Professional Documents
Culture Documents
Knowledge-Based Decision Support System For Prefabricated Prefinished Volumetric Construction
Knowledge-Based Decision Support System For Prefabricated Prefinished Volumetric Construction
Knowledge-Based Decision Support System For Prefabricated Prefinished Volumetric Construction
Automation in Construction
journal homepage: www.elsevier.com/locate/autcon
A R T I C LE I N FO A B S T R A C T
Keywords: As prefabricated prefinished volumetric construction (PPVC) has gained considerable attention worldwide in the
Prefabrication past few years, decision-making for implementing PPVC becomes critical. As a result, this study aims to (1)
Prefabricated prefinished volumetric identify the key decision-making factors (DMFs) for the adoption of PPVC, (2) propose a scoring approach that
construction can assess the feasibility of using PPVC for a given project, and (3) develop a Knowledge-Based Decision Support
Decision-making
System for Prefabricated Prefinished Volumetric Construction (KBDSS-PPVC) that can facilitate the decision-
Decision support system
making for PPVC implementation. To achieve these goals, a comprehensive literature review and pilot inter-
Singapore
views with industry experts were conducted first, followed by a structured questionnaire administered to 41
construction organizations in Singapore. Results of the questionnaire reported 19 DMFs of PPVC, which were
then used to create the PPVC scoring approach. Subsequently, the KBDSS-PPVC was developed using the created
PPVC scoring approach. Lastly, a panel of industry experts validated the developed KBDSS-PPVC, by utilizing the
tool for their current construction projects. Validation results showed that the developed system could provide
reliable recommendations for the industry practitioners on the decision-making of PPVC. Existing literature has
seldom addressed the decision-making of PPVC, therefore, this study bridges the knowledge gap and contributes
to the current body of knowledge. Furthermore, the developed KBDSS-PPVC would be useful to the industry
practitioners as well, because it can help them achieve a better and easier decision-making of PPVC.
⁎
Corresponding author.
E-mail addresses: bdghbg@nus.edu.sg (B.-G. Hwang), ming.shan@csu.edu.cn (M. Shan).
https://doi.org/10.1016/j.autcon.2018.06.016
Received 8 December 2017; Received in revised form 14 June 2018; Accepted 26 June 2018
Available online 04 July 2018
0926-5805/ © 2018 Elsevier B.V. All rights reserved.
B.-G. Hwang et al. Automation in Construction 94 (2018) 168–178
Table 1
Potential DMFs of PPVC.
No. Potential DMF A B C D E F G H I J K L M N O P Q R S
Note: A = Construction Industry Institute [46]; B = Murtaza, et al. [34]; C = Gibb and Isack [37]; D = Blismas, et al. [47]; E = Pan, et al. [38]; F = Tam, et al. [45];
G = Pan, et al. [39]; H = Scofield, et al. [48]; I = Chen, et al. [35]; J = McGraw Hill [49]; K = Pan and Sidwell [30]; L = Azhar, et al. [41]; M = Rahman [43];
N = Elnaas [50]; O = Zhai, et al. [36]; P = Mao, et al. [12]; Q = Lee and Kim [51]; R = Liu, et al. [52]; S = Wong, et al. [40].
Table 2 Table 3
DMFs affecting the adoption of PPVC method. Profile of the respondents.
Code Decision-making factors Respondent Profile Frequency Percentage (%) Cumulative
Percentage (%)
DMF1 Suitability of design for PPVC
DMF2 Presence of repetitive layout in design Respondent's Institution
DMF3 Number of stories Governmental agency 1 3 3
DMF4 Structural stability of individual and assembled modules Developer 5 12 15
DMF5 Need for inspection/supervision of manufactured units Construction company 21 51 66
DMF6 Lead time for PPVC modules Consultancy 6 15 81
DMF7 PPVC module's size Architectural firm 3 7 88
DMF8 Site layout (e.g. availability of storage space for PPVC modules/ Quantity Surveyor 2 5 93
availability of sufficient space to unload) PPVC Supplier/ 3 7 100
DMF9 Site location Manufacturer
DMF10 Availability of skilled labor and experienced supervising team Respondent's occupation
DMF11 Availability of transportation equipment Architect 4 10 10
DMF12 Availability of construction equipment Engineer 13 32 42
DMF13 Organization's familiarity with PPVC method Project Manager 7 17 59
DMF14 Use of information and communication technology (e.g. BIM) Quantity Surveyor 9 22 81
DMF15 Size and type of the project (i.e. project scope) Facility Manager 8 19 100
DMF16 Shorter construction duration at site Years of experiences in the
DMF17 Improved construction safety construction industry
DMF18 Reducing environmental impact through the reduction of site activities 1–4 years 16 39 39
DMF19 Reducing neighborhood disruption and noise 5–9 years 7 17 56
> 10 years 18 44 100
Years of experiences with
PPVC projects
construction methods. Second, how would these factors affect the de-
< 1 year 27 66 66
cision-making of PPVC? Third, is it possible to develop a convenient, 2 years 9 22 88
computer-aided tool for industry practitioners to facilitate their deci- 3 years 5 12 100
sion-making of PPVC? To answer these questions, this study identified
the key decision-making factors (DMFs) affecting the use of PPVC,
proposed a scoring approach that could assess the feasibility of using (KBDSS-PPVC).
PPVC for a given project, and developed a Knowledge-Based Decision This research study was conducted in the context of Singapore.
