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

net/publication/325924502

Knowledge dynamics-integrated map as a blueprint for system development:

Applications to safety risk management in Wuhan metro project

Article  in  Automation in Construction · September 2018

DOI: 10.1016/j.autcon.2018.05.014

CITATIONS READS

0 114

5 authors, including:

Chao Dong Fan Wang

Huazhong University of Science and Technology The Hong Kong Polytechnic University
2 PUBLICATIONS   0 CITATIONS    20 PUBLICATIONS   82 CITATIONS   

SEE PROFILE SEE PROFILE

Heng Li Hanbin Luo

The Hong Kong Polytechnic University Huazhong University of Science and Technology
396 PUBLICATIONS   9,168 CITATIONS    93 PUBLICATIONS   645 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

System reliability analysis of shield tunneling under incomplete probability information View project

Megaproject management View project

All content following this page was uploaded by Chao Dong on 22 June 2018.

The user has requested enhancement of the downloaded file.

 Automation in Construction 93 (2018) 112–122

Contents lists available at ScienceDirect

Automation in Construction
journal homepage: www.elsevier.com/locate/autcon

Knowledge dynamics-integrated map as a blueprint for system development: T

Applications to safety risk management in Wuhan metro project
⁎
Chao Donga,b,c, Fan Wanga,c,d, , Heng Lic, Lieyun Dinga,b, Hanbin Luoa,b
a
School of Civil Engineering & Mechanics, Huazhong University of Science & Technology, Wuhan, China
b
Hubei Engineering Research Center for Virtual, Safe and Automated Construction (ViSAC), HUST, China
c
Department of Building and Real Estate, The Hong Kong Polytechnic University, Hong Kong
d
School of Civil Engineering and Architecture, Wuhan Institute of Technology, Wuhan, China

A R T I C LE I N FO A B S T R A C T

Keywords: Safety has always been a persistent problem in the construction industry, particularly for tunnel construction
Safety knowledge projects due to the inherent uncertainty in geotechnical conditions and the complexity of the tunnel construction
Knowledge map process. Literatures and practices highlight the importance of safety knowledge to the construction safety.
Knowledge flow dynamics However, safety knowledge is not always provided in a convenient and timely manner. This paper develops a
Knowledge management
Knowledge-dynamics Integrated Map (KIM) to visualize safety knowledge flow in tunnel construction safety risk
Tunnel construction project
management. The KIM highlights the what, the who, and the why of knowledge flow by portraying its dynamics
associated to the working processes. With the use of KIM, the safety knowledge flow barriers are identified so
that knowledge flow facilitators can be designed. The KIM has been applied in Wuhan metro project as a
blueprint for systematically developing effective safety knowledge management systems.

1. Introduction To overcome this barrier, knowledge management serves as an ap-

proach to leverage and harness knowledge in most organizations
Statistical data on accident rates associated with tunneling projects [6–10]. There are several definitions of knowledge management in
indicated the high risks involved [1]. Risk management has been different fields [7,11]. In construction industry, knowledge manage-
adopted in the majority of tunneling projects through identifying, as- ment aims to effectively and systematically transfer and share knowl-
sessing, analyzing and controlling project risks. Successful im- edge and experiences among engineers, through which the learned
plementation of risk management requires substantial practical ex- knowledge from previous projects can be applied to solve problems in
perience and sound theoretical knowledge to perform various risk new projects, which effectively encourages organizational learning
management activities such as site inspection, risk prioritization, and through continuous feedback and knowledge updates based on in-
expert consultation [2]. However, due to the high staff turnover rates formation and communication technologies [12,13].
and frequent reassignment of experts and engineers, the project safety Despite many efforts have been put into knowledge management for
risk-related knowledge cannot be made full reuse among different construction organizations, barriers still exist in capturing and reusing
projects [3,4]. The communication of construction project risks within safety knowledge to perform safety risk management activities, which
an organization and among stakeholders also tends to be poor, in- may result in omissions or violations where accidents are originated
complete and inconsistent [5]. This situation has become a major bar- and incubated. Therefore, a method for identifying and overcoming the
rier for the success of tunneling projects. As noted by Cárdenas et al. barriers in knowledge sharing within safety risk management frame-
[3], “failures in ground-related construction projects are mainly due to work is required. This paper applied the knowledge dynamics-in-
shortcomings in the use of available and relevant knowledge rather tegrated map (KIM) to visualize the safety knowledge and their re-
than the result of uncertainty due to unknown factors. These short- lationships in the safety risk management of Wuhan metro project. The
comings reflected situations in which tacit or explicit information, such KIM is used to clarify the fluidity of the knowledge through the risk
as technological knowledge, design assumptions, monitoring records, management process and aid the identification of knowledge flow
thresholds and tolerances, and risks, was ignored, improperly used, barriers. The aim of the paper is to answer the following questions:
rejected, or not passed on by someone in the project.” (p. 92) What safety knowledge is required to perform a risk management

⁎
Corresponding author at: Department of Building and Real Estate, The Hong Kong Polytechnic University, Hong Kong.
E-mail addresses: [email protected] (C. Dong), [email protected] (F. Wang), [email protected] (H. Li), [email protected] (L. Ding), [email protected] (H. Luo).

https://doi.org/10.1016/j.autcon.2018.05.014
Received 23 October 2017; Received in revised form 13 March 2018; Accepted 9 May 2018
0926-5805/ © 2018 Elsevier B.V. All rights reserved.
C. Dong et al. Automation in Construction 93 (2018) 112–122

