka and A. Lupicka

. forcing the internationalization of carriers—who operate today in many foreign

markets from Poland,
. an increase in the concentration level—as a result of the collapse of small enter-
prises that do not have an adequate share of transport in the country and on foreign
markets in their activities.
It should be noted that with the low level of margins in the road haulage trans-
port segment, an increase in costs may lead to the elimination of the least effective
entities from the market. In turn, the scientific community focuses on the growing
requirements in the field of environmental protection regarding lower CO2 emis-
sions and other pollutants, the use of more ecological drives and local restrictions
on the movement of internal combustion vehicles, which for entrepreneurs may also
contribute to an increase in costs. Attention is also drawn to the insufficient level of
development of national law in the field of new technologies.
It is also worth recognizing the area of implementing intelligent tools facilitating
transport management, which could be helpful in overcoming the above-mentioned
problems and supporting carriers in maintaining a competitive advantage in the inter-
national dimension. In addition, transport processes connect all links in the supply
chain and are of great importance in the EU in the light of the Sustainable and Smart
Mobility Strategy. The aim of the article is to propose a theoretically based concep-
tual framework for studying the impact of identified intelligent solutions on the TFL
industry. The study used a qualitative descriptive method referring to a critical anal-
ysis of domestic and foreign studies in the field of intelligent solutions, and made
a short review of selected theoretical approaches in the field of costs and benefits.
The potential benefits and costs of introducing intelligent solutions in the transport
industry were also identified.
The presented issues, due to their topicality and dynamics of development as
well as very poor research in the field of science, require in-depth research. The
conclusions obtained are of practical importance, as they allow for the identification
and assessment of benefits and costs for entities in the TFL industry, which are
brought by solutions of intelligent supply chains. Although research in the field of
transport is common and more and more studies also deal with industry 4.0, the
review of Polish literature carried out in 2020 (Młody and Weiner 2021) shows that
no studies linking the subject of road haulage transport with the latest IT solutions
have been conducted, 7 out of 68 of analyzed publications was related to logistics and
supply chains. There is also a lack of studies using quantitative methods and case
studies. Taking up a popular but new, poorly recognized, dynamically developing
topic allows to signal possible economic, social and environmental consequences,
becoming the basis for a more adequate socio-economic policy and creating strategies
for entrepreneurs in the TFL industry.
The article is theoretical and constitutes a kind of research reconnaissance. The
method of systematic literature review was used for its preparation. The following
databases were selected for the study: WoS, Scopus and Google Scholar. In the first
phase, the literature was reviewed in terms of solutions used in intelligent supply
chains, then the identified solutions were confronted with the transport activity related
The Impact of the Use of Intelligent Supply Chain Tools … 267

STAGE 1. Identification of intelligent solutions in supply chains and selection of those related
to transport (based on a literature review from the Scopus database)

STAGE 2. Identification of selected intelligent solutions used in road haulage transport,

selection of those most often described in the literature (based on Scopus and WoS
databases)

STAGE 3. Identification of theories about benefits and costs

STAGE 4. Identification of potential benefits and costs resulting from the use of selected
(most often described in road haulage transport) intelligent solutions (Google Scholar,
UEP multi-search database)

Fig. 1 Diagram of the identification steps carried out as part of the literature review. Source Own
study

to the transport of goods. The Google Scholar database was used to review the
literature in terms of the theory of benefits and costs and to search for potential
benefits and costs for the transport industry from the use of selected IT solutions.
Details of the search keywords used are provided in each subsection. The method
of operation aimed at achieving the set goal is reflected in the theoretical model
presented in Fig. 1.

2 Intelligent Solutions Used in Road Transport

According to the Dictionary of the Polish Language (www.sjp.pl 2021), intelligence

is “the ability to understand, learn and use the possessed knowledge and skills in
new situations”. Therefore, it can be ascribed not only to personal entities, but also
to organizations that are constantly operating in new situations today. Apart from the
term “intelligent organization”, there is also a “learning organization”. According to
Penc (2003), the main difference between a learning organization and an intelligent
organization is the desire (the last of the mentioned ones) to fully use the intellectual
potential of all employees.
Smart organizations are characterized by: agility (speed in action), adaptability,
self-regulation and self-optimization (adapting key business indicators to short-term
changes in the business climate), smooth borders and a network structure (enabling
restructuring and changing the scale of the organization’s operations depending on
the needs, e.g., by outsourcing or insourcing), as well as the ability to transform into
new, better forms in the long term (Kemal et al. 2002).
It seems that intelligent organizations should strive to maintain a balance between
acquired knowledge and shared knowledge. The formation of an intelligent organi-
zation is a process that consists in constantly increasing knowledge and changing
the mentality of its members, who learn to be more responsible, independent and
entrepreneurial related to their tasks (Kaczmarek 2013). The development of such an
enterprise may result from taking independent actions by managers and members of
268 S. Konecka and A. Lupicka