Support System for Prefabricated Prefinished Volumetric Construction Being a city-state with a limited working age population [22], Singa-
pore gives significant value to production approaches that require less
169
B.-G. Hwang et al. Automation in Construction 94 (2018) 168–178
Table 4
Assessment of the DMFs of PPVC.
Code DMF Rank Mean Shapiro-Wilk test (p-value) Kruskal-Wallis test (p-value)
⁎
DMF2 Presence of repetitive layout in design 1 4.59 0.000 0.179
DMF1 Suitability of design for PPVC 2 4.34 0.000⁎ 0.151
DMF4 Structural stability of individual and assembled modules 3 4.05 0.000⁎ 0.770
DMF6 Lead time for PPVC modules 4 4.05 0.000⁎ 0.712
DMF8 Site layout 5 3.98 0.000⁎ 0.513
DMF7 PPVC module's size 6 3.95 0.000⁎ 0.283
DMF16 Shorter construction duration on site 7 3.95 0.000⁎ 0.588
DMF10 Availability of skilled labor and experienced supervising team 8 3.90 0.000⁎ 0.760
DMF13 Organization's familiarity with PPVC method 8 3.90 0.000⁎ 0.682
DMF19 Reducing neighborhood disruption and noise 10 3.85 0.000⁎ 0.848
DMF3 Number stories 11 3.80 0.000⁎ 0.385
DMF12 Availability of construction equipment 11 3.80 0.000⁎ 0.137
DMF17 Improved construction safety 11 3.80 0.000⁎ 0.195
DMF18 Reducing environmental impact through the reduction of site activities 11 3.80 0.000⁎ 0.779
DMF14 Use of information and communication technology (e.g. BIM) 15 3.76 0.000⁎ 0.504
DMF15 Size and type of the project (i.e. project scope) 15 3.76 0.000⁎ 0.205
DMF5 Need for inspection/supervision of manufactured units 17 3.73 0.000⁎ 0.945
DMF11 Availability of transportation equipment 17 3.73 0.000⁎ 0.919
DMF9 Site location 19 3.56 0.000⁎ 0.935
⁎
The Shapiro-Wilk test was significant at the significance level of 0.05, suggesting the data were not normally distributed.
manpower. PPVC is a lean production approach requiring less work- which addresses the safety concern of construction projects successfully
force than conventional construction approaches [23] and hence, PPVC [23]. Despite criticism that the overuse of PPVC would result in a loss of
is highly recognized by the Singapore government and has been ac- aesthetic value for buildings, the advantages of PPVC make it an at-
tively promoted in the country for the past few years [24,25]. To date, tractive proposition for the global construction community [19,31].
several PPVC projects have been successfully executed in Singapore.
Examples include but are not limited to the Crown Plaza Changi Airport 2.2. PPVC in Singapore
Hotel Extension [26], and the North Hill Residential Halls of Nanyang
Technological University [27]. Moreover, as planned by the Singapore Since 2010, BCA Singapore has launched two rounds of
government, more PPVC projects are expected to be implemented in Construction Productivity Roadmap (1st in 2010 and 2nd in 2015)
Singapore in the near future. This is because the local authority has aimed at improving productivity in the local construction industry [32].
aimed to construct nearly 35% of the newly launched public housing Particularly, the 2nd Construction Productivity Roadmap BCA high-
projects in Singapore using the PPVC approach by 2019 [28]. As a re- lighted prefabrication as one of the main directions that the local
sult, Singapore is the right and suitable context for conducting this construction industry should pursue [19]. In recent years, BCA has been
research. actively promoting the PPVC approach. For example, BCA stipulated
PPVC is an emerging off-site construction method and in-depth re- that from 1 November 2014 the adoption of PPVC method would be one
search of this innovative approach in the existing literature is lacking. of the land sale conditions for some selected land parcels under the
Thus, this study can contribute to the current body of knowledge on off- Government Land Sales Program [24]. In 2015, BCA [33] further re-
site construction research through the addition of literature on PPVC. quired that, for the developments of these selected land parcels, the
This study would also be beneficial to the practice, as the study findings minimum level of use of PPVC should be 65% of the total constructed
can help industry practitioners enhance their understanding of floor area (including all floors except the basement and any floor area
PPVC·Furthermore, the KBDSS-PPVC developed in this study can built for use as a roof or car park). Moreover, BCA [24] has set up an
greatly facilitate the decision-making of PPVC for the industry practi- acceptance framework recently to certify each PPVC supplier and
tioners. manufacturer in order to ensure the reliability and durability of the
different PPVC systems used at mandated development sites. In addi-
2. Background tion to BCA, the Housing and Development Board (HDB) of Singapore is
another local authority who has an interest in PPVC. HDB deems PPVC
2.1. PPVC as a productive and helpful instrument for the development of HDB
apartments [19]. Furthermore, it is reported that HDB has carried out a
According to the definition of the Building and Construction series of research initiatives funding the development of an integrated,
Authority (BCA) [24], PPVC is a construction method whereby free- automatic PPVC production system [19]. In summary, PPVC is being
standing volumetric modules (complete with finishes for walls, floors, actively promoted in Singapore with patronage from the local autho-
and ceilings) are manufactured and assembled in an accredited fabri- rities.