activity? Why is the desired safety knowledge difficult to acquire? How advantage of the visualization and navigation in locating and pub-
can these difficulties be overcome? The proposed method was applied lishing knowledge, knowledge maps have been employed for various
to the Wuhan metro project to reveal the barriers that hinder safety applications [32,33].
improvements; several systems were designed to capture, share and Ontology is a formal, explicit representation of a shared con-
reuse safety knowledge and facilitate safety knowledge flows. ceptualization, representing knowledge as a set of concepts within a
domain and the relationships between those concepts [34]. In the
2. Related studies construction safety research, ontology has been adopted in safety
knowledge representation, sharing, exchanging, and reusing [35]. Le
2.1. Safety knowledge et al. [36] integrated ontology with social network to enhance a better
communication of construction safety knowledge among all stake-
As accidents (re)occur during tunneling, accident causation models holders of construction projects and organizations. Other researchers
have been proposed to analyze the contributors to accidents. The most have employed ontology to model and reuse the knowledge on con-
notable one is Reason's Swiss cheese model [14], in which the safety struction safety rules for automated construction safety checking [35],
culture is believed to contribute to the potential accident trajectory as a job hazard analysis [22,37,38] and safety risk analysis [39]. As on-
root cause. An effective or positive safety culture can reduce the risk of tology is convenient in representing knowledge that is embedded in the
accidents in the system [15]. According to Reason [16], a safety culture actual practices (i.e. the cases) in an organization, ontology-based
is an informed culture that by definition is “one in which those who knowledge maps have been built for managing the knowledge in-
manage and operate the system has current knowledge about the ventory of a manufacturing company [31], assisting knowledge process
human, technical, organizational and environmental factors that de- outsourcing decisions for the healthcare organization [40], improving
termine the safety of the system as a whole” (p. 294). From the per- the business process re-engineering efficiency [41], and improving the
spective of knowledge, the lack of safety knowledge is perceived to efficiency of knowledge sharing and exchanging in a virtual environ-
contribute to accidents. It can thus be inferred that effective safety ment [42]. Thus, it is necessary to embed the ontology dimension into
knowledge management has a positive relationship with a “well-in- the knowledge map for knowledge representation.
formed” safety culture [17]. Bearing this in mind, it is argued that For the knowledge-intensive nature of the construction industry,
tunnel construction safety can be improved by assisting practitioners to several researchers applied knowledge maps to capture and represent
acquire the needed safety knowledge, through identifying the safety construction project knowledge [43,44], or to assist knowledge transfer
knowledge and facilitate the knowledge flow in the safety risk man- and reuse [26,45]. However, the majority of knowledge maps are static,
agement workflow. that is, the knowledge flow dynamics are not presented in the map,
Knowledge is defined as a justified belief that increases an entity's which may hinder the development of effective knowledge-based sys-
capability for effective action [18], while safety is defined as the control tems. For example, the knowledge required to assess and control ex-
of recognized hazards to attain an acceptable level of risk [19]. Thus, cessive tunneling-induced settlement can be acquired through two
safety knowledge can be defined as a justified belief that increases an methods: sharing among colleagues or learning from a training course.
entity's capability to effectively control the recognized hazards to attain The knowledge is created from tacit to tacit through socialization in the
an acceptable level of risk. Several researchers have addressed the first scenario whereas the knowledge is created from explicit to tacit
importance of safety knowledge to improve construction safety through internalization in the second scenario. However, classical
[20–23]. The capture, dissemination and reuse of safety knowledge knowledge maps cannot distinguish the knowledge type nor the
provide continuous feedback of what was learned and implements the knowledge creation mode in the flow of knowledge (refer to Fig. 1) and
lessons learned [24,25], which enhances the ability to perceive, re- may overlook the need for different knowledge management strategies.
cognize and control hazards. However, tunnel construction is a For the former case, regular meetings would be effective, whereas an
knowledge-rich industry in terms of the knowledge it creates, the ex- online technical course system or e-learning system may be appropriate
change among participants, and the information absorbed from ex- for the latter case. Therefore, the knowledge flow model should be in-
ternal sources [26]. Without an overview of the required safety tegrated into knowledge maps to provide insights into the knowledge
knowledge and the associated relationships in the risk management flow dynamics and constraints.
activity, their use can be confusing. This gap necessitates the use of
knowledge maps as a knowledge management tool, which reveals 2.3. Knowledge flow model
knowledge sources as well as their underlying relationships using a map
metaphor for visualization [27]. Knowledge can be generally classified into two types: tacit and ex-
plicit [46,47]. Tacit knowledge refers to personal context-specific
2.2. Knowledge map knowledge that resides within people's minds, and is difficult to for-
malize, express or articulate to another person. Tacit knowledge, such
A knowledge map is defined as a diagrammatic and graphic re- as the “know-how” of experienced staff, can be obtained from inter-
presentation of knowledge, which portrays the sources, flows, con- views with experts or managers. Therefore, tacit safety knowledge is a
straints and terminations of knowledge within an organization [13]. personal hard-to-reach justified belief, that increases his or her cap-
Using nodes and links, knowledge maps enable the visualization of the ability of effectively and efficiently control the recognized hazards to
conception and correlation of knowledge, which can be easily ex- attain an acceptable level of risk. In contrast, explicit knowledge is the
amined, refined and shared by non-expert knowledge users [28]. Be- knowledge that can be formalized and stored, including standard op-
cause knowledge maps typically highlight people as well as documents erating procedures, best practice guides, reports and manuals [7,48].
and repositories, they are frequently employed to describe the knowl- Knowledge creation and transfer is a spiral process of interactions be-
edge that is required in a risk management process, who owns what or tween tacit knowledge and explicit knowledge through socialization
where it resides, which supports the acquisition, connecting, and (tacit to tacit), externalization (tacit to explicit), combination (explicit
sharing of knowledge in a convenient manner [29]. The knowledge to explicit), and internalization (explicit to tacit). The knowledge-con-
map serves as an important role in implementing knowledge manage- version creates the organizational knowledge through the trajectory,
ment not only because it provides access to desired knowledge and from the individual level to the group level, and finally to the organi-
demonstrates directions of contents in a timely manner, but also due to zational or even inter-organizational level [18]. Although this knowl-
the simplicity of the notation that enables users to interpret and im- edge flow model has ontological dimension, which facilitates the
prove the understanding of the knowledge [30,31]. Due to the knowledge representation in the maps, Nissen and Levitt [49,50]

113
C. Dong et al. Automation in Construction 93 (2018) 112–122

supportive knowledge,
specification, experience
from former projects, etc.

knowledge of controlling
excessive settlement

Fig. 1. Illustration of different knowledge flow dynamics that are represented by the same knowledge map.

argued that the spiral model remains static, which tries to capture the tacit knowledge stored in an individual's mind, or the reports and
knowledge flow dynamics by snapshots. A third dimension was added drawings store in the organization. The “know-what” and “know-who”
to extend the spiral model; thus, the representation of knowledge flow is clearly shown in this plane.
dynamics was subsequently extended to three dimensions: explicitness,
organizational reach and life cycle. The three-dimensional representa-
3.1.2. Extension to the life cycle dimension
tion enabled the visualization of a diversity of patterns of enterprise
The links between different types of knowledge and different levels
knowledge flows in a discernable way, because it defines a vector space
indicate the knowledge flow dynamics, which yields the need to specify
and enables the plotting of dynamic trajectories for each knowledge
the life cycle of each process. Thus, the “know-why” is reflected by
flow.
explaining how the safety knowledge is obtained and shared. This paper
It is unlikely to design effective flow-enhancing interventions until
adapted representation of knowledge life cycle proposed by Rodríguez-
the phenomenon of knowledge flow is understood [49]. Therefore, it is
Elias et al. [54], because they provided a straightforward connection to
important to adopt the knowledge flow model to delineate the safety
the spiral knowledge creation model [18]. Therefore, seven stages i.e.,
knowledge flow in tunnel construction, in this way, the knowledge flow
acquire, apply, generate, organize, store, retrieve and share, were
barriers can be identified, and appropriate strategies to enhance the
identified. In acquisition stage, knowledge is captured by someone or
knowledge flow can be designed, finally the safety knowledge “well-
within an organization. By applying the acquired knowledge, new
informed” culture can be achieved.
knowledge may be generated or created through comprehension or
cognition. The new knowledge can be disseminated both informally
3. Knowledge-dynamics integrated map (KIM) (shared by personal contact, namely socialization) and formally (or-
ganized and stored in the knowledge base and subsequently retrieved
This section proposes the basics of knowledge-dynamics integrated when needed, namely combination). Externalization and internaliza-
map. The KIM was specifically designed for tunnel construction safety tion are used to group the stages that are involved in the conversion of
risk management. The major difference between the proposed KIM and tacit knowledge into explicit knowledge, and vice versa. Note that
extant knowledge maps is the representation that characterizes the iterations and feedback loops may exist between stages and some of
dynamic nature of knowledge flow. these stages may exist simultaneously [49,55].

