See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/351629432

Cost Optimization for Transportation Using Linear Programming

Chapter · May 2021

DOI: 10.1007/978-981-16-0976-3_2

CITATIONS READS

8 4,412

3 authors, including:

Divya Sharma S G
Amrita Vishwa Vidyapeetham
6 PUBLICATIONS 53 CITATIONS

SEE PROFILE

All content following this page was uploaded by Divya Sharma S G on 11 June 2021.

The user has requested enhancement of the downloaded file.

Cost Optimization for Transportation
Using Linear Programming

A. Vamsikrishna, Vigneshwar Raj, and S. G. Divya Sharma

1 Introduction

A company’s transportation services and methods are really important in building a

larger customer group and if an effective logistics or transportation model is designed
then it can make a really good turnover for the company. Jonsson [1] (a logistics
and supply chain expert) claim that 10–30% of turnover or profit can be made by
optimizing the direct cost section of the logistics department. In this present situation
where the population is increasing drastically and where the world is becoming a
global village, it is really important for a company to have a cost-efficient and time-
efficient transportation system which strives to produce its services to each and every
corner of the world.
Since transportation is a very important key factor which plays a major role in
the growth of a company, many logistics and supply chain experts have researched
various aspects of the transportation system. Ronald et al. [2] have generated a
genetic algorithm that considers the geographic information system (GIS) which
solves problems regarding the transportation routes. Dorer and Calisti [3] have
worked on the concept of adaptive transportation network which automatically opti-
mizes the existing transportation system accordingly. Authors of [4] have discussed
various methods and aspects of linear programming and its applications. In [5],
Chandrakantha has provided a solution for optimization problems using the EXCEL
solver option. Jonsson [1] have discussed logistics and supply chain management
and the part which deals with direct costing in logistics was really helpful in opti-
mizing the transportation system and reducing its cost. Caris et al. [6] have researched
intermodal freight transport systems and have provided solutions considering routes,

A. Vamsikrishna · V. Raj · S. G. Divya Sharma (B)

Department of Mechanical Engineering, Amrita School of Engineering, Amrita Vishwa
Vidyapeetham, Bengaluru, India
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 11
K. Jha et al. (eds.), Recent Advances in Sustainable Technologies,
Lecture Notes in Mechanical Engineering,
https://doi.org/10.1007/978-981-16-0976-3_2
12 A. Vamsikrishna et al.

climate, and the traffic parameters. Authors of [7] have mainly concentrated on
increasing the profits of a company by working on their transportation model using
linear programming method. Authors of [8] have worked on a transportation model
where mosquito coils have to be transported from warehouse to distributors and
they provided a cost-efficient transportation model using EXCEL solver function.
Tatavarthy and Sampangi [9] have worked on reverse supply chain and provided a
better transportation model by incorporating genetic algorithms on reverse supply
chain. Authors of [10] have worked on public transport fare and they have provided
a solution for this by using a simulation which increased the operator profit and
also benefited public transport commuters. Authors of [11] have focused on solving
problems regarding fuzzy transportations where the need for a number of prod-
ucts is uncertain and they provided a solution for fuzzy transportations problems by
generating an algorithm. Martínez-López et al. [12] focused on providing a motor-
ways system of transportation for shipping companies considering the environmental
conditions. Mathur et al. [13] have worked on providing a solution for fuzzy trans-
portation problems by generating algorithms. Authors of [14] have worked on public
transport networks and made a transportation model or design that used the ideology
of integrated optimization that improved urban mobility. Sharma and Tatavarthy [15]
have focused on working on a disaster management problem regarding how to opti-
mize the supply chain of necessary needs during a disaster and they achieved this
by surveying various papers. Mangala et al. [16] have discussed about operational
excellence for a sustainable supply chain. Also, they have mentioned that integra-
tion of various technologies like blockchain, big data, etc., could help create a more
efficient supply chain system.
In this paper, the cost of transportation was tried to be reduced to the cheapest
transportation model while taking all the constraints into consideration during the
process of designing the model. A model of a flavors and fragrance company had to
be optimized which consisted of three plants, three warehouses, and nine markets,
and the optimization of the transportation model was done using the EXCEL solver
function.