cation facility in accordance with any accredited fabrication method,
and then transported to site and installed to form a building. In com- 2.3. Decision-making factors for the adoption of PPVC
parison with conventional construction methods where design, en-
gineering, and construction activities are consecutive, PPVC allows Given the limited amount of research on PPVC, the decision-making
works to be done in parallel [24]. Thus, it can expedite the project factors (DMFs) for the adoption of PPVC can hardly be identified from
implementation significantly and thereby increase productivity [29]. literature directly. However, there is considerable documentation on
Furthermore, PPVC can bring tangible cost savings through a reduction general off-site construction methods, which is highly relevant to PPVC.
in on-site manpower and equipment requirement when compared to As a result, this study conducted a comprehensive literature review of
conventional construction methods [30]. Moreover, by carrying out the the research conducted on general off-site construction methods, to
majority of the work in a controlled, factory environment, PPVC is also identify the DMFs that may affect the adoption of PPVC.
able to provide a healthier and safer working environment for workers, The literature review showed that several studies had done excellent
170
B.-G. Hwang et al. Automation in Construction 94 (2018) 168–178
Table 6
0.000⁎
0.003⁎
0.031⁎
0.018⁎
0.005⁎
0.032⁎
DMF9
0.136
0.058
0.135
0.256
0.260
0.132
0.349
0.360
0.450
0.464
0.480
0.231
PPVC scores of different DMFs at different assessment levels.
–
DMF Mean Maximum Score of PPVC Assessment Levels of DMF
DMF11
0.000⁎
0.006⁎
0.044⁎
0.070
0.153
0.145
0.285
0.353
0.377
0.506
0.860
0.477
0.637
0.593
0.861
0.878
0.987
1 2 3 4 5
–
DMF1 4.34 6 1 2 3 5 6
DMF2 4.59 6 1 2 4 5 6
0.000⁎
0.003⁎
0.035⁎
DMF5
0.051
0.174
0.230
0.349
0.271
0.490
0.527
0.793
0.665
0.785
0.749
0.969
0.873
DMF3 3.80 5 1 2 3 4 5
DMF4 4.05 5 1 2 3 4 5
–
DMF5 3.73 5 1 2 3 4 5
DMF6 4.05 5 1 2 3 4 5
DMF15
0.000⁎
0.005⁎
0.171
0.171
0.202
0.277
0.293
0.404
0.565
0.685
0.748
0.802
0.914
0.834
0.904
DMF7 3.95 5 1 2 3 4 5
DMF8 3.98 5 1 2 3 4 5
–
DMF9 3.56 5 1 2 3 4 5
DMF10 3.90 5 1 2 3 4 5
DMF14
0.000⁎
0.009⁎
DMF11 3.73 5 1 2 3 4 5
0.062
0.119
0.287
0.185
0.469
0.463
0.435
0.541
0.906
0.744
0.692
0.654
DMF12 3.80 5 1 2 3 4 5
–
DMF13 3.76 5 1 2 3 4 5
DMF14 3.76 5 1 2 3 4 5
DMF18
0.000⁎
0.016⁎
DMF15 4.34 5 1 2 3 4 5
0.135
0.165
0.375
0.378
0.555
0.746
0.678
0.695
0.967
0.905
0.888
DMF16 3.95 5 1 2 3 4 5
–
The Wilcoxon Signed-Ranks test was significant at the significance level of 0.05, suggesting the two compared variables were statistically different. DMF17 3.80 5 1 2 3 4 5
DMF18 3.80 5 1 2 3 4 5
DMF17
DMF19 3.85 5 1 2 3 4 5
0.000⁎
0.007⁎
0.242
0.114
0.357
0.367
0.448
0.650
0.711
0.770
0.981
1.000
DMF20 3.90 5 1 2 3 4 5
–
Total 100
DMF12
0.000⁎
0.012⁎
0.159
0.094
0.311
0.330
0.513
0.669
0.494
0.707
0.905
0.224
0.139
0.388
0.412
0.329
0.519
0.485
0.709
method and categorized them into five groups: plant location, en-
–
0.001⁎
0.026⁎
market attributes, and local regulations. Zhai, et al. [36] examined the
factors affecting the off-site production of housing construction in
DMF13
0.000⁎
0.008⁎
0.292
0.330
0.706
0.749
0.768
0.864
China and found that the six principal factors were regarding con-
–
0.000⁎
0.018⁎
0.385
0.351
0.646
0.582
0.788
to help identify the DMFs of PPVC. Gibb and Isack [37] investigated the
drivers that would affect the use of pre-assembly construction method
0.000⁎
0.016⁎
DMF7
0.275
0.432
0.975
these drivers were: less congestion on site, time, cost, quality and
productivity benefits through minimization of on-site operations and
Results of the Wilcoxon Signed-Ranks Test on DMFs of PPVC.