3.1. Associating knowledge map with knowledge flow dynamics 3.2. Graphical representation and developing procedures of KIM

Effective knowledge maps consider the who, the what and the why The 3D representation is not convenient for the map development,
of the knowledge mapping process [7,51]. Classic knowledge maps use we modified the classic knowledge map components to show the 3D
nodes to represent the concepts of knowledge, and links for de- KIM in a two-dimensional manner. As shown in Fig. 3, KIM consists of
pendencies between knowledge. To portray the knowledge flow dy- three components:
namics, we extend the knowledge map to three dimensions, as shown in (1) Node: Rectangular object denotes safety knowledge. The blank
Fig. 2. nodes and shadowed nodes represent explicit knowledge and tacit
knowledge, respectively;
3.1.1. The explicitness-organizational reach dimensions [52] (2) Symbol: Three symbols are used to illustrate the organizational
This paper follows Nissen et al. [53] who use “explicitness” and reach;
“organizational reach” to represent the two dimensions originally la- (3) Link: The nodes and symbols are connected by arrows, and la-
beled “epistemological” and “ontological” [18]. The explicitness in- beled with the stages of the life cycle.
dicates the type of safety knowledge that is applied to perform a risk Fig. 4 displays a KIM representing the same case mentioned in
management activity, whether explicit or tacit. The organizational Section 2.2, which is to assess the risks of excessive settlement. In one
reach indicates who should possess the required safety knowledge to case, the supportive knowledge (e.g., standards etc.) or experience was
conduct a risk management activity or where to find the required safety stored in the organization as documents or manuals that can be re-
knowledge. In this paper, organizational reach contains three levels, trieved and acquired by the users. The knowledge for controlling ex-
i.e., the individual level, the group level and the organizational level. cessive settlement was subsequently generated by absorbing the sup-
Therefore, the explicitness-organizational reach plane shows the portive knowledge and adapting it to the new project. In the other case,
knowledge type, who possess it and stores in which form. Such as the the same supportive knowledge or experience is transferred among the

114
C. Dong et al. Automation in Construction 93 (2018) 112–122

K6
explicit
linkage
K5
node

K1

K4 K2

tacit
K3

individual group organizational

te
ra

ext
ne

ern
ge

aliz
sh l y

ati
e
p

on
ar
e

socialization
ire p
iz
qu a
n
a
e g

ac
v or

combination
ie

intern
tr

alizat
re
re

ion
sto

Fig. 2. Association of the knowledge map components with the dimensions of knowledge flow dynamics.

engineers, e.g., discussion in a business meeting. The two cases imply 3.3. Procedures for KIM development and application
two kinds of dynamics in knowledge flow. In the former case, the
knowledge flow from the organizational level to the individual level Fig. 5 presents the five steps of building a KIM for system devel-
and convert explicit knowledge to tacit knowledge, namely knowledge opment:
internalization. In the latter case, the knowledge flows from group level (1) Workflow identification: the analysis of workflow activities and
to individual level and the socialization is manifested in the tacit to tacit knowledge requirements is important because all requisite knowledge
conversion. must be integrated prior to the completion of knowledge-intensive work
[56]. Therefore, the first step is to identify the risk management process
of tunnel construction. This step can be achieved by interviewing and

Fig. 3. Basic components of a KIM.

115
C. Dong et al. Automation in Construction 93 (2018) 112–122

Organizational level

knowledge
Risk group of engineers
base Group level

organize
retrieve
assessment Explicit
knowledge
supportive knowledge, supportive knowledge,
specification, experience specification, experience
from former projects, etc. from former projects, etc.

acquire

share
Individual level
Stages of
lifecycle
engineer engineer

generate

generate
Tacit
Risk knowledge
control knowledge of controlling knowledge of controlling
excessive settlement excessive settlement

Fig. 4. Examples of representing different knowledge flow dynamics using KIM.

Identify risk identified. Safety knowledge can be linked based on their dependencies
management process of to create a classic knowledge map.
tunnel construction (3) Map adaptation: the classic knowledge map is modified in terms
of the type of the safety knowledge, the level in which the safety
knowledge resides, and the phases of the life cycle of the safety
knowledge flow. However, by associating the safety knowledge with a
specific person (group, or organization), additional safety knowledge
Identify safety knowledge that the person uses to perform a risk management activity may be
and construct classic identified, which results in the need to link the new knowledge with the
knowledge maps existent knowledge in the map. The new linkage between safety
KIM knowledge induces modification of the map. Therefore, the KIM can be
development KIM
constructed by iteratively implementing Step 2 and Step 3 until a
adaptation
complete map that incorporates all required safety knowledge is
Associate the knowledge achieved.
map with three dimensions (4) Knowledge flow barriers analysis: safety knowledge flow bar-
of knowledge flow dynamics
riers can impede or delay the provision of knowledge relevant to po-
tential risks to the users. The key to designing effective knowledge
management strategies lies in the understanding of the knowledge flow
barriers. Lin [57,58] categorized knowledge flow barriers into five
Analyze the knowledge flow types: knowledge source barriers (e.g., senior engineers may be un-
processes and identify the willing to share their safety knowledge because of inadequate rewards
knowledge flow barriers
KIM or the fear of weakening their competitive advantage), knowledge re-
KIM updating ceiver barriers (e.g., workers with insufficient education may have
application difficulty in absorbing and transforming some safety knowledge),
knowledge characteristics barriers (e.g., the uncertainties of the geo-
Design strategies for
technical conditions can produce ambiguity of the knowledge), con-
eliminating each
knowledge flow barrier textual barriers (e.g., workers are not adequately trained and a high
staff turnover rate), and inadequate/lack of mechanisms (e.g., blocked
channel between knowledge providers and knowledge seekers).
(5) Systems development: by identifying the knowledge flow bar-
Fig. 5. The procedures for applying KIM. riers, appropriate systems can be designed to overcome these barriers.
For example, an insufficient mechanism can be largely overcome by IT-
surveying the stakeholders about the activities they conduct. A diagram centric strategies, such as safety data warehouse. Note adaptations of
of risk management process can be created based on the sequence of the the developed systems may be needed during their applications due to
activities and the participants involved. new knowledge requirements. Then the KIM needs to be reviewed and
(2) Knowledge elicitation: based on the risk management process, updated by repeating the previous steps.
the safety knowledge required to implement a specific activity can be

116
C. Dong et al. Automation in Construction 93 (2018) 112–122

Wuhan metro safety early-warning system

Legend:
Metro line
completed
construction

Metro line under

construction

Fig. 6. Layout of Wuhan metro network.

4. Empirical evidence from the Wuhan metro project In the construction phase, the focus is to minimize the risks through
the planning and implementation of risk eliminating and/or risk miti-
The Wuhan metro project started in early 2000s. A total number of gating initiatives [2]. Various participants are involved in this phase.
25 lines (1045 km) will be operational before 2049. Up to now, 8 lines The working process contains monitoring, assessing, analyzing, and
(237 km) have been completed and in operation. Other 8 lines (168 km) controlling risks. The field measurements, were taken by contractors
are under construction and will be completed by the year 2021. The and third-party monitors. The inspections or audits of safety weak-
layout of the lines is shown in Fig. 6. Most of the alignment of the nesses, such as cracks on the surface of the structures or unsafe beha-
tunnels was arranged in the densely populated area to provide con- viors, are conducted by the consultants to supplement the safety related
venient access for the public. Thus, tunnel construction may cause da- information. Based on field measurements, inspections and audits, de-
mage to adjacent structures, facilities and pipelines, and this is the ductive analysis can be adopted to judge whether the risks identified in
prime risk to be managed through collaborations among stakeholders. the pre-construction phase are incubated and new risks are incurred.
The main stakeholders of the lines currently under construction include The results can support the contractors, supervisors and consultants to
the owner (Wuhan metro group), 12 contractors, 5 safety auditors, and make risk mitigation suggestions. If the risk level is negligible, no fur-
4 third-party monitors. The associated risks and multiple stakeholders ther consideration of the risk is required. If the risk level is acceptable
imply the complexity of safety management. or unwanted, the risks need be addressed by the contractors with the
As the appointed consultant, we have access to relevant materials owner's approval of the control measures. If the risk level is un-
and directly participated in the whole process of risk management of acceptable, the owner invites domain experts to diagnose the problems
Wuhan metro project. This enabled us to implement the KIM framework and recommend preventive measures. After addressing the risks, the
and directly evaluate its effectiveness. During this process, the safety identified risks will be updated in terms of risk planning and the control
knowledge flow barriers are revealed using the KIM. Strategies have effects were monitored with feedback to the relevant participants.
been designed and implemented for the Wuhan metro lines as safety
knowledge flow facilitators. In the following subsections, the Wuhan 4.2. Safety knowledge identification and KIM construction
metro Line 2, which has been fully completed and successfully operated
for 5 years, was adopted to illustrate the implementation. Each risk management activity requires specific safety knowledge.
The granularity of a knowledge map should contain fewer nodes while
4.1. Identify risk management working process providing the functionality for which it was intended [32]. This study
constructed the KIM in the granularity of directly applied safety
In the pre-construction phase, the focus of risk management is to knowledge. The supportive knowledge that underpinned the applied
identify potential hazards posed to the construction activities, which is safety knowledge was not presented in the map with respect to parsi-
largely dependent on the experts' experiences [2]. As there are no site mony.
activities in this phase, risk identification relies mainly on the deductive A total of 21 people (5 domain experts, 14 engineers, and 2 site
inference based on the comparison between project profiles (project managers) with at least three years of related experience were invited
plan, design drawings, specifications, and guidelines) and previous si- to list the specifications they refer to, the reports they provide, and any
milar projects. Therefore, the working process in this phase begins with related safety knowledge they used to complete each activity. The
the selection of experienced experts, followed by analyzing risks, managers were further interviewed to describe the way they conduct
prioritizing risks, and formulating risk report. these activities. Also, the manuals and policies about working