2 Problem

The network consisted of three manufacturing units or the plants, three warehouses,
and nine markets. The product flow was such that it was manufactured in the plants
and sent to the warehouses and then from the warehouses it was sent to the market.
Cost of transportation did not remain the same for every transport situation. The
existing model in company approximately costed about Rs. 3,800,000 and was
seeking for an improved model. Hence, out of all possible combinations, it was
necessary to find the cheapest transporting route with the optimal quantities to be
transported as it will benefit the company.
Cost Optimization for Transportation Using Linear Programming 13

Fig. 1 Methodology
flowchart

3 Methodology

Initially, the transportation costs for each of the cases were collected and tabulated.
Then the cost function or the target function was formulated for the given case. All
the existing constraints were listed out and then the target function was minimized
using EXCEL solver. Figure 1 shows a simple flowchart of the methodology adopted
in solving the problem.

4 Results and Discussions

Since it is necessary to understand actual product flow, a network diagram was drawn
as shown in Fig. 2.
In Table 1, the cost of distribution for various cases can be seen. P1, P2, and P3
stand for the three plants, W 1, W 2, and W 3 stand for the three warehouses, and M1
to M9 stand for the nine markets. The cell corresponding to a plant and a warehouse
denotes the transportation cost per unit from the plant to the warehouse while the
cell corresponding to a market and a warehouse denotes the transportation cost per
unit from the warehouse to the market. As mentioned above, this data was collected
from a flavors and fragrance manufacturing industry.
In Table 2, the variables to be determined can be viewed, i.e., number of units to
be transported. For example, x(P1, W 1) represents the no. of units to be transported
from the plant 1 to the warehouse 1 while x(W 3, M8) represents no. of units to be
transported from warehouse 3 to the market 8.
Further, the cost function was formulated. Cost function or the target function
signifies the total transportation cost. In this case, it is nothing but the summation of
the product of the corresponding cells of Tables 1 and 2, i.e.,

{6 × x(P1, W 1) + 5 × x(P2, W 1) + · · · + 3 × x(W 1, M8) + 2 × x(W 1, M9)

+ 3 × x(P1, W 2) + 4 × x(P2, W 2) + · · · + 4 × x(W 2, M8) + 5 × x(W 2, M9)
14 A. Vamsikrishna et al.

Fig. 2 Network diagram of

the product

Table 1 Distribution cost of the product (per unit) in Rs

Distribution costs per unit (Rs)
P1 P2 P3 M1 M2 M3 M4 M5 M6 M7 M8 M9
W1 6 5 1 3 8 4 7 9 4 3 3 2
W2 3 4 5 6 6 5 7 9 4 3 4 5
W3 4 3 2 5 8 2 4 4 6 4 4 6

+ 4 × x(P1, W 3) + 3 × x(P2, W 3) + · · · + 4 × x(W 3, M8)

+ 6 × x(W 3, M9)} = F(x)

The Min (F(x)), i.e., minimum of the function F(x) is to be determined. The values
of x(P1, W 1), x(P2, W 1), …, x(W 3, M8), x(W 3, M9) for which the minimum value
of F(x) can be obtained will be the required solution. But for finding the minimum
of F(x), it is necessary to formulate all the constraints. Hence, the company gave the
expected quantities from each market after performing the demand forecast. It can
be seen in Table 3.
It is also necessary to note that every plant has a manufacturing capacity. Plants
2 and 3 can manufacture maximum 100,000 and 150,000 units per annum while the
Table 2 No. of units to be transported
No of units to be transported
P1 P2 P3 M1 M2 M3 M4 M5 M6 M7 M8 M9
W1 x(P1, x(P2, x(P3, x(W 1, x(W 1, x(W 1, x(W 1, x(W 1, x(W 1, x(W 1, x(W 1, x(W 1, M9)
W 1) W 1) W 1) M1) M2) M3) M4) M5) M6) M7) M8)
W2 x(P1, x(P2, x(P3, x(W 2, x(W 2, x(W 2, x(W 2, x(W 2, x(W 2, x(W 2, x(W 2, x(W2, M9)
W 2) W 2) W 2) M1) M2) M3) M4) M5) M6) M7) M8)
W3 x(P1, x(P2, x(P3, x(W 3, x(W 3, x(W 3, x(W 3, x(W 3, x(W 3, x(W 3, x(W 3, x(W3, M9)
W 3) W 3) W 3) M1) M2) M3) M4) M5) M6) M7) M8)
Cost Optimization for Transportation Using Linear Programming
15
16 A. Vamsikrishna et al.