0.001⁎
DMF8
0.065
0.577
0.618
0.891
0.096
duced labor inputs, safer work environment, and the reduction of waste
–
and pollution. Azhar et al. [41] identified and assessed the constraints
associated with the adoption of modular construction, and the top six
0.012⁎
DMF1
DMF12
DMF17
DMF18
DMF14
DMF15
DMF11
Table 5
DMF2
DMF1
DMF4
DMF6
DMF8
DMF7
DMF3
DMF5
DMF9
171
B.-G. Hwang et al. Automation in Construction 94 (2018) 168–178
User Interface
Table 7 skills and experience, market demands, tools and standards, industry-
Proposed recommended action items. related issues, inflexibility, and project-specific issues. Chiang et al. [8]
Computed PPVC Score Action Items
investigated the barriers that hinder off-site construction method to be
adopted in the public housing construction sector of Hong Kong. These
< 60 (Basic) Consider the use of conventional construction method were identified to be: difficulties in transporting heavy and bulky pre-
60–80 (Intermediate) Revisit building design and modify accordingly if fabricated modules, gaining access to buildings for large modules, ad-
needed
Appoint PPVC specialist builders and experts with
ditional labor training, and poor jointing of prefabricated. Jaillon and
relevant experiences Poon [44] surveyed construction professionals in Hong Kong who ex-
Use past projects (local and/or overseas) as references pressed that site storage and access, increased time for design, and the
> 80 (Advanced) Ensure active participation and commitments of project lack of flexibility were the major limitations to the adoption of pre-
stakeholders in early stage
fabrication. Tam et al. [45] investigated the hindrances in applying
Arrange site visits and factory tours for better
understanding prior to construction and production prefabrication, and the top three hindrances were found to be inflexible
Appoint PPVC specialist builders and experts with for design changes, lack of research information, and higher initial
relevant experiences construction cost. Based on the comprehensive literature review pre-
Establish a consistent quality assurance and quality sented above, 32 DMFs affecting the adoption of PPVC were identified
control system to make sure that the modules
prefabricated overseas are of the required standard and
as shown in Table 1.
the predetermined dimensions.
2.4. KBDSS
between on-site demand and off-site construction, higher maintenance
Technically, KBDSS is an integration of an expert system and a de-
costs for end users caused by low utilization of expensive manu-
cision support system [53]. An expert system is a computer program
facturing resources, and inconsistent quality due to high variety of
that includes a knowledge base containing experts' knowledge for a
customized products and inefficient floor control.
particular problem domain and a reasoning mechanism for generating
Kamali and Hewage [42] reviewed the challenges for using the
inferences over the knowledge base, emulating the decision-making
modular construction method and found that the main challenges were
ability of a human expert [54]. In comparison, decision support system
attributable to project planning, transportation restraints, negative
is an interactive, computer-based information system that intended to
perceptions from the public, site constraints, high initial cost, and in-
help decision makers compile useful information from a combination of
creased coordination and communication required. Mao et al. [12]
raw data, documents, and personal knowledge, or business models to
examined the major barriers that impede the use of offsite construction
identify and solve problems and make decisions [55]. Integrating the
method in the Chinese construction industry. They found that the top
expert system with the decision support system, KBDSS is a computer
three major barriers were the absence of government incentives and
information system that provides information and methodological
regulations, the extremely high upfront cost, and practitioners' depen-
knowledge using analytical decision models to support decision makers
dence on conventional construction methods. Similarly, Rahman [43]
in making effective judgements in a complex and ill-structured problem
conducted a critical review to investigate the barriers to the im-
domain [56,57]. KBDSS has considerable merits, for example, it em-
plementation of modern construction methods (mostly are off-site
phasizes flexibility and adaptability to accommodate changes in the
construction methods). This study found that the barriers mainly lie in
environment and the decision-making approach of the user [58]. Fur-
eight perspectives including cost-related issues, motivation and culture,
thermore, it usually has a friendly interactive user interface which
172
B.-G. Hwang et al. Automation in Construction 94 (2018) 168–178
enables easy usage for even non-computer-proficient people [59]. based on the 19 DMFs, a questionnaire consisting of two sections was
Given these merits, KBDSS has been widely used to tackle a variety of developed. The first section was to record the respondents' background
tasks in the building and construction domain, including the selection information including their employer types and their years of experi-
of soil improvement method [60], site selection for residential land ence in the construction industry and the PPVC area. The second section
development [61], prequalification of design-builder [62], construction asked the respondents to assess the significance of the DMFs in deciding
enterprise risk management [53,56] and urban planning management the use of PPVC method, using a five-point rating scale (i.e., 1 = very
[63,64]. Moreover, KBDSS has also found applications in building en- insignificant, 2 = insignificant, 3 = slightly significant, 4 = significant,
velop assessment [65], variation order management [58], roadway and 5 = very significant).
safety analysis [66], and also home energy retrofits [67]. As a result, The potential respondents for the questionnaire are 100 institutions
KBDSS would be a good choice to developing a decision-making tool of (identified from the BCA Contractor Registration Systems) having ex-
PPVC. perience in prefabrication projects in Singapore [68]. Soft copies of the
questionnaire were sent to these institutions via email in January 2017.
3. Research methodology and data presentation To ensure a high response number, phone calls and email reminders
were sent every week if the dispatched questionnaires were not re-
The research process of this study consisted of the following six turned before the due date of the submission (31 January 2017). Fi-
steps. Step 1 was to conduct a comprehensive literature review to nally, 41 responses were received, representing a response rate of 41%.