Table 1 Number of
The selected tool of intelligent supply Number of articles
publications on intelligent
chain
supply and transport chains
(Scopus database) RFID (Radio-frequency identification) 50
ITS (Intelligent transport systems) 35
IoT (Internet of things) 27
ICT (Information and communication 14
technologies)
AI (Artificial intelligence) 13
Automation 12
TMS (Transport management systems) 1
Source Own study

the organization, but it may also take place as a result of implementing the concept
of a virtual organization, strategic alliances or coopetition. (Romaniuk 2017)
The literature review shows that intelligent solutions applicable in a broader
context than a single enterprise, i.e., supply chains, are discussed in the literature on
the subject in such areas as types of algorithms—genetic or fuzzy-logical systems,
in the field of reverse logistics, intelligent systems in intermodal transport, in trans-
port urban and extra-urban, ITS system and security of their use, intelligent ware-
house management systems, Cloud of Things, RFID, autonomous vehicles (Konecka
2019). Initial research and literature review indicate that a significant part of these
tools relates to transport activities. The literature most often states that transport
companies can use evolutionary versions of ITS, RFID, cloud of things, blockchain,
IoT (Konecka and Maryniak 2020) (Table 1).
Knowledge of the impact of technology on the activities of companies in the TFL
industry is based mainly on industry reports. Only in one Polish publication appeared
a direct connection of transport activities with solutions of the so-called industry 4.0
in the form of the use of the concept of transport 4.0 (Brach 2019). There are few
publications dealing only with transport, in particular road transport, and only goods,
with the exception of passenger transport.
The preliminary assessment of the state of knowledge based on the number of
publications in the databases: Web of Science (WoS) and Scopus in the last five
years, i.e., from 2017 to 2021, shows that for the entry in the title or summary “road
AND transport” a pool of 43 902 publications was obtained in WoS, while looking
for the “road transport” version—2,624. In the Scopus database it was 13,898 and
3440 publications, respectively. Among these results, the following keywords were
checked: “Industry 4.0”, “Telematics”, “Internet of Things”, “Automation”, “Big
Data”, “Information and Communication Technologies”, “Artificial Intelligence”.
Only one article concerned a study on the impact of digitization on road haulage
transport using the example of Sweden, the study was carried out using the scenario
method and based on the opinions of 52 experts. The results of the number of
publications obtained are presented in Table 2.
The Impact of the Use of Intelligent Supply Chain Tools … 269

Table 2 Number of publications on road transport and selected smart tools

Tool/Solution WoS Scopus
Road and transport Road transport Road and transport Road transport
Industry 4.0 75 4 110 24
Automation 461 53 1259 294
Telematics 91 19 230 65
IoT (Internet of 179 11 457 77
things)
Big Data 610 31 828 168
ITS (Intelligent 967 145 1446 240
transport systems)
ICT (Information 56 11 280 70
and communication
technologies)
AI (Artificial 323 16 1358 272
intelligence)
Source Own study

The industry literature review shows that in practice, carriers use the following IT
solutions: freight exchanges, cargo tracking systems, TMS systems, tachographs,
electronic toll collection systems, telematics, sensors, GPS, translators, cargo
securing and monitoring, information on CO2 emissions. Thus, it can be assumed that
the solutions indicated in the literature are used rather by public entities or industry
leaders and not by carriers, but e.g., by logistics operators. Before the potential bene-
fits and costs resulting from the use of intelligent solutions are identified, those most
often discussed in the literature on the subject were selected, i.e., automation (for
transport, autonomous vehicles will be considered here), ITS, AI, Big Data and IoT.

3 The Concept of Benefit and Cost in the Light of Selected

Theories

There is no doubt that the emerging intelligent IT tools affect the concept of human
capital, production, consumption and exchange of goods both in theoretical and
practical terms. Human labor ceased to be the basic factor of production, and human
capital takes on an intellectual dimension. There is a concern among economists
about the need for humans to fight robots for jobs (Rifkin 1995). A reference to the
revolution based on ICT techniques and the current economic thought is the concept
of the process of creative destruction of Schumpeter. Since “information techniques
(ICT) are a fundamental element of most systems, the feature of which is constant
and dynamic development, often of a subversive nature” (disruptive technologies),
a Schumpter approach to innovation suggests itself (Goliński 2013). Currently, a
270 S. Konecka and A. Lupicka

scientific theory is expected to be able to explain the processes taking place and to
be reliable in predicting the future. It is also expected to be able to properly analyze
emerging trends and use a huge amount of data, which fits perfectly into the concept
of the latest revolution. The equilibrium models that have been in force for many
years, being a certain photograph of the optimal state on the market or, more broadly,
in the economy, cease to be applicable in the world of constant changes (Olender-
Skorek 2017). Hence the idea to take a closer look at the concept of benefits and
costs in theory in order to compare them to the benefits and costs of implementing
intelligent solutions in road transport of goods. Similarly, if modern IT solutions are
to be considered revolutionary in the transport industry, then one should also refer to
the classic concept of Kuhn’s revolution (2009). Nevertheless, first the considerations
were focused on the concept of benefits and costs, wondering who will benefit from
this revolution and who will lose.
When looking for a definition of benefits in economic terms, we can meet the two
most popular approaches:
. external benefits, i.e. benefits obtained from the production and consumption
of goods by people not directly involved in their production, consumption or
exchange, and
. economies of scale that occur in the long term and relate to the phenomenon of
lowering the average total costs along with increasing the scale of production and
introducing new technologies.
For example, according to Bentham, the essence of pleasure and unpleasantness
obtained from various sources is the same. The only difference between them lies in
the circumstances surrounding their creation, and therefore they should be the subject
of a felicific calculus of happiness for the individual and society. The property of an
object or activity by which it promotes the production of benefit or the prevention
of harm to the interested party, Bentham called utility. The benefit derived from the
decisions made should be regarded as an expression of the utility of an action or
object, rather than its utility, since the latter is merely a property of the object or
action by which it favors the production of benefits (Betham 1958/1780).
The meaning of the term “favorable decision” is related to the scope of the
expression “rational decision”. And making a rational decision is associated with the
demands of perfect information, consistency and maximization. The first of them,
assuming that the model decision maker knows all the possibilities of using the avail-
able resources, and that the expected benefit (utility) will be (utility) actually realized
as a result of taking these actions, is of particular importance in the context of the
current development of “big data”. In addition, Bentham’s original idea concerns the
concept of utility, allowing the benefits achieved by market participants to be located
both in the sphere of consumption and in the sphere of production. By analogy with
consumers benefiting from the consumption of goods, one can speak of producers
benefiting from their production. Exposing the benefits as a category expressing the
most universal goal of actions undertaken by market participants is a procedure that
precisely formulates the starting point for further analysis of the behavior of market
entities. The benefits that people strive for are a polymorphic category and very
The Impact of the Use of Intelligent Supply Chain Tools … 271