Table 3 Demand forecast

Forecast values
values for the markets per
annum M1 50,000
M2 40,000
M3 100,000
M4 80,000
M5 30,000
M6 35,000
M7 45,000
M8 70,000
M9 65,000

plant 1 can manufacture up to 800,000 units per annum as it is a very big plant.
It is also assumed that the number of units reaching the warehouse will leave the
warehouse and go to the market for the simplicity of calculation.
Considering the plants, it is known that the sum of all the quantities from the
plant must be less than or equal to its capacity. Hence, the formulation can be seen
as follows in Eqs. (1a)–(1c).

x(P1, W 1) + x(P1, W 2) + x(P1, W 3) ≤ 800,000 (1a)

x(P2, W 1) + x(P2, W 2) + x(P2, W 3) ≤ 100,000 (1b)

x(P3, W 1) + x(P3, W 2) + x(P3, W 3) ≤ 150,000 (1c)

With respect to the markets, the sum of all the quantities reaching the market must
be equal to the forecast values of Table 3. So, in total, nine equations are obtained,
one for each market as follows in Eqs. (2a)–(2i).

x(W 1, M1) + x(W 2, M1) + x(W 3, M1) = 50,000 (2a)

x(W 1, M2) + x(W 2, M2) + x(W 3, M2) = 40,000 (2b)

x(W 1, M3) + x(W 2, M3) + x(W 3, M3) = 100,000 (2c)

x(W 1, M4) + x(W 2, M4) + x(W 2, M4) = 80,000 (2d)

x(W 1, M5) + x(W 2, M5) + x(W 3, M5) = 30,000 (2e)

x(W 1, M6) + x(W 2, M6) + x(W 3, M5) = 35,000 (2f)

Cost Optimization for Transportation Using Linear Programming 17

x(W 1, M7) + x(W 2, M7) + x(W 2, M7) = 45,000 (2g)

x(W 1, M8) + x(W 2, M8) + x(W 3, M8) = 70,000 (2h)

x(W 1, M9) + x(W 2, M9) + x(W 3, M9) = 65,000 (2i)

It should be noted that the number of quantities entering the warehouse equals the
number of quantities leaving the warehouse. There are three warehouses and hence
three equations are formulated for the same as follows in Eqs. (3a)–(3c).

x(P1, W 1) + x(P2, W 1) + x(P3, W 1) = {x(W 1, M1) + x(W 1, M2)

+ x(W 1, M3) + x(W 1, M4)
+ x(W 1, M5) + x(W 1, M6)
+ x(W 1, M7) + x(W 1, M8)
+x(W 1, M9)} (3a)

x(P1, W 2) + x(P2, W 2) + x(P3, W 2) = {x(W 2, M1) + x(W 2, M2)

+ x(W 2, M3) + x(W 2, M4)
+ x(W 2, M5) + x(W 2, M6)
+ x(W 2, M7) + x(W 2, M8)
+x(W 2, M9)} (3b)

x(P1, W 3) + x(P2, W 3) + x(P3, W 3) = {x(W 3, M1) + x(W 3, M2)

+ x(W 3, M3) + x(W 3, M4)
+ x(W 3, M5) + x(W 3, M6)
+ x(W 3, M7) + x(W 3, M8)
+x(W 3, M9)} (3c)

After defining all the constraints and the target functions, these were calculated in
the EXCEL spreadsheet using existing functions like SUM, AVERAGE, etc. Further,
these equations were solved using EXCEL solver by defining all the necessary inputs.
Figure 3 shows the dialog box for EXCEL solver.
The cell that corresponds to the target function should be set as the objective.
Since this is a cost function and is to be minimized, the radio button that says ‘Min’
was selected. As defined earlier, the LPP has a lot of constraints. These constraints
need to be added in the ‘Subject to the Constraints’ box. Solving method should be
chosen as ‘Simplex LP’. After setting all these, the solve button was clicked.
18 A. Vamsikrishna et al.

Fig. 3 Solver parameter dialog box for solving the LPP

After the solutions are fount to converge, the values for optimal distribution quan-
tity is obtained for the transportation from each plant to each warehouse and from
each warehouse to each market as shown in Table 4.
For the above condition, the transportation cost estimation was Rs. 3,035,000.
Previously as stated by the company, the cost was approximately Rs. 3,800,000.
Therefore, savings of about Rs. 765,000 per annum was calculated if the new
transportation model was adopted.