identify the DMFs of PPVC. Step 2 was to carry out pilot interviews with It is noteworthy that this response rate is relatively high as the normal
industry experts to verify the DMFs identified by literature review. Step response rate for surveys in the construction industry is between 20 and
3 was to administer a questionnaire to industry practitioners in order to 30% [69]. The profiles of the respondents and their institutions were
capture their assessments of the verified DMFs. At Step 4, a PPVC presented in Table 3. It is noteworthy that the respondents were from
scoring approach that can assess the feasibility of using PPVC for a different types of institutions such as governmental agencies, devel-
given project was created, using the data collected by the questionnaire. opers, construction companies, consultancies, architectural firms,
At Step 5, a KBDSS-PPVC was developed in the platform of Microsoft quantity surveyors, and PPVC supplier/manufacturers. In addition, the
Excel, using the programing language of Microsoft Visual Basic for respondents were from different occupational backgrounds including
Applications (VBA) and the PPVC scoring approach created at Step 4. project manager, architect, quantity surveyor, engineer, and facility
Lastly, at Step 6, validation of the developed KBDSS-PPVC was per- manager. The diversity and heterogeneity of the respondent panel can
formed with industry experts. The six research steps are elaborated ensure the quality and reliability of the collected data and thereby yield
further below. some findings that are more reliable and convincing [70]. Although it
First, a comprehensive literature review was conducted to identify could be found from Table 3 that the majority of the respondents had
the DMFs of PPVC and subsequently 32 DMFs were identified, as pre- limited years of experience in PPVC projects, this could be explained by
sented in Table 1. Following the literature review, pilot interviews were the fact that PPVC was not introduced to Singapore until very recently
conducted with three industry experts having experiences in local PPVC [24].
projects to verify the identified DMFs. According to the results of the Using the data collected by the questionnaire, this study created a
pilot interviews, 19 out of 32 DMFs were found to be relatively more PPVC scoring approach that can assess the feasibility of using PPVC for
relevant to the decision-making of PPVC and thus were retained as the a given project. Based on the PPVC scoring approach created, this study
DMFs of PPVC, as presented in Table 2. This result supported the hy- then developed a KBDSS-PPVC in the platform of Microsoft Excel using
pothesis of this study, that there are some particular DMFs for PPVC the programing language of Microsoft Visual Basic for Applications
different from those for general off-site construction methods. Then, (VBA). The combined use of Microsoft Excel and Microsoft VBA were
173
B.-G. Hwang et al. Automation in Construction 94 (2018) 168–178
due to the following considerations. First, Microsoft Excel is a basic and Therefore, the four industry experts were familiar with PPVC and thus
essential office software widely available to the public and can thereby were qualified for the validation. During the validation, the four experts
make the Microsoft Excel-based KBDSS-PPVC easily accessible to mass were first asked to rate the 19 DMFs on a scale of 1 to 5 for their current
users [53]. Second, Microsoft VBA is a powerful programing language projects using the developed KBDSS-PPVC. Then, they were requested
that has been highly optimized to support rapid application develop- to comment on the results provided by the KBDSS-PPVC. Apart from
ment, and most importantly, it is the programing language unique to that, the experts were also requested to provide suggestions for possible
Microsoft Excel and other Office programs [71]. Furthermore, the Mi- enhancements to the KBDSS-PPVC.
crosoft Excel-based KBDSS-PPVC will update automatically whenever
Microsoft updates its Excel software, which can save programmers from
the monotonous work of maintaining and upgrading the software [72]. 4. Key DMFs for PPVC and development of PPVC score
After the KBDSS-PPVC was developed, four industry experts from
the Building Innovation Panel of Singapore were invited for system 4.1. Key DMFs for PPVC
validation. The Building Innovation Panel is a local government orga-
nization established to seek improvement in construction productivity In construction engineering and management research, mean score
through various innovative technologies including PPVC [73]. ranking is a widely used approach to prioritize variables [74,75]. Thus,
this study utilized mean score ranking technique to prioritize DMFs in
174
B.-G. Hwang et al. Automation in Construction 94 (2018) 168–178
this study as presented in Table 4. It was found that DMF2 “presence of 4.2. Development of the PPVC scoring approach
repetitive layout in design”, DMF1 “suitability of design for PPVC”,
DMF4 “structural stability of individual and assembled modules”, Following Caldas, et al. [83] who proposed a method to assess the
DMF6 “lead time for PPVC modules”, and DMF8 “site layout” were the implementation level of the best productivity practices in industry
top five key DMFs. Furthermore, a series of statistical analysis methods projects, the research team developed a PPVC scoring approach that can
were employed to analyze the data collected by the questionnaire. As a assess the feasibility of using PPVC for a given project. The PPVC score
large number of statistical analysis tests require data to fall in a normal was generated from the assessments of the 19 DMFs given to a specific
distribution, Shapiro-Wilk test was conducted first to check the nor- project. The maximum attainable PPVC score for the project was set to
mality of the data, following the suggestions of Chou, et al. [76] and 100, representing the project best suited for the adoption of PPVC
Gel, et al. [77]. According to the test results presented in Table 4, the method (i.e., all DMFs have an assessment of 5); while the lowest score
data were not normally distributed, implying that only non-parametric was set to 19 (i.e., all DMFs have an assessment of 1), representing the
statistical methods could be used for data analysis. As the respondents project completely unsuitable for PPVC.
of the questionnaire had different backgrounds, it is necessary to con- As indicated in Table 4 earlier, the 19 DMFs were perceived dif-
duct an inter-group comparison to check whether significant differences ferently by the respondents in terms of importance, suggesting that
exist among the respondents. As a result, Kruskal-Wallis test, a rank- each of the DMFs influenced the adoption of PPVC differently. Thus, the
based non-parametric statistical test method widely used to compare research team calculated the maximum PPVC scores of different DMFs
the differences between two or more groups [78,79], was conducted. in accordance with their mean scores in Table 4, following the approach
According to the test results in Table 4, no significant differences were of Caldas, et al. [83]. Specifically, the maximum PPVC score for the ith
found among the respondents, suggesting that the data could be treated DMF (i.e., Si5, the score corresponding to the assessment of 5) was
as a whole for further analysis [78]. Furthermore, following the re- calculated with its mean score (i.e., Mi) using Eq. (1).