often completely immeasurable quantitatively, economists over the centuries have

tried to find replacement variables for which they assumed that they changed in the
same direction as satisfaction, satisfaction or benefit resulting from the undertaken
activities. For example, there are proxy variables such as wealth, utility, preferences,
profit, income, reciprocity, entropy, and well-being.
In the analysis of the result achieved by the producer from business activity,
two elements are juxtaposed: the costs of undertaken activities and the resulting
benefits. The costs are subjective and their magnitude depends on the personal char-
acteristics of the decision-maker. Hidden income is an expression of satisfaction
(benefits) resulting from running a business and may include, for example, such
elements as: social prestige, a sense of social security (peaceful life), membership
of environmental organizations such as associations, clubs or chambers, a sense of
professional independence and decision-making freedom as well as opportunities
for self-realization.
Among the theories regarding the concept of benefits and costs, there are also:
. Homans’ theory of social exchange (1967), which is a sociological and anthro-
pological theory, referring to behaviorism and economics, from which the notion
of exchange was adopted, social relations explained by the exchange of goods
(rewards and punishments) between individuals participating in the interaction;
. the theory of consumption chains of goals and means, developed by Newell and
Simon (1972), the idea of which is that the attributes of products inform the
consumer whether the desired goals, namely values and direct benefits, will be
met by the products;
. the theory of public goods and the theory of groups, in which Olson (2012) presents
the fundamental problem of the limitations of the effectiveness of collective action
undertaken by various types of groups or organizations. Rejects as untrue the
judgment that if a group has any collective goal or benefit, the individuals making
up that group will automatically seek to achieve it;
. the theory of comparative advantages, formulated by Ricardo and Hartwell (1971)
and based on the claim that international exchange can be beneficial for both
partners, when one produces most of the goods cheaper than the other. To obtain
such benefits, it is sufficient that there are relative differences in the production
costs in both countries. Trade between two countries can be beneficial for both
countries if each of them exports a good in the production of which he has a
comparative (relative) advantage;
. the theory of absolute costs formulated by Smith, the theory of international trade,
according to which the basis for the development of international specialization
and the source of benefits from international trade is the fact that there are absolute
differences between production costs and labor inputs. The benefits of exchange
mean an increase in the production capacity of both trade partners from interna-
tional exchange and explain the reasons why individual countries trade with each
other (Milewski 2003).
. the theory of distributive justice, the most general concept of the fair distribution
of Sena’s goods regarding equality and benefits (Kwarciński 2011) and
272 S. Konecka and A. Lupicka

. cost–benefit analysis.
Cost–benefit analysis (CBA) is perhaps the most useful in the context of using
intelligent solutions in transport. It is a method of assessing the effectiveness of invest-
ments and projects, taking into account all expected benefits and costs, including
quantitative and qualitative elements, allowing to determine the degree of invest-
ment effectiveness in a complex environment. The cost–benefit analysis also takes
into account social, cultural and environmental aspects, expressed as external costs.
CBA is particularly useful in assessing projects, the implementation of which is
associated with a large number of benefits and costs incurred by various groups of
stakeholders, and where the main selection criterion is not necessarily the maximiza-
tion of financial profit. The theoretical basis of the CBA is welfare economics (Becla
et al. 2012).
According to Paprocki (2017), the development of technology and technology, as
well as changes in the behaviour of participants in social and economic processes,
are factors causing the need to correct the theory describing the functioning of trans-
port. Changes in social and economic policy, as well as in climate, ecological and
energy policy, necessitate further research work devoted to the analysis of possible
and acceptable scenarios. Thus, the deliberate behavior of researchers is to indicate
possible system solutions together with a description of the anticipated benefits and
costs that may occur in the case of applying any of them.

4 Potential Benefits of Implementing Intelligent Transport

Solutions

The implementation of intelligent supply chain solutions has both potential bene-
fits and costs. They can be divided into three groups—social, environmental and
economic. They can be perceived differently by small companies and large logistics
operators, employers and employees. In the article, the authors will try to answer
research questions: Are intelligent supply chain tools perceived as a source of benefits
or costs in companies from the TFL industry?