117
C. Dong et al. Automation in Construction 93 (2018) 112–122

Acquire Approval of
Select experts/ Analysis of the Compile risk
Working process relevant
engineers associated risk
the risk
report
materials prioritization

acquire generate acquire

- knowledge of - suggestions - decision theory

e
ev
tri
Project manager who-knows-what - recommendations Project manager - project management

re

e
ar

r
sh

fe
generate

ns
tra
ze
ani
- experts/engineers list org
Knowledge - specifications of risk
org
ani - risk priority
- revised report ze
dynamics map - risk response plan
analysis

sto
Knowledge

re
sh
Group of

ar
base transfer experts/engineers

e
generate
re
sto Knowledge
- suggestions base
- Project plan - recommendations
- knowledge of
- Design drawings ze
retrie apply ani risk analysis
- Investigation profile ve org retrieve
- Risk list from similar projects
Experts/engineers - risk management

retrieve
acquire - tunneling-related knowledge
- understanding drawings
- standards - experience from similar
- specifications projects
- guidelines
Safety risk identification systems 4-D visualization system

Fig. 7. Working process and KIM in the pre-construction phase.

procedures (e.g., procedures of risk inspection and risk response) were The knowledge contextual barrier also exists in the tunnel construc-
reviewed as a complement to identify and organize the safety knowl- tion risk management, because the construction industry has an in-
edge according to each risk management process. Subsequently, the trinsic high staff turnover rate, which results in substantial knowledge
KIM for the pre-construction phase and the construction implementa- loss as line workers resigned. The newly hired line workers have to be
tion phase were constructed with reference to the corresponding trained from scratch before they are capable of implementing risk
working processes, as shown in Figs. 7 and 8, respectively. control measures according to the manager's requirements.
It cannot be ignored that, most line workers in the construction
4.3. Analysis of knowledge flow barriers industry are not well-educated [76], which creates the knowledge re-
ceiver barrier in the safety knowledge flow. The knowledge source
The knowledge flow dynamics were clarified in the KIM by high- barrier does not exist in this case. The reason is that the tunnel con-
lighting the know-what, know-who and know-why of safety knowledge. struction project has a significant impact on the public, thus the con-
Therefore, the knowledge flow barriers can be identified by questioning tractors that win the bid are highly experienced.
the knowledge users about the difficulties when acquiring required
safety knowledge to perform a specific risk management activity. Seven 4.4. Design of knowledge flow facilitators
major knowledge flow barriers were identified in the construction of
the Wuhan metro project, as illustrated in Table 1. The knowledge flow facilitators can then be designed after the
It can be observed that the knowledge mechanism barrier is the most analysis of barriers. Several software and tools were developed and
crucial one in Wuhan metro project. The main reason is that the tun- deployed in the Wuhan metro project to address the barriers, which will
neling project risks relate to many parties, not only the project itself and be presented below.
stakeholders, but also the third parties. All risk factors related to them
need to be identified, monitored, and analyzed dynamically. In this 4.4.1. Safety risk identification system (SRIS)
circumstance, the related safety knowledge needs to be acquired and Two safety risk identification systems were designed to auto-
disseminated in a timely manner. However, the documents that parti- matically identify safety hazards and assess the risks based on con-
cipants usually refer to (e.g. project planning, design drawing, safety struction drawings [60] and BIM models [61] (Fig. 7), respectively.
statements, etc.) are always paper-based, poorly organized [59] and Both systems contain three modules, namely engineering parameters
hard to retrieve, let alone the huge amount of daily generated mon- recognition module, safety risk knowledge base, and risk identification
itoring data for decision supporting is scattered in different partici- module. They adopted different algorithms to extract engineering
pants. Moreover, experienced domain experts are often needed while parameters (e.g., soil type, depth of foundation pit) from drawings and
they are always a scarce resource who may not be able to provide real- BIM models. Corresponding reasoning rules were adopted to match the
time guidance due to time constraints. engineering parameters with the knowledge base. The safety risks were
The next one is knowledge characteristics barrier, which is caused by identified subsequently.
the intrinsic uncertainties of the tunnel construction associated with the
complicated geotechnical conditions. For example, the foundation pit 4.4.2. 4-Dimensional (4D) visualization system
edge of Xunlimen station (one station of Wuhan metro Line 2) is only To visualize the safety hazards that might be concealed in the dy-
1.25 m away from the nearest bridge pier of an existing light rail. The namic construction process, a 4D visualization system was developed
construction has great impacts on the piers and can affect the operation for the entire construction phase (Figs. 7 and 8). In the pre-construction
of the light rail. However, it's tough to control the deformation of bridge phase, a rule-based tool automatically detected potential risks and
piers, because the impact of various construction works on the bridge provided instructions for correction. In the construction implementa-
piers was unclear. tion phase, actual site monitoring data were continuously compared

118
C. Dong et al. Automation in Construction 93 (2018) 112–122

Recommend/
Select experts Analyze risk suggest control
measures
Working process Unacceptable

Acceptable or unwanted Approve risk

Determine risk
level
control Owner
Monitor risk measures
On-site
inspection Implement
Identify risk Assess risk Analyze risk control Contractor
Field measures
measurements