commendation of Roberts, et al. [80], Wilcoxon Signed-Rank test was
175
B.-G. Hwang et al. Automation in Construction 94 (2018) 168–178
Mi was > 80, the decision support system would recommend some actions
Si5 = 19
× 100
∑i = 1 Mi (1) that could improve the implementation of PPVC method, such as en-
suring active participation and commitments of various project stake-
Then, a score of 1 was given to the assessment of 1 for all the DMFs. holders, arranging manufacturing plant tours and site visits, and ap-
After that, the scores for the assessments of 2, 3, and 4 were calculated pointing PPVC specialists. In this study, the majority of these
using Eqs. (2) to (4) successively. Lastly, the PPVC score of the given recommended actions were identified from existing off-site construc-
project was calculated using Eq. (5), i.e., by summing the scores of all tion research (e.g., [41,43,45,84–86]) and then modified to suit the
the 19 DMFs. To facilitate readers' understanding of the PPVC scoring context of the PPVC method. Apart from that, some actions were ob-
approach, Table 6 presented the PPVC scores of different DMFs at all tained from the validations of the KBDSS-PPVC after it was developed.
assessment levels.
Si5 − 1 5.2.3. User interface
Si2 = +1
4 (2) The user interface allowed users to interact with the KBDSS-PPVC. It
was Excel-based and consisted of two sections. The first section pro-
Si5 − 1
Si3 = + Si2 vided a brief introduction of the KBDSS-PPVC as well as a concise user
4 (3)
guide. In addition, it allowed users to key in the project profile in-
Si5 − 1 formation as shown in Fig. 2. The second section allowed the users to
Si 4 = + Si3
4 (4) assess the 19 DMFs in the context of their project using a five-point
Likert rating scale. It also reported the computed PPVC score for the
19
project as shown in Fig. 3. Furthermore, the second section contained a
PPVC Score = ∑ Si dialogue message box that displayed the different recommended action
i=1 (5)
items for the users based on the computed PPVC score, as shown in
Fig. 4.
5. KBDSS-PPVC
5.3. Validation
5.1. Objectives of the KBDSS-PPVC
As introduced earlier, four experts from the Building Innovation
The specific objectives of the KBDSS-PPVC were to provide an op- Panel of Singapore were invited to validate the developed KBDSS-PPVC.
portunity for a project team to assess the feasibility of using PPVC for a The experts were working on three different building construction
given project and provide the project team with some corresponding projects at the time of validation. Specifically, Experts A and B were
recommendations based on the assessment results. It is expected that working on a residential building construction project at the Nanyang
this KBDSS-PPVC would aid the project team in obtaining a firm un- Technological University, Singapore (NTU Project for short thereafter),
derstanding of the suitability of their project for PPVC, thereby facil- which had adopted the PPVC method. Expert C was working on a multi-
itating their decision-making on the adoption of PPVC. Furthermore, if purpose sports hall construction project at the National University of
the project team decides to use PPVC, it is also expected that the re- Singapore (NUS Project for short thereafter) which did not adopt PPVC
commendations provided by the KBDSS-PPVC could also help the method. Expert D was working on a condominium construction project
project team achieve a better implementation of PPVC in their project. (Condominium Project for short thereafter) which did not adopt PPVC
either. The experts were requested to rate their projects using the de-
5.2. Architecture of the KBDSS-PPVC veloped KBDSS-PPVC. Table 8 presented the PPVC scores computed by
the KBDSS-PPVC. It was noted that the scores obtained from Experts A
The developed KBDSS-PPVC consisted of three main components: a and B were > 80, indicating the suitability of PPVC and the readiness of
knowledge base, a decision support system, and a user interface, as the project team to embrace the method. Conversely, the scores ob-
shown in Fig. 1. tained from Experts C and D were < 60, suggesting the inappropriate-
ness of using PPVC method and the preference for the conventional
5.2.1. Knowledge base construction method. It was noted that the assessment results were in
The knowledge base was a repository of the knowledge and ex- line with the project teams' choices in reality, indicating that the de-
perience of experts about the PPVC method. It included the 19 DMFs veloped KBDSS-PPVC was able to provide reliable recommendations for
identified by the literature review and pilot interviews. It also included the users in the industry.
a group of recommended action items identified from existing litera- Besides, the experts' opinion was elicited on the design of KBDSS-
ture. Furthermore, some recommendations obtained during the sub- PPVC, alongside a request for their suggestions for further enhance-
sequent system validation exercise were also included in the knowledge ments. All the experts recognized the developed KBDSS-PPVC and said
base. it did help them develop a better understanding of what they lack for
the implementation of PPVC and what can be done before the im-
5.2.2. Decision support system plementation. The experts also agreed with the proposed re-
Based on the project team's inputs to the 19 DMFs, the decision commendations in Table 7, especially the active participation among
support system would calculate the PPVC score for the given project project teams and factory tours. In addition to the compliment, the
using the scales listed in Table 6. It would also provide corresponding experts also provided some comments and suggestions. Expert A em-
action plans based on the computed PPVC score. The computed PPVC phasized the importance of appointing manufacturers with capability,
score was categorized into three levels by the decision support system and Expert B suggested that a consistent quality assurance and quality
as shown in Table 7. If the computed PPVC score was < 60, the decision control system should be established to ensure modules prefabricated
support system would recommend the project to use conventional overseas are of the required standard and predetermined dimensions.