4.1 Internet of Things

In the report “IoT in the Polish economy” (2019), prepared by the Working Group
for the IoT at the Ministry of Digital Affairs, it was estimated that in 2022 almost
USD 590 million will be allocated to IoT in transport. The Internet of Things is used
in many areas of logistics, including traffic management and transport fleet. Then
the telematics systems play the main role. The foundation of their functioning is the
skilful use of GPS, GSM, BLE and Wi-Fi technologies. The solutions based on them
offer the opportunity to improve both the efficiency and safety of road transport.
The Impact of the Use of Intelligent Supply Chain Tools … 273

They enable the collection and analysis of data on the efficiency and effectiveness
of traffic and transport fleet management in the enterprise, and thus, reducing the
costs of vehicle operation, detecting potential problems and more effective planning
of subsequent routes. IoT supports such processes as: property protection, employee
safety, control and optimization.
The Internet of Things is also used in supply chain management. For example, in
the case of shipment tracking, RFID tags connected to the data cloud, thanks to which
the only activity performed by employees is loading the goods. Also when it comes
to the course of delivery, the use of GPS and RFID technology allows to minimize
the risk of delays in transport resulting from unfavourable weather conditions or
poor road condition. Drivers are informed on an ongoing basis about the expected
difficulties, which gives them the opportunity to take action to prevent the extension
of deliveries. In addition, a network of sensors installed in trucks monitors their
technical parameters, such as load stability and tire pressure.
According to (Trzop 2020), the most important benefits of implementing IoT
solutions include:
. monitoring the state of material resources of the company and the status of
shipments as well as supervision over employees,
. measuring the effectiveness of the use of material and human resources of the
company,
. controlling the work of the company’s resources and influencing its course,
. automation of business processes,
. optimization and coordination of mutual cooperation of people, systems and
material resources,
. learning about areas that can be improved and applying best practices.

4.2 Big Data

In a global economy where data is expected to be used to make better decisions and
allocate resources faster, algorithms with large data sizes can perform tasks more effi-
ciently than humans, the transportation industry has a lot to gain, if only because it is
fragmented. Transport resources, senders, recipients, intermediaries form a network
that is difficult to handle effectively using manual processes. Transport generates
huge amounts of data that can be used to optimize its performance. The use of
telematics allows you to send information about the location and condition of vehi-
cles, TMS collect data on the content, quality and purpose of the load, beacons are
installed in the loads themselves, data is collected by toll collection devices, drivers’
mobile phones are integrated with electronic exchanges, electronic documents are
used, e.g., e-CMR, intelligent tachographs control the driver’s working time, driving
speed, data on the driving style of individual drivers, the degree of tire wear and fuel
are collected. These data, properly processed by algorithms, can contribute to a better
use of existing resources: matching loads to means of transport, using the capacity
274 S. Konecka and A. Lupicka

of semi-trailers and planning the length of routes. Especially when they cannot be
enlarged enough.
Possibilities resulting from the use of big data analytics include (Trzop 2020):
. comprehensive evaluation of the course of processes,
. process planning improvement,
. optimization of transport processes,
. improving risk prediction and management processes,
. the ability to define priorities in everyday tasks,
. improving customer relations.

4.3 Intelligent Transport Systems

The use of ITS is one of the priorities in the implementation of the city’s transport
policy. The most important effects of using ITS include:
. general improvement in the efficiency of the transport system operation (in
particular, shortening the travel time, reduction of the number of stops and road
incidents),
. increasing the level of road safety, and
. reduction of the emission of harmful exhaust components, dust and noise.
Regardless of the development of ITS solutions related to road infrastructure,
electronic systems for communicating vehicles with the environment, including other
vehicles, the so-called C-ITS systems (Cooperative ITS). As a development of this
concept, works are also carried out on Connected and Automated Driving vehicles,
which will cooperate with road ITS solutions in order to effectively use data, access
current information and respond to current needs and threats.
The concept of MaaS (Mobility as a Service) is also being developed, which
consists in providing a transport service with the use of various means of transport,
but without the need to have any of them. The goal of the ongoing development of
MaaS is to better respond to market demand by linking booking, purchasing and
payment systems in the transport chains and providing real-time information on
timetables, weather and traffic conditions, as well as available carrying capacity and
available solutions. MaaS is therefore the user digitized transport interface. Its aim
is also to optimize the use of transport capacity. This solution reduces travel costs
and air pollution, especially where it is very important—in cities. It also significantly
increases the use of means of transport, which under traditional solutions (vehicle
ownership) are used on average in about 5%. The vehicles are not used for the rest
of the day (Kamiński 2020).
Basically, the above-mentioned solutions concern the concept of digitization,
collecting large amounts of data and their use or sharing via the cloud and creating
IoT would not be possible without it. TMS and FMS class systems, online service
purchase platforms, ITS based on telematics can reduce the time-consuming admin-
istrative activities of drivers and office workers and are the result of digitization.
The Impact of the Use of Intelligent Supply Chain Tools … 275

There are several forms of transport digitization. Today, digital information is already
used in many ways in vehicles, including technologies and services that support car
driving, rail traffic control, aviation and vessel traffic management. The digitization
of passenger and cargo information is another area that is used on a daily basis.
Finally, robots are widely used in terminal operations in freight logistics.
The digitalisation of mobility of people and goods transport has many potential
benefits, such as
. ease of access and convenience for passengers,
. increasing efficiency, productivity and safety in freight transport,
. greater traffic safety, especially advanced automation increases transport safety
by reducing the frequency of human error.
. emission reduction.
The impact of transport on the climate and the environment depends on many
factors. Replacing fossil fuels with low carbon fuels, electricity or hydrogen is another
important way to reduce emissions. The implementation of electric vehicles and the
expansion of smart electric grids are closely related to transport automation. The
environmental impact is not only related to transport, but also to the life cycle of
vehicles. Repatriating production and implementing a circular economy and close
loop supply chains approach reduce the impact of the life cycle.