On-site
Assess risk Analyze risk Safety auditor

Negligible
auditing

On-site
inspection
Assess risk Analyze risk Consultant

Field Third party

measurements monitor

- decision theory

acquir
r
- project management tranfe
- experts engineers list - risk management
- knowledge of who-

e
- problem specification Group of
- tunneling-related
knows-what knowledge
- risk management experts/engineers
- risk management - experience from
- tunneling-related te
- specification genera

te
similar projects Owner

acq
knowledge

genera
- experience from

uir
- experience from transfer - recommended
similar projects

e
similar projects

er
- knowledge of actions

sf
e - risk evaluation

n
Knowledge uir acquire Decision-maker

tra
field measurement te
acq genera - recommended

gen
- knowledge of te - decisions on risk
dynamics map genera
ge

actions - project management

ne

era
safety weaknesses level and control
ra

Project manager Project manager - tunneling-related

te

te
measures
knowledge
4-D visualization system - suggestions transfer acquire

transfer
- monitoring plan e
uir - recommendations - risk classification
acq transf
share - design of control er - reports of control Contractor

ze
tra
er

nsf measures effects

transfer

ani
Project manager
nsf

er

org
- inspection results generate
tra

generate

organize
retrieve

generate
fer

acquire- experience and

ns

- plan of control transf

tra

Line worker er
organ Group of engineers measures knowledge of
ize implementing
- measurement data er
transf Line worker control measures
- knowledge of te - risk evaluation
era
safety weaknesses gen - recommended actions
sto er
- Risk priority re transf - risk plan modification
Knowledge acquire
- Risk response plan te Senior - tunneling-related - newly identified risk
base genera - auditing
transf

safety auditor
Safety auditor
re

results acquire knowledge

tri
ev

- experience from
er
e

gen

Safety auditor
- suggestions similar projects Web-
retrie

gen
era

based
ge

- recommendations
Software
ne

Workplace safety
te

- risk management
era
ve

safety
ra

- inspection transf
te

- specification er - risk evaluation intervention systems

te

er share
- experience from
acq Line worker results transf
Senior engineer - recommended risk early Consultant
u ire
similar projects - risk evaluation actions warning
r e

system
transf

Group of experts - recommended

generate
Project manager - measurement actions
generate data
Hardware

Monitoring
te - knowledge of Third party
- monitoring plan era system
gen acquir
e field measurement monitoring
based on
transfer
Line worker
I-o-T
store

Fig. 8. Working process and KIM in the construction implementation phase.

with the 4D model during the construction process. Therefore, the maintain the safety knowledge of the line workers. When workers
safety status of related components could be continuously visualized in breach the safety barriers (e.g. unsafe acts), warnings would be sent to
the system as potential risk develops. related workers via the safety barrier warning system [74], and the
incidents would be recorded in the Big-Data-based platform of workers'
4.4.3. Safety risk early warning system behavior [64]. In addition, safety engineers conducted inspections on
The web-based safety risk early warning system ([73], Fig. 8) acted the construction sites and recorded the near miss incidents through the
as a collaborative platform, which facilitated the communication and near miss management system [66]. With detailed descriptions and
collaboration among the organizations in Wuhan metro risk manage- photos of these incidents, the Big-Data-based platform of workers' be-
ment. The multi-source monitoring data were obtained by the Internet- havior and the near-miss management system also served as vivid
of-Things-based monitoring system ([63], Fig. 8) and transmitted to the materials for line workers to learn and perceive what is dangerous.
data warehouse [65] of the safety risk early warning system. The in-
spection data was also uploaded to the platform to support risk analysis.
Then, the safety risk of the project was analyzed by integrated nu- 4.5. Discussion and lessons learned
merical simulation [67,68], data mining [69,70] or hybrid approaches
[71,72], which was further proceeded to the experts for decision A complete validation of the benefits of applying the KIM could be
making. Finally, counter-measures would be designed and disseminated made by measuring the benefits obtained from the implementation of
to related individuals immediately to accelerate the risk response, such the safety knowledge flow facilitators in the Wuhan metro project.
as sending warning signals to workers who were entering high risk area However, it is difficult to measure how much the KIM has contributed
by a portable protection unit (PPU) and other acoustic/optic warning to those results, as has been pointed out in [77], because it is not clear
devices or disseminating warning messages to contractors if the dis- what kind of metrics should be adopted to evaluate the knowledge
placement of diaphragm wall exceeded its tolerance. management strategies [78,79]. Therefore, we have limited our eva-
luation to the testing of the KIM's applicability to the risk management
in the Wuhan metro construction, and extracting experiences from this
4.4.4. Workplace safety intervention systems case study.
Workplace safety intervention systems (Fig. 8) were designed to

119
C. Dong et al. Automation in Construction 93 (2018) 112–122

1. Effects on the organization: In July 15, 2009, an incident of foun-

Web-based safety risk early warning system [73]; 4D

Data warehouse [65]; near miss management system

Numerical simulation [67,68]; data mining [69,70];

dation pit wall leakage in the Pangxiejia station site occurred. With

Automatic and continuous monitoring system [63];

[66]; Big-Data-based platform of workers' behavior

Big-Data-based platform of workers' behavior [64]

Workplace safety intervention systems [64,74,75]

visualization system [62]; safety barrier warning
emergency supplies, equipment and personnel prepared in advance
based on the risk report of SRIS, the event has been effectively and
Safety risk identification systems [60,61]

efficiently controlled. The systems and software provided a con-

venient platform for safety knowledge sharing (timely and wide-
spread among stakeholders), as the KIM aims to facilitate a fostering
of safety knowledge “well-informed” culture.

hybrid risk modeling [71,72]

4D visualization system [62]
Knowledge flow facilitators

2. Effects on the working group and individuals: KIM can facilitate the
knowledge acquisition and reuse for the end users. Up to now, 2197
recordings of workers' unsafe behavior and > 1124 recordings of
near miss incidents were stored in the Big-Data-based platform of

system [74]
workers' behavior and near miss management system, respectively.
Based on these systems, the workers' unsafe behaviors or unsafe
[64]

operations were effectively intervened, as the number of detected

warnings is decreased by 77.3% in the first ten weeks. The systems
Contextual barrier; knowledge

apparently provide a more efficient way to overcome the receiver

barriers by safety knowledge internalization rather than socializa-
Knowledge characteristics

Knowledge characteristics

tion.
Contextual barrier;
Mechanism barrier

Mechanism barrier

Mechanism barrier

Mechanism barrier

receiver barrier

The KIM serves as a reference for system development in large

projects. The taxonomy of knowledge flow barrier is built on the vi-
Barrier type

sualized knowledge dynamics, which assists better understanding of the

barrier

barrier

inefficiency of safety management and supporting effective system de-

sign. Unfortunately, a quantitative method for evaluating the effec-
tiveness of the KIM in terms of safety improvement is not yet available
Intrinsic uncertainties associated with the geotechnical parameters
The knowledge for analyzing the risk was tacit, difficult to capture

Extensive measurement and observation data were not stored in

Time constraints on the risk information dissemination; lack of

Knowledge loss due to high staff turnover; some hired workers

The process for identifying risks from construction drawings is

Some high-risk sites (e.g., cross-river tunnel section, Xunlimen

and this will be part of future research plan.

coordination between participants due to the geographical
station) and workers' behavior required high-frequency

an explicit, structured manner, and difficult to retrieve

5. Conclusions
lacked considerable education and experience

Safety has always been a persistent problem in the construction

industry, particularly in underground construction projects due to
complex geotechnical conditions. The importance of safety knowledge
has been frequently highlighted; yet, a systemic method for knowledge
management system development is lacking. This paper developed the
KIM to visualize the safety knowledge flow in risk management. In the
monitoring and judgment
Knowledge flow barriers

tunnel construction risk management, what safety knowledge was re-

quired, who owned it were highlighted by the KIM, and why was it hard
time-consuming

to acquire was answered as the knowledge flow barriers were identi-

fied. Appropriate knowledge flow facilitators were later designed to
dispersion
and share

enhance the safety knowledge flow in tunnel construction risk man-

agement.
The KIM has been applied in the Wuhan metro project for nearly ten
From engineers to risk evaluation

years. Several major knowledge flow barriers were identified and

From domain knowledge to line
measurement/observation data

Interaction among participants

observation data to data users
From experts to risk priority

overcome by developing various systems. To date, no major accident

Summary of the knowledge flow barriers and designed facilitators.