construction method instead of PPVC method. If the computed PPVC
score was between 60 and 80, the decision support system would re- 6. Conclusions and recommendations
commend the project team: to revisit and modify the building design,
making it more suitable for PPVC method; to refer more to the similar This study addressed the decision-making of PPVC, a typical type of
past projects using PPVC method; and to appoint PPVC specialists to off-site construction method. Based on a comprehensive literature re-
help the implementation of PPVC method. If the computed PPVC score view and pilot interviews with three experienced experts in the
176
B.-G. Hwang et al. Automation in Construction 94 (2018) 168–178
177
B.-G. Hwang et al. Automation in Construction 94 (2018) 168–178
[38] W. Pan, A.G. Gibb, A.R. Dainty, Perspectives of UK housebuilders on the use of [63] H.M. Osman, T.E. El-Diraby, Knowledge-enabled decision support system for
offsite modern methods of construction, Constr. Manag. Econ. 25 (2) (2007) routing Urban Utilities, J. Constr. Eng. Manag. 137 (3) (2011) 198–213, http://dx.
183–194, http://dx.doi.org/10.1080/01446190600827058. doi.org/10.1061/(ASCE)CO.1943-7862.0000269.
[39] W. Pan, A.G. Gibb, A.R. Dainty, Leading UK housebuilders' utilization of offsite [64] Y. Wang, Z. Zou, Spatial decision support system for urban planning: case study of
construction methods, Build. Res. Inf. 36 (1) (2008) 56–67, http://dx.doi.org/10. Harbin City in China, J. Urban Plann. Dev. 136 (2) (2010) 147–153, http://dx.doi.
1080/09613210701204013. org/10.1061/(ASCE)0733-9488(2010)136:2(147.
[40] P.S.P. Wong, C. Zwar, E. Gharaie, Examining the drivers and states of organizational [65] N. Singhaputtangkul, S.P. Low, A.L. Teo, B.-G. Hwang, Knowledge-based decision
change for greater use of prefabrication in construction projects, J. Constr. Eng. support system quality function deployment (KBDSS-QFD) tool for assessment of
Manag. 143 (7) (2017) 04017020, , http://dx.doi.org/10.1061/(ASCE)CO.1943- building envelopes, Autom. Constr. 35 (Supplement C) (2013) 314–328, http://dx.
7862.0001309. doi.org/10.1016/j.autcon.2013.05.017.
[41] S. Azhar, M.Y. Lukkad, I. Ahmad, An investigation of critical factors and constraints [66] J.-H. Jo, J.-S. Lee, Y. Ouyang, F. Peng, Integrated decision support for roadway
for selecting modular construction over conventional stick-built technique, Int. J. safety analysis, J. Comput. Civ. Eng. 25 (1) (2010) 50–56, http://dx.doi.org/10.
Constr. Educ. Res. 9 (3) (2013) 203–225, http://dx.doi.org/10.1080/15578771. 1061/(ASCE)CP.1943-5487.0000066.
2012.723115. [67] M. Syal, D. Duah, S. Samuel, M. Mazor, Y. Mo, T. Cyr, Information framework for
[42] M. Kamali, K. Hewage, Life cycle performance of modular buildings: a critical re- intelligent decision support system for home energy retrofits, J. Constr. Eng. Manag.
view, Renew. Sust. Energ. Rev. 62 (2016) 1171–1183, http://dx.doi.org/10.1016/j. 140 (1) (2014) 04013030, , http://dx.doi.org/10.1061/(ASCE)CO.1943-7862.
rser.2016.05.031. 0000773.
[43] M.M. Rahman, Barriers of implementing modern methods of construction, J. [68] S. BCA, BCA Contractors Registration System (CRS), https://www.bca.gov.sg/
Manag. Eng. 30 (1) (2013) 69–77, http://dx.doi.org/10.1061/(ASCE)ME.1943- ContractorsRegistry/others/CRS_FAQ.pdf, (2017) , Accessed date: 5 December
5479.0000173. 2017.
[44] L. Jaillon, C.S. Poon, Design issues of using prefabrication in Hong Kong building [69] R. Jia, Q. Li, X. Deng, X. Zhao, J. Yuan, Entry mode taxonomy and choice of Chinese
construction, Constr. Manag. Econ. 28 (10) (2010) 1025–1042, http://dx.doi.org/ international construction companies, J. Manag. Eng. (2016) 04016058, , http://dx.
10.1080/01446193.2010.498481. doi.org/10.1061/(ASCE)ME.1943-5479.0000506.