4.4 Artificial Intelligence

The area of application of AI in logistics is supply chain planning (SCP) systems

supported by machine learning. In the area of transport, AI solutions are also used.
An example is autonomous trucks that independently control their work and react to
external conditions, and even form integrated convoys (truck platooning).
When it comes to AI, the main benefits of its application in the area of logistics
include (Trzop 2020):
. increasing the efficiency of processes,
. shortening the time of performing tasks,
. streamlining the information flow process,
. improving the customer service process,
. the ability to prioritize tasks,
. avoiding empty runs: both in internal and external transport processes,
. optimization of the way the goods are distributed.
276 S. Konecka and A. Lupicka

4.5 Automation

The benefits of automation in transport processes include, among others, reducing

the shortage of employees. It is estimated that after 2025, autonomous vehicles will
reduce transport costs by up to 28% compared to 2016. By replacing labor costs
(drivers) with software costs, telematics and remote control. Most economic, social
or political and legal factors are likely to drive up costs in the industry, so autonomy
may be one of the few ways to reduce them. However, it will take 5–10 years for
technological solutions in this area (PwC 2017). As many as 78% of representatives
of global transport and logistics companies plan to take actions to automate tasks and
positions in order to ensure the effective achievement of goals. Kantar’s researchers
interviewed 1300 people over the age of 15 from Polish cities over 50,000. residents.
They asked, among others about the advantages of autonomous cars. Among the
greatest advantages, half of the respondents indicated greater mobility of people
who cannot drive a car, e.g., young people and the elderly. Another benefit is the
reduction of parking spaces (21% of responses) if the autonomous car returns to the
owner’s garage after reaching its destination or, as in the case of shared vehicles, will
pick up another person, he will not need a parking space in the city center. In addition,
the popularization of shared mobility may cause many people to give up driving their
own cars in favor of autonomous cars. 8% of city residents indicated that autonomous
vehicles will provide the opportunity to rest or work while traveling (2%). In any
case, they will allow more efficient use of travel time. Autonomous mobility can also
make car users healthier and calmer, as driving in a crowded city can be a source of
stress. Meanwhile, in the case of autonomous cars, the vehicle itself takes care of
everything. (www.transport-publiczny.pl)
The results of pilot research carried out among road transport operators in Greater
Poland in 2019 for the purposes of the thesis will be presented below (Różniewicz
2019). According to 58% of respondents, by 2030 autonomous cars will normally
functioning on Polish roads, only 16% were skeptical about such implementation
prospects. 76% of respondents declared their willingness to participate in courses
enabling better understanding of the topic of autonomous vehicles. 82% of respon-
dents are willing to replace trucks and vans in their own fleet with self-steering ones.
Among the benefits that the respondents saw in the introduction of autonomous cars,
cost optimization was 58% first. It was to consist of getting rid of the costs of drivers’
salaries, faster covering the route—without mandatory breaks in accordance with the
Act on the driver’s working time, timely delivery. 32% of respondents noticed the
benefit of being able to constantly monitor a vehicle, its location and, technical
parameters. Only 14% of respondents saw the benefit of eliminating a significant
number of accidents caused by drunk drivers.
Questions about the potential benefits of using autonomous vehicles as well as
threats were multiple-choice questions. Respondents perceived threats more often.
They considered the biggest threat to the lack of trust in the 36% system, they are
also afraid of software errors and the inability to respond to such obstacles from
vehicles. 64% of respondents indicated that autonomous vehicles were not adapted
The Impact of the Use of Intelligent Supply Chain Tools … 277