From design documents to

occurred during the construction, implying the potential of KIM as a

Ends of knowledge flow

From measurement and

From line workers to

promising blueprint for the knowledge-intensive industries. Although

experts/engineers

the development of KIM is illustrated in the context of tunneling pro-

jects, it is generally applicable to other types of construction projects.
The main difference lies in the elicitation of the risk management
workers
results
report

process, which plays an important role in the KIM development and has
different definitions, components, tools, and acceptance criteria in
different construction areas.
Analysis of the possible

Determine risk level

Implement control
Working process

Acknowledgements
Monitor risk

Assess risk
(high) risk

measures

The presented work has been supported by the National Natural

Science Foundation of China (NSFC) under Grant No. 51608399 and
51678265, the National 12th Five-Year Plan Major Scientific and
Pre-construction

Implementation

Technological Issues (NFYPMSTI) under Grant No. 2015BAK33B04, the

Research Grants Council of Hong Kong Grant No. PolyU 152093/14E,
and the Fundamental Research Funds for the Central Universities,
Table 1

Phase

HUST: cx15-011.

120
C. Dong et al. Automation in Construction 93 (2018) 112–122

References dss.2004.03.008.
[28] S. Pyo, Knowledge map for tourist destinations—needs and implications, Tour.
Manag. 26 (4) (2005) 583–594, http://dx.doi.org/10.1016/j.tourman.2004.03.001.
[1] H.H. Einstein, D. Veneziano, Risk Analysis for Tunneling Projects (Ph.D.), Dept. of [29] J.-Y. Lai, C.-T. Wang, C.-Y. Chou, How knowledge map fit and personalization affect
Civil and Environmental Engineering, Massachusetts Institute of Technology, 2010 success of KMS in high-tech firms, Technovation 29 (4) (2009) 313–324, http://dx.
retrieved from https://dspace.mit.edu/handle/1721.1/58282 (last access date < doi.org/10.1016/j.technovation.2008.10.007.
11/01/2018 >). [30] A. Gómez, A. Moreno, J. Pazos, A. Sierra-Alonso, Knowledge maps: an essential
[2] S. Degn Eskesen, P. Tengborg, J. Kampmann, T. Holst Veicherts, Guidelines for technique for conceptualisation, Data Knowl. Eng. 33 (2) (2000) 169–190, http://
tunnelling risk management: international tunnelling association, working group dx.doi.org/10.1016/S0169-023X(99)00050-6.
no. 2, Tunn. Undergr. Space Technol. 19 (3) (2004) 217–237, http://dx.doi.org/10. [31] S. Kim, E. Suh, H. Hwang, Building the knowledge map: an industrial case study, J.
1016/j.tust.2004.01.001. Knowl. Manag. 7 (2) (2003) 34–45, http://dx.doi.org/10.1108/
[3] I.C. Cárdenas, S.S. Al-jibouri, J.I. Halman, F.A. van Tol, Capturing and integrating 13673270310477270.
knowledge for managing risks in tunnel works, Risk Anal. 33 (1) (2013) 92–108, [32] J.L. Gordon, Creating knowledge maps by exploiting dependent relationships,
http://dx.doi.org/10.1111/j.1539-6924.2012.01829.x. Knowl.-Based Syst. 13 (2) (2000) 71–79, http://dx.doi.org/10.1016/S0950-
[4] H.C. Tan, C.J. Anumba, P.M. Carrillo, D. Bouchlaghem, J. Kamara, C. Udeaja, 7051(00)00048-4.
Capture and Reuse of Project Knowledge in Construction, John Wiley & Sons, 2009 [33] F.-r. Lin, C.-m. Hsueh, Knowledge map creation and maintenance for virtual com-
(ISBN: 1444315455). munities of practice, Inf. Process. Manag. 42 (2) (2006) 551–568, http://dx.doi.
[5] J. Tah, V. Carr, Knowledge-based approach to construction project risk manage- org/10.1016/j.ipm.2005.03.026.
ment, J. Comput. Civ. Eng. 15 (3) (2001) 170–177, http://dx.doi.org/10.1061/ [34] T.R. Gruber, Toward principles for the design of ontologies used for knowledge
(ASCE)0887-3801(2001)15:3(170). sharing? Int. J. Hum. Comp. Stud. 43 (5–6) (1995) 907–928, http://dx.doi.org/10.
[6] S. Kivrak, G. Arslan, I. Dikmen, M.T. Birgonul, Capturing knowledge in construction 1006/ijhc.1995.1081.
projects: knowledge platform for contractors, J. Manag. Eng. 24 (2) (2008) 87–95, [35] Y. Lu, Q. Li, Z. Zhou, Y. Deng, Ontology-based knowledge modeling for automated
http://dx.doi.org/10.1061/(ASCE)0742-597X(2008)24:2(87). construction safety checking, Saf. Sci. 79 (2015) 11–18, http://dx.doi.org/10.1016/
[7] T. Mezher, M.A. Abdul-Malak, I. Ghosn, M. Ajam, Knowledge management in me- j.ssci.2015.05.008.
chanical and industrial engineering consulting: a case study, J. Manag. Eng. 21 (3) [36] Q.T. Le, D.Y. Lee, C.S. Park, A social network system for sharing construction safety
(2005) 138–147, http://dx.doi.org/10.1061/(ASCE)0742-597X(2005)21:3(138). and health knowledge, Autom. Constr. 46 (2014) 30–37, http://dx.doi.org/10.
[8] Y. Rezgui, C.J. Hopfe, C. Vorakulpipat, Generations of knowledge management in 1016/j.autcon.2014.01.001.
the architecture, engineering and construction industry: an evolutionary perspec- [37] H.-H. Wang, F. Boukamp, T. Elghamrawy, Ontology-based approach to context
tive, Adv. Eng. Inform. 24 (2) (2010) 219–228, http://dx.doi.org/10.1016/j.aei. representation and reasoning for managing context-sensitive construction in-
2009.12.001. formation, J. Comput. Civ. Eng. 25 (5) (2011) 331–346, http://dx.doi.org/10.
[9] G. Peng, H. Wang, H. Zhang, Y. Zhao, A.L. Johnson, A collaborative system for 1061/(ASCE)CP.1943-5487.0000094.
capturing and reusing in-context design knowledge with an integrated representa- [38] H.-H. Wang, F. Boukamp, Ontology-based representation and reasoning framework
tion model, Adv. Eng. Inform. 33 (2017) 314–329, http://dx.doi.org/10.1016/j.aei. for supporting job hazard analysis, J. Comput. Civ. Eng. 25 (6) (2011) 442–456,
2016.12.007. http://dx.doi.org/10.1061/(ASCE)CP.1943-5487.0000125.
[10] Y. Chen, S.J. Treado, J.I. Messner, Building HVAC control knowledge data schema – [39] L. Ding, B. Zhong, S. Wu, H. Luo, Construction risk knowledge management in BIM
towards a unified representation of control system knowledge, Autom. Constr. 72 ( using ontology and semantic web technology, Saf. Sci. 87 (2016) 202–213, http://
(2016) 174–186, http://dx.doi.org/10.1016/j.autcon.2016.08.036. dx.doi.org/10.1016/j.ssci.2016.04.008.
[11] Y. Malhotra, Knowledge Management and Virtual Organizations, IGI Global, 2000 [40] G. Mansingh, K.-M. Osei-Bryson, H. Reichgelt, Building ontology-based knowledge
(ISBN: 187828973X). maps to assist knowledge process outsourcing decisions, Knowl. Manag. Res. Pract.