5 Conclusion

A transportation optimization model was developed. It consisted of three plants, three

warehouses, and nine markets and was optimized using the EXCEL solver function
to design the transportation model. The transportation cost was reduced in the best
possible way by using linear programming. The cost obtained was Rs. 3,035,000. A
savings of about Rs. 765,000 per annum was estimated comparing the cost of new
Table 4 Optimal distribution strategy
Optimal distribution strategy (No. of units)
P1 P2 P3 M1 M2 M3 M4 M5 M6 M7 M8 M9
W1 0 0 150,000 50,000 0 0 0 0 0 0 35,000 65,000
W2 155,000 0 0 0 40,000 0 0 0 35,000 45,000 35,000 0
W3 110,000 100,000 0 0 0 100,000 80,000 30,000 0 0 0 0
Cost Optimization for Transportation Using Linear Programming
19
 20 A. Vamsikrishna et al.

transportation model to that of the previous one. This paper further creates scope for
research on designing transportation models and solving them with the help of linear
programming method which simplifies the cost optimization technique. However, the
scope of the work extends to consider multivariable problems and use EXCEL solver
to consider different optimization techniques to solve multiple problems related to
transportation.

References

1. Jonsson P (2008) Logistics and supply chain management. McGraw-Hill, New York
2. Ronald S, Kirkby S, Eklund P (1997) Multi-chromosome mixed encodings for heterogeneous
problems. In: Proceedings of 1997 IEEE international conference on evolutionary computation
(ICEC’ 97). IEEE, pp 37–42
3. Dorer K, Calisti M (2005) An adaptive solution to dynamic transport optimization. In: Proceed-
ings of the fourth international joint conference on autonomous agents and multiagent systems,
pp 45–51
4. Eiselt HA, Sandblom CL (2007) Linear programming and its applications. Springer Science
& Business Media, Berlin
5. Chandrakantha L (2008) Using EXCEL solver in optimization problems. Mathematics and
Computer Science Department, New York
6. Caris A, Macharis C, Janssens GK (2008) Planning problems in intermodal freight transport:
accomplishments and prospects. Transp Plan Technol 31(3):277–302
7. Fagoyinbo IS, Akinbo RY, Ajibode IA, Olaniran YOA (2011) Maximization of profit in manu-
facturing industries using linear programming techniques: Geepee Nigeria Limited. Mediterr
J Soc Sci 2(6):97–97
8. Khan MA (2014) Using linear programming transportation cost optimization. In: International
conference on mechanical, industrial and energy engineering, pp 2241–2245
9. Tatavarthy SR, Sampangi G (2015) Solving a reverse supply chain TSP by genetic algorithm.
In: Applied mechanics and materials, vol 813. Trans Tech Publications Ltd., Switzerland, pp
1203–1207
10. Kaddoura I, Kickhöfer B, Neumann A, Tirachini A (2015) Agent-based optimisation of
public transport supply and pricing: impacts of activity scheduling decisions and simulation
randomness. Transportation 42(6):1039–1061
11. Chandran S, Kandaswamy G (2016) A fuzzy approach to transport optimization problem.
Optim Eng 17(4):965–980
12. Martínez-López A, Sobrino PC, González MM (2016) Influence of external costs on the opti-
misation of container fleets by operating under motorways of the sea conditions. Int J Shipping
Transp Logistics 8(6):653–686
13. Mathur N, Srivastava PK, Paul A (2018) Algorithms for solving fuzzy transportation problem.
Int J Math Oper Res 12(2):190–219
14. Fuchs F, Corman F (2019) An open toolbox for integrated optimization of public transport.
In: 2019 6th international conference on models and technologies for intelligent transportation
systems (MT-ITS). IEEE, pp 1–7
15. Sharma SD, Tatavarthy SR (2019) Development of optimized solution for a generic disaster
management problem through construction of responsive supply chain a review. In: AIP
conference proceedings, vol 2148, no 1. AIP Publishing LLC, USA, p 030049
16. Mangla SK, Kusi-Sarpong S, Luthra S, Bai C, Jakhar SK, Khan SA (2020) Operational
excellence for improving sustainable supply chain performance. Resour Conserv Recycl
162:105025

View publication stats