to the psychological aspects of other road users—for example, the behavior of drunk
pedestrians. Another threat is what has also been indicated as an advantage, i.e.,
reduced employment of drivers—44%, also in the context of retraining or acquiring
new skills, piloting an autonomous vehicle.
The number of technological solutions used in enterprises employing respon-
dents may also affect the perception of automation. The vast majority of respondents
declared that their enterprises use the most popular solutions, i.e., GPS (Global Posi-
tioning System)—satellite navigation system, working time controller and driving
speed controller. 23 people indicated that they worked with the help of cargo tracking,
and three people less—with a fuel consumption meter. The least respondents got
acquainted with the functioning of the tire pressure sensor and temperature sensor.
Awareness and ability to use individual tools also results in answering previous ques-
tions. People who worked with the fuel consumption meter, work time controller,
GPS indicated that they are in favor of the introduction of autonomous cars because
it will facilitate their work because the above devices will cease to need them.
96% of respondents would like to use the latest technologies at work. 60% of
respondents believe that autonomous transport is an expensive solution. In this issue,
the most important were the responses of employees who use the latest technologies at
work. They are the ones who have the greatest awareness and have broader knowledge
than others. After careful analysis, it was this group of people who fully opted for
autonomous transport. And 40% of those surveyed believe that an investment in
autonomous rolling stock will pay off and begin to bring benefits in a few years.
From a bibliometric analysis it follows that (Astarita et al. 2019; Bechtsis et al.
2018; Khondaker and Kattan 2015; Li et al. 2015; McGinley 2014; Sovacoola and
Axsenc 2018; Vleugel and Bal 2018; Yigitcanlar, Wilson and Kamruzzaman 2019):
. the future is bringing increasing demands for greater efficiency and for more
sustainable designs in cargo handling technologies;
. automated transport systems also play an important role in the creation of urban
mobility and have an important influence on the shaping of urban space;
. attention should also be paid to the safety of autonomous vehicle traffic (AV) as
part of both ITS and sharing economics;
. information security management is critical to protect the users of autonomous
vehicles;
. the potential of autonomous vehicles to implement sustainable development can
also be seen in the wider context of linking autonomous vehicles to sustainable
supply chain development.
278 S. Konecka and A. Lupicka

5 Potential Costs of Implementing Intelligent Transport

Solutions

One of the most frequently indicated potential social costs related primarily to
automation is the elimination of human labor. The change in the nature of human work
will certainly lead to the elimination of professions in transport, such as: driver, train
driver or operator of reloading equipment—crane, forklift. However, automation does
not have to lead to the elimination of work, as the dynamics of economic processes
must be taken into account. According to Götz and Gracel (2017), the effects of digi-
tization on the labor market should be positive on balance, as in Germany one-third
of enterprises subject to digitization processes plan to increase employment, and
only one-tenth expect a reduction in employment as a result of digitization. On the
other hand, OECD studies show that as a result of automation, about 12% of jobs in
Germany are at risk, and about 7% in Poland (Brandt 2016). Workstations where 70%
of work can be replaced by machines are considered such. In Great Britain this figure
reaches 10%, in the USA—9%, and in Japan—7%. Some studies even mention the
disappearance of more than 200 professions, new ones will surely emerge. According
to Paprocki (2017), just as computerization did not cause structural unemployment,
robotization does not have to cause this phenomenon.
However, the barriers to implementing the above-mentioned tools should be taken
into account: high investment costs, lack of experts, well-qualified and understanding
market needs, mentality, data security, fragmentation—95% of Polish transport
companies are micro-enterprises with a fleet of up to 20 vehicles. Such companies
have approx. 60% of the fleet resources in Poland.
Both in terms of individual users and enterprises, the development of intelli-
gent technologies in transport and the resulting costs should be viewed in terms of
the development of social media and the concept of Zuboff (2015) surveillance capi-
talism. And by analogy, consider the possibilities of extending it also to the discussed
solutions. Structural asymmetries of knowledge and rights made it impossible for
people and small organizations to learn about practices of leading tech companies.
In this concept “the new tools, networks, apps, platforms, and media thus became
requirements for social participation”. In the same way, the use of intelligent transport
tools becomes a prerequisite for the continued existence of transport companies. In
this new regime, a global architecture of computer mediation turns the electronic text
of the bounded organization into an intelligent world-spanning organism that Zuboff
(2015) call Big Other. One of the main challenges is the detailed recognition of a
new business model in which service providers, i.e., carriers of people and goods, are
completely cut off by virtual platform operators from their customers, i.e., passen-
gers and shippers. When analyzing this model, it is necessary to take into account
a specific scenario in which one operator occupies an unprecedented position in the
market, becoming a monopsony. In economic theory, the operation of monopsons
has so far been analyzed in a very narrow range. It is possible to emancipate virtual
platform operators, who will take over the role of architects of the entire mobility
support system for individuals and the entire society, as well as for the distribution
The Impact of the Use of Intelligent Supply Chain Tools … 279