[12] Y.-C. Lin, H.-Y. Lee, Developing project communities of practice-based knowledge 7 (1) (2009) 37–51, http://dx.doi.org/10.1057/kmrp.2008.37.
management system in construction, Autom. Constr. 22 (2012) 422–432, http://dx. [41] L. Rao, G. Mansingh, K.-M. Osei-Bryson, Building ontology based knowledge maps
doi.org/10.1016/j.autcon.2011.10.004. to assist business process re-engineering, Decis. Support. Syst. 52 (3) (2012)
[13] S.Y.L. Yin, H.P. Tserng, M.D. Tsai, A model of integrating the cycle of construction 577–589, http://dx.doi.org/10.1016/j.dss.2011.10.014.
knowledge flows: lessons learned in Taiwan, Autom. Constr. 17 (5) (2008) [42] J.-S. Gao, D.-P. Li, X.-D. Wang, Study on knowledge map construction for virtual
536–549, http://dx.doi.org/10.1016/j.autcon.2007.10.002. team based on ontology, IT in Medicine & Education, ITIME'09. IEEE International
[14] J. Reason, Managing the Risks of Organizational Accidents, Routledge, 2016 (ISBN: Symposium on, Vol. 1 IEEE, Jinan, China, 2009, pp. 360–365, , http://dx.doi.org/
1134855354). 10.1109/ITIME.2009.5236397 (14–16 Aug. 2009, ISBN: 1424439280).
[15] I.J. Obadia, M.C.R. Vidal, P.F.F.F. e Melo, An adaptive management system for [43] Y.-C. Lin, L.-C. Wang, H.P. Tserng, Enhancing knowledge exchange through web
hazardous technology organizations, Saf. Sci. 45 (3) (2007) 373–396, http://dx.doi. map-based knowledge management system in construction: lessons learned in
org/10.1016/j.ssci.2006.07.002. Taiwan, Autom. Constr. 15 (6) (2006) 693–705, http://dx.doi.org/10.1016/j.
[16] J. Reason, Achieving a safe culture: theory and practice, Work Stress. 12 (3) (1998) autcon.2005.09.006.
293–306, http://dx.doi.org/10.1080/02678379808256868. [44] J.-B. Yang, Developing a knowledge map for construction scheduling using a novel
[17] R.M. Choudhry, D. Fang, S. Mohamed, The nature of safety culture: a survey of the approach, Autom. Constr. 16 (6) (2007) 806–815, http://dx.doi.org/10.1016/j.
state-of-the-art, Saf. Sci. 45 (10) (2007) 993–1012, http://dx.doi.org/10.1016/j. autcon.2007.02.005.
ssci.2006.09.003. [45] J.-H. Woo, M.J. Clayton, R.E. Johnson, B.E. Flores, C. Ellis, Dynamic knowledge
[18] I. Nonaka, A dynamic theory of organizational knowledge creation, Organ. Sci. 5 (1) map: reusing experts' tacit knowledge in the AEC industry, Autom. Constr. 13 (2)
(1994) 14–37, http://dx.doi.org/10.1287/orsc.5.1.14. (2004) 203–207, http://dx.doi.org/10.1016/j.autcon.2003.09.003.
[19] P. Mitropoulos, T.S. Abdelhamid, G.A. Howell, Systems model of construction ac- [46] I. Nonaka, The Knowledge-Creating Company, Harvard Business Review Press,
cident causation, J. Constr. Eng. Manag. 131 (7) (2005) 816–825, http://dx.doi. 2008 (ISBN: 1633691373).
org/10.1061/(ASCE)0733-9364(2005)131:7(816). [47] I. Nonaka, H. Takeuchi, The Knowledge-Creating Company: How Japanese
[20] B. Hadikusumo, S. Rowlinson, Capturing safety knowledge using design-for-safety- Companies Create the Dynamics of Innovation, Oxford University Press, 1995
process tool, J. Constr. Eng. Manag. 130 (2) (2004) 281–289 (10.1061/(ASCE) (ISBN: 0199879923).
0733-9364(2004)130:2(281)). [48] P. Carrillo, P. Chinowsky, Exploiting knowledge management: the engineering and
[21] Y.Z. Toh, Y.M. Goh, B.H. Guo, Knowledge, attitude, and practice of design for construction perspective, J. Manag. Eng. 22 (1) (2006) 2–10, http://dx.doi.org/10.
safety: multiple stakeholders in the singapore construction industry, J. Constr. Eng. 1061/(ASCE)0742-597X(2006)22:1(2).
Manag. 143 (5) (2016) 04016131, http://dx.doi.org/10.1061/(ASCE)CO.1943- [49] M. Nissen, R. Levitt, Dynamic Models of Knowledge-Flow Dynamics, (2002) re-
7862.0001279. tieved from http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.95.3090&
[22] S. Zhang, F. Boukamp, J. Teizer, Ontology-based semantic modeling of construction rep=rep1&type=pdf (last access date < 11/01/2018 >).
safety knowledge: towards automated safety planning for job hazard analysis [50] M. Nissen, R. Levitt, Toward Simulation Models of Knowledge-Intensive Work
(JHA), Autom. Constr. 52 (2015) 29–41, http://dx.doi.org/10.1016/j.autcon.2015. Processes, Center for Integrated Facility Engineering, (2002) retrieved from https://
02.005. www.researchgate.net/profile/Raymond_Levitt/publication/242414417_Toward_
[23] P.X. Zou, P. Lun, D. Cipolla, S. Mohamed, Cloud-based safety information and Simulation_Models_of_Knowledge-Intensive_Work_Processes/links/
communication system in infrastructure construction, Saf. Sci. 98 (2017) 50–69, 548881200cf2ef3447909d5c/Toward-Simulation-Models-of-Knowledge-Intensive-
http://dx.doi.org/10.1016/j.ssci.2017.05.006. Work-Processes.pdf (last access date < 11/01/2018 >).
[24] D.K. Chua, Y.M. Goh, Incident causation model for improving feedback of safety [51] M.N. Wexler, The who, what and why of knowledge mapping, J. Knowl. Manag. 5
knowledge, J. Constr. Eng. Manag. 130 (4) (2004) 542–551, http://dx.doi.org/10. (3) (2001) 249–264, http://dx.doi.org/10.1108/EUM0000000005868.
1061/(ASCE)0733-9364(2004)130:4(542). [52] I. Nonaka, A dynamic theory of organizational knowledge creation, in: N. W.C.B.
[25] M. Zhang, D. Fang, A continuous behavior-based safety strategy for persistent safety Choo (Ed.), The Strategic Management of Intellectual Capital and Organizational
improvement in construction industry, Autom. Constr. 34 (2013) 101–107, http:// Knowledge, Oxford University Press, Oxford, 2002, pp. 437–462 (ISBN: 0-19-
dx.doi.org/10.1016/j.autcon.2012.10.019. 513866-X (pbk.)).
[26] N. Forcada, M. Casals, A. Fuertes, M. Gangolells, X. Roca, A web-based system for [53] M.E. Nissen, An extended model of knowledge-flow dynamics, Commun. Assoc. Inf.
sharing and disseminating research results: the underground construction case Syst. 8 (1) (2002) 18 retrieved from http://aisel.aisnet.org/cais/vol8/iss1/18 (last
study, Autom. Constr. 19 (4) (2010) 458–474, http://dx.doi.org/10.1016/j.autcon. access date < 12/03/2018 >).
2009.12.018. [54] O.M. Rodríguez-Elias, A.I. Martínez-García, A. Vizcaíno, J. Favela, M. Piattini, A
[27] T.-H. Ong, H. Chen, W.-k. Sung, B. Zhu, Newsmap: a knowledge map for online framework to analyze information systems as knowledge flow facilitators, Inf.
news, Decis. Support. Syst. 39 (4) (2005) 583–597, http://dx.doi.org/10.1016/j. Softw. Technol. 50 (6) (2008) 481–498, http://dx.doi.org/10.1016/j.infsof.2007.