system of things and goods devoid of material form. Nowadays, the source of the
leader’s advantage is the intellectual advantage obtained thanks to the use of the
so-called narrow AI. (Paprocki 2017) So, perhaps we should consider the scenario
of a specific exclusion of enterprises that do not efficiently use the most modern IT
solutions.
The need to introduce new types of digital platform-based business models is
linked to the need to create a supportive and enabling environment and, through a
regulatory framework, a level playing field for businesses. Market leaders (logis-
tics operators) are already implementing technological innovations, but there are no
regulatory provisions. When introducing intelligent IT tools, concerns are raised not
only about work, but also security, privacy and the environment.
When it comes to security, there is a growing concern about cybersecurity, which
will be one of the key aspects of transport security. It applies not only to vehicles,
but also to the infrastructure that supports, controls and manages them. The main
challenge may be to de facto accountability in the event of an accident, given the
role of digital systems and the involvement of several entities such as vehicle manu-
facturers and owners and infrastructure managers. New safety risks are due to the
limited ability of sensors to recognize shapes, malfunctions, internet disruptions, and
new types of human error such as software error. Of course, all these protections will
generate costs, but the final effect is still assessed as positive.
Concerning the need for privacy and greater data exchange, concerns are raised
about continuous monitoring, shape recognition. The management of non-personal
data, especially data generated by sensors and smart devices, also requires expla-
nation and regulation. When considering the issue of ease of access and re-use of
data, it is worth noting that generally it is not the data itself that gives a competitive
advantage, but rather it is improved by tools, innovation resources and market power.
Paradoxically, autonomous transport may lead to more use of private cars due to
greater passenger comfort. But car sharing (along with the use of autonomous public
transport) can reduce the number of private cars. Consumer preferences will play a
decisive role. They will depend on the possibility of travel planning, encouraging
passengers to make environmentally friendly choices, price incentives for the use of
electricity infrastructure. However, it should not be forgotten that new digital and
digitized infrastructure, new services are needed to enable access to traffic informa-
tion as well as for reservations and payments for mobile services and they require
electricity, the way to get it will be environmental impact (EKES 2017).
The implementation of intelligent IT solutions is associated with large financial
outlays, which not every enterprise, especially small or medium-sized, can decide
to do. In addition, their use is often associated with the need to entrust an external
entity with valuable information resources of the enterprise. There is also no doubt
that the implementation of such advanced technologies is associated with significant
changes in the organization of processes, requiring appropriate change management
skills from the managerial staff. However, it is clear that this process cannot succeed
if technology alone is driving progress. Ideally, development should be driven by
societal demand.
280 S. Konecka and A. Lupicka

6 Conclusions

Technology offers many opportunities, but progress cannot be based solely on it. It
requires an appropriate degree of availability, ease of access and free flow of data. At
the same time, adequate data protection must be ensured. In order to be able to respond
to new developments, it is also necessary to strengthen cybersecurity capacities and
to address liability issues. Close links with other policy areas such as digital single
market, energy, industrial development, innovation and skills policies. As climate
change mitigation objectives and requirements also contribute to the development of
digital transport, there is also a close link with environmental sustainability.
It is difficult to unequivocally identify and estimate the scale of benefits and the
scale of threats posed by the described IT solutions. The problem of the redistribution
of these benefits and costs remains open, both between enterprises in the country
and in enterprise networks and in the international dimension. A question arises—
Are these benefits and costs the same for all entities? It is worth considering the
dependence of perceiving the same tools as a cost or benefit depending on the type
of entity, e.g., its size in the transport industry.
Each of the analyzed solutions—digitization, autonomization, IoT, big data,
telematics, could be studied separately in a more detailed way. In the sciences of
management and quality, it is expected to continue work on dynamic changes in the
behavior of market participants. It is also possible to conduct research as part of the
so-called Uberization and amazonization taking place in the supply channels, they
adapt to the requirements imposed on them by entities gaining market dominance
thanks to the use of the latest digital technologies.
As part of megatrends, e.g., connected living or big data clouds, it is possible
to study the decision-making model at the enterprise level, which will have the
opportunity to achieve effects so far in theory considered achievable only on a macro
scale. The topics that require separate research are: electromobility, including vans
and trucks, sharing economy and instant pricing, blockchain and cybersecurity in
transport.
However, it is important to adequately address these structural changes by devel-
oping strategies to ensure a fair and efficient transition, reduce negative social impact
and respond to skills mismatches, with appropriate monitoring of progress, that is,
considering the achievements and consequences of intelligent IT transport solutions
from an end-to-end perspective. society, including enterprises, workers, consumers,
and looking at the balance of benefits and losses in both economic, social and
environmental terms.
The Impact of the Use of Intelligent Supply Chain Tools … 281

References

Astarita V, Giofrè VP, Guido G, Vitale A (2019) A single intersection cooperative-competitive