121
C. Dong et al. Automation in Construction 93 (2018) 112–122

07.002. metro construction in complex environments, Int. J. Proj. Manag. 29 (7) (2011)
[55] H.C. Tan, P.M. Carrillo, C.J. Anumba, N. Bouchlaghem, J.M. Kamara, C.E. Udeaja, 797–807, http://dx.doi.org/10.1016/j.ijproman.2011.04.006.
Development of a methodology for live capture and reuse of project knowledge in [68] L. Ding, L. Zhang, X. Wu, M.J. Skibniewski, Y. Qunzhou, Safety management in
construction, J. Manag. Eng. 23 (1) (2007) 18–26, http://dx.doi.org/10.1061/ tunnel construction: case study of Wuhan metro construction in China, Saf. Sci. 62
(ASCE)0742-597X(2007)23:1(18). (2014) 8–15, http://dx.doi.org/10.1016/j.ssci.2013.07.021.
[56] H.P. Tserng, S.Y.L. Yin, R.J. Dzeng, B. Wou, M.D. Tsai, W.Y. Chen, A study of on- [69] L. Ding, F. Wang, H. Luo, M. Yu, X. Wu, Feedforward analysis for shield-ground
tology-based risk management framework of construction projects through project system, J. Comput. Civ. Eng. 27 (3) (2012) 231–242, http://dx.doi.org/10.1061/
life cycle, Autom. Constr. 18 (7) (2009) 994–1008, http://dx.doi.org/10.1016/j. (ASCE)CP.1943-5487.0000219.
autcon.2009.05.005. [70] Y. Zhou, W. Su, L. Ding, H. Luo, P.E.D. Love, Predicting safety risks in deep foun-
[57] C. Lin, B. Tan, S. Chang, An exploratory model of knowledge flow barriers within dation pits in subway infrastructure projects: support vector machine approach, J.
healthcare organizations, Inf. Manag. 45 (5) (2008) 331–339, http://dx.doi.org/10. Comput. Civ. Eng. 31 (5) (2017), http://dx.doi.org/10.1061/(ASCE)CP.1943-5487.
1016/j.im.2008.03.003. 0000700.
[58] C. Lin, J.-C. Wu, D.C. Yen, Exploring barriers to knowledge flow at different [71] F. Wang, L. Ding, P.E. Love, D.J. Edwards, Modeling tunnel construction risk dy-
knowledge management maturity stages, Inf. Manag. 49 (1) (2012) 10–23, http:// namics: addressing the production versus protection problem, Saf. Sci. 87 (2016)
dx.doi.org/10.1016/j.im.2011.11.001. 101–115, http://dx.doi.org/10.1016/j.ssci.2016.01.014.
[59] D. Lieyun, X. Jie, A review of metro construction in China: organization, market, [72] F. Wang, L.Y. Ding, H.B. Luo, P.E.D. Love, Probabilistic risk assessment of tun-
cost, safety and schedule, Front. Eng. Manag. 4 (1) (2017) 4–19, http://dx.doi.org/ neling-induced damage to existing properties, Expert Syst. Appl. 41 (4) (2014)
10.15302/J-FEM-2017015. 951–961, http://dx.doi.org/10.1016/j.eswa.2013.06.062.
[60] L.Y. Ding, H.L. Yu, H. Li, C. Zhou, X.G. Wu, M.H. Yu, Safety risk identification [73] L.Y. Ding, C. Zhou, Development of web-based system for safety risk early warning
system for metro construction on the basis of construction drawings, Autom. Constr. in urban metro construction, Autom. Constr. 34 (2013) 45–55, http://dx.doi.org/
27 (2012) 120–137, http://dx.doi.org/10.1016/j.autcon.2012.05.010. 10.1016/j.autcon.2012.11.001.
[61] L. Zhang, X. Wu, L. Ding, M.J. Skibniewski, Y. Lu, BIM-based risk identification [74] C. Zhou, L.Y. Ding, Safety barrier warning system for underground construction
system in tunnel construction, J. Civ. Eng. Manag. 22 (4) (2016) 529–539, http:// sites using Internet-of-Things technologies, Autom. Constr. 83 (2017) 372–389,
dx.doi.org/10.3846/13923730.2015.1023348. http://dx.doi.org/10.1016/j.autcon.2017.07.005.
[62] Y. Zhou, L.Y. Ding, L.J. Chen, Application of 4D visualization technology for safety [75] Q.Z. Yu, L.Y. Ding, C. Zhou, H.B. Luo, Analysis of factors influencing safety man-
management in metro construction, Autom. Constr. 34 (2013) 25–36, http://dx.doi. agement for metro construction in China, Accid. Anal. Prev. 68 (2014) 131–138,
org/10.1016/j.autcon.2012.10.011. http://dx.doi.org/10.1016/j.aap.2013.07.016.
[63] L.Y. Ding, C. Zhou, Q.X. Deng, H.B. Luo, X.W. Ye, Y.Q. Ni, P. Guo, Real-time safety [76] V. Sousa, N.M. Almeida, L.A. Dias, Risk-based management of occupational safety
early warning system for cross passage construction in Yangtze Riverbed Metro and health in the construction industry – part 1: background knowledge, Saf. Sci. 66
Tunnel based on the internet of things, Autom. Constr. 36 (2013) 25–37, http://dx. (2014) 75–86, http://dx.doi.org/10.1016/j.ssci.2014.02.008.
doi.org/10.1016/j.autcon.2013.08.017. [77] N. Anquetil, K.M. de Oliveira, K.D. de Sousa, M.G.B. Dias, Software maintenance
[64] S.Y. Guo, L.Y. Ding, H.B. Luo, X.Y. Jiang, A Big-Data-based platform of workers' seen as a knowledge management issue, Inf. Softw. Technol. 49 (5) (2007)
behavior: observations from the field, Accid. Anal. Prev. 93 (2016) 299–309, 515–529, http://dx.doi.org/10.1016/j.infsof.2006.07.007.
http://dx.doi.org/10.1016/j.aap.2015.09.024. [78] F. Folkens, M. Spiliopoulou, Towards an evaluation framework for knowledge
[65] L. Ma, H.B. Luo, H.R. Chen, Safety risk analysis based on a geotechnical in- management systems, in: D. Karagiannis, U. Reimer (Eds.), Practical Aspects of
strumentation data warehouse in metro tunnel project, Autom. Constr. 34 (2013) Knowledge Management. PAKM 2004, Lecture Notes in Computer Science, vol.
75–84, http://dx.doi.org/10.1016/j.autcon.2012.10.009. 3336, Springer, Berlin, Heidelberg, 2004, , http://dx.doi.org/10.1007/978-3-540-
[66] C. Zhou, L. Ding, M.J. Skibniewski, H. Luo, S. Jiang, Characterizing time series of 30545-3_3 (ISBN: 978-3-540-30545-3).
near-miss accidents in metro construction via complex network theory, Saf. Sci. 98 [79] M. Jennex, L. Olfman, Assessing knowledge management success, Int. J. Knowl.
(2017) 145–158, http://dx.doi.org/10.1016/j.ssci.2017.06.012. Manag. 1 (2) (2005) 33–49, http://dx.doi.org/10.4018/jkm.2005040104.
[67] L.Y. Ding, X.G. Wu, H. Li, H.B. Luo, Y. Zhou, Study on safety control for Wuhan

122

View publication stats