paradigm in real time traffic signal settings based on floating car data. Energies 12:409
Bechtsis D, Tsolakis N, Vlachos D, Srai JS (2018) Intelligent autonomous vehicles in digital supply
chains: a framework for integrating innovations towards sustainable value networks. J Clean Prod
181(20):60–71
Becla A, Czaja S, Zielińska A (2012) Analiza kosztów-korzyści w wycenie środowiska naturalnego,
Difin, Warszawa
Betham J (1958/1780) Wprowadzenie do zasad moralności i prawodawstwa, Państwowe
Wydawnictwo Naukowe, Warszawa
Brach J (2019) Kształtowanie si˛e Transportu 4.0 i Systemu Transportu 4.0 w kontekście wpływu
Rewolucji 4.0 na współczesny transport drogowy. Economics of the 21st Century 1(21):87–101
Brandt M (2016) Automatisierung bedroht Millionen Arbeitsplätze. http://ap-verlag.de/automatis
ierung-bedroht-millionen-arbeitsplaetze/23457/
Europejskiego Komitetu Ekonomiczno-Społecznego (EKES) (2017) Wpływ cyfryzacji i robotyzacji
transportu na kształtowanie polityki UE. TEN/632, Bruksela
Goliński M (2013) Informacja i techniki informacyjne jako przyczyna zmiany. Nierówności
Społeczne a Wzrost Gospodarczy 32:37–48
Götz M, Gracel J (2017) Przemysł czwartej generacji (Industry 4.0) – wyzwania dla badań w
kontekście mi˛edzynarodowym. Kwartalnik Naukowy Uczelni Vistula 1(15):217–235
Grupa robocza do spraw Internetu Rzeczy przy Ministerstwie Cyfryzacji (2019) IoT w polskiej
gospodarce
Homans G (1967) The nature of social science. New York
https://www.gov.pl/web/cyfryzacja/grupa-robocza-ds-internetu-rzeczy-internet-of-things-iot
Kaczmarek B (2013) Tworzenie organizacji inteligentnej jako nowej wartości firmy, Zeszyty
Naukowe Uniwersytetu Szczecińskiego, 786. Finanse, Rynki Finansowe, Ubezpieczenia
64(1):157–162
Kamiński T (2020) Kooperacyjne Inteligentne Systemy Transportowe (C-ITS) jako rozwi˛azania
podnosz˛ace bezpieczeństwo i efektywność transportu drogowego. Gospodarka Materiałowa i
Logistyka, LXXII 6:10–18
Kemal A, Delic KA, Dayal U (2002) The rise of the intelligent enterprise, Ubiquity, 6
Khondaker B, Kattan L (2015) Variable speed limit: a microscopic analysis in a connected vehicle
environment. Transp Res Part C Emerg Technol 58:146–159
Konecka S (2019) Intelligent supply chains from the perspective of benefits and costs for transport-
forwarding-logistics industry. Ekonomika i Organizacja Logistyki 4(4):49–59
Konecka S, Maryniak A (2020) Intelligent and innovative solutions in supply chains, advances in
intelligent systems and computing. Springer 1026:854–859
Kuhn TS (2009) Struktura rewolucji naukowych. Aletheia, Warszawa
Kwarciński T (2011) Równość i korzyść: Amartyi Kumar Sena: koncepcja sprawiedliwości
dystrybutywnej, Monografie: prace doktorskie, Uniwersytet Ekonomiczny w Krakowie, 11
Li Z, Chitturi MV, Yu L, Bill AR, Noyce DA (2015) Sustainability effects of next-generation
intersection control for autonomous vehicles, special issue on smart and sustainable transport,
15, 30(3):342–352
McGinley K (2014) Preparing port container terminals for the future: making the most of intelligent
transport systems (ITS), Urban Transport XX
Milewski R (2003) Podstawy ekonomii. PWN, Warszawa, p 585
Młody M, Weiner A (2020) The importance of technological anxiety for the digital transformation
of industrial processing companies in Poland. In: Contemporary organisation and management.
Challenges and trendsPublisher, Łódź University Press
Newell A, Simon HA (1972) Human problem solving. Prentice Hall, Englewood Cliffs
Olender-Skorek M (2017) Czwarta rewolucja przemysłowa a wybrane aspekty teorii ekonomii.
Nierówności Społeczne a Wzrost Gospodarczy 51:38–49
282 S. Konecka and A. Lupicka

Olson M (2012) Logika działania zbiorowego. Dobra publiczne i teoria grup, Wydawnictwo
Naukowe Scholar
Paprocki W (2017) Role of virtual platform operators in transforming consumer goods market. J
Manag Financ Sci XI(27):25–37
Penc J (2003) Zarz˛adzanie w warunkach globalizacji, Difin, Warszawa
PriceWaterhouseCoopers (2017) A decade of digital keeping pace with transformation, 2017 Global
Digital IQ® Survey: 10th anniversary edition. https://www.pwc.com/sk/en/publikacie/assets/
2017/pwc-digital-iq-report.pdf
Ricardo D, Hartwell RM (1971) On the principles of political economy, and taxation. Penguin
Books, Harmondsworth
Rifkin J (1995) The end of work. The decline of the global labor force and the dawn of the post-market
era, New York
Retrieved June 20, 2021, from https://www.transport-publiczny.pl/wiadomosci/jakie-korzysci-z-
auto-autonomicznych-dostrzegaja-polacy-61220.html
Retrieved July 30, 2021, from https://sjp.pwn.pl/szukaj/inteligencja.html
Romaniuk K (2017) Implikacje koopetycji dla organizacji inteligentnej w perspektywie teorii gier
i teorii kosztów transakcyjnych. Przegl˛ad Organizacji 7(930):11–18
Różniewicz E (2019) The prospect of autonomous transport development, praca dyplomowa pod
kierunkiem Konecka S., UEP, Poznań
Schumpeter JA (2008) Capitalism, socialism and democracy. Harper Collins Publ, UK
Sovacoola BK, Axsenc J (2018) Functional, symbolic and societal frames for automobility:
implications for sustainability transitions. Transp Res Part A 118:730–746
Trzop A (2020) Przegl˛ad rozwi˛azań z zakresu przemysłu 4.0 stosowanych w obszarze logistyki.
Organizacja i Zarz˛adzanie 81:233–247
Vleugel JM, Bal F (2018) More space and improved living conditions in cities with autonomous
vehicles. Int J Des Nat Ecodyn 12(4):505–515
Yigitcanlar T, Wilson M, Kamruzzaman M (2019) Disruptive impacts of automated driving systems
on the built environment and land use: an urban planner’s perspective. J Open Innov Technol Mark
Complex 5:24
Zuboff S (2015) Big other: surveillance capitalism and the prospects of an information civilization.
J Inf Technol 30:75–89
Zysińska M (2019) Ocena przedsi˛ebiorstw TSL działaj˛acych w Polsce – dylematy metodyczne
badań. Studia i Prace Kolegium Zarz˛adzania i Finansów 173:141–162