[45] V.W.Y. Tam, C.M. Tam, S.X. Zeng, W.C.Y. Ng, Towards adoption of prefabrication [70] D. Cao, H. Li, G. Wang, X. Luo, X. Yang, D. Tan, Dynamics of project-based colla-
in construction, Build. Environ. 42 (10) (2007) 3642–3654, http://dx.doi.org/10. borative networks for BIM implementation: analysis based on stochastic actor-or-
1016/j.buildenv.2006.10.003. iented models, J. Manag. Eng. (2016) 04016055, , http://dx.doi.org/10.1061/
[46] U. Construction Industry, Institute, MODEX, Automated decision support system for (ASCE)ME.1943-5479.0000503.
modular construction, https://www.construction-institute.org/scriptcontent/ [71] Z. Zhou, Q. Li, W. Wu, Developing a versatile subway construction incident data-
more/29_1_more.cfm, (1992) , Accessed date: 10 May 2018. base for safety management, J. Constr. Eng. Manag. 138 (10) (2012) 1169–1180,
[47] N. Blismas, M. Pendlebury, A. Gibb, C.L. Pasquire, Constraints to the use of off-site http://dx.doi.org/10.1061/(ASCE)CO.1943-7862.0000518.
production on construction projects, Archit. Eng. Des. Manag. 1 (3) (2005) [72] P. Pishdad-Bozorgi, M. de la Garza Jesus, B. Austin Robert, Readiness assessment for
153–162, http://dx.doi.org/10.1080/17452007.2005.9684590. flash tracking, J. Constr. Eng. Manag. 142 (12) (2016) 06016005, , http://dx.doi.
[48] R. Scofield, S. Wilkinson, R. Potangaroa, F. Rotimi, Driving Innovative Offsite org/10.1061/(ASCE)CO.1943-7862.0001190.
Construction Techniques in New Zealand, http://12219-portal.memberconnex. [73] S. BCA, Building Innovation Panel, https://www.bca.gov.sg/cpc/BIP.html, (2017) ,
com/PDFs/CSA/Driving%20Innovative%20Offsite%20Construction Accessed date: 27 March 2017.
%20Techniques%20in%20NZ.pdf, (2009) , Accessed date: 10 May 2018. [74] C.K. Hon, A.P. Chan, M.C. Yam, Empirical study to investigate the difficulties of
[49] C. McGraw Hill, Prefabrication and Modularization: Increasing Productivity in the implementing safety practices in the repair and maintenance sector in Hong Kong,
Construction Industry, https://www.nist.gov/sites/default/files/documents/el/ J. Constr. Eng. Manag. 138 (7) (2011) 877–884, http://dx.doi.org/10.1061/
economics/Prefabrication-Modularization-in-the-Construction-Industry-SMR- (ASCE)CO.1943-7862.0000497.
2011R.pdf, (2011) , Accessed date: 10 May 2018. [75] Y. Ke, S. Wang, A.P. Chan, E. Cheung, Understanding the risks in China's PPP
[50] E. Elnaas, The Decision to Use Off-Site Manufacturing (OSM) Systems for House projects: ranking of their probability and consequence, Eng. Constr. Archit. Manag.
Building Projects in the UK, http://eprints.brighton.ac.uk/14673/, (2014) , 18 (5) (2011) 481–496, http://dx.doi.org/10.1108/09699981111165176.
Accessed date: 10 May 2018. [76] Y.M. Chou, A.M. Polansky, R.L. Mason, Transforming non-normal data to normality
[51] J.-S. Lee, Y.-S. Kim, Analysis of cost-increasing risk factors in modular construction in statistical process control, J. Qual. Technol. 30 (2) (1998) 133–141, http://dx.
in Korea using FMEA, KSCE J. Civ. Eng. 21 (6) (2017) 1999–2010 https://link. doi.org/10.1080/00224065.1998.11979832.
springer.com/article/10.1007/s12205-016-0194-1. [77] Y.R. Gel, W. Miao, J.L. Gastwirth, Robust directed tests of normality against heavy-
[52] G. Liu, K. Li, D. Zhao, C. Mao, Business model innovation and its drivers in the tailed alternatives, Comput. Stat. Data Anal. 51 (5) (2007) 2734–2746, http://dx.
Chinese construction industry during the shift to modular prefabrication, J. Manag. doi.org/10.1016/j.csda.2006.08.022.
Eng. 33 (3) (2017) 04016051, , http://dx.doi.org/10.1061/(ASCE)ME.1943-5479. [78] E.E. Ameyaw, Y. Hu, M. Shan, A.P. Chan, Y. Le, Application of Delphi method in
0000501. construction engineering and management research: a quantitative perspective, J.
[53] X. Zhao, B.-G. Hwang, S.P. Low, An enterprise risk management knowledge-based Civ. Eng. Manag. 22 (8) (2016) 991–1000, http://dx.doi.org/10.3846/13923730.
decision support system for construction firms, Eng. Constr. Archit. Manag. 23 (3) 2014.945953.
(2016) 369–384, http://dx.doi.org/10.1108/ECAM-03-2015-0042. [79] A. Tixier, M. Hallowell, A. Albert, L. van Boven, B. Kleiner, Psychological ante-
[54] S.-H. Liao, Expert system methodologies and applications—a decade review from cedents of risk-taking behavior in construction, J. Constr. Eng. Manag. 140 (11)
1995 to 2004, Expert Syst. Appl. 28 (1) (2005) 93–103, http://dx.doi.org/10.1016/ (2014) 04014052, , http://dx.doi.org/10.1061/(ASCE)CO.1943-7862.0000894.
j.eswa.2004.08.003. [80] A.C. Roberts, G.I. Christopoulos, J. Car, C.-K. Soh, M. Lu, Psycho-biological factors
[55] H. Guler, Decision support system for railway track maintenance and renewal associated with underground spaces: what can the new era of cognitive neu-
management, J. Comput. Civ. Eng. 27 (3) (2013) 292–306, http://dx.doi.org/10. roscience offer to their study? Tunn. Undergr. Space Technol. 55 (2016) 118–134,
1061/(ASCE)CP.1943-5487.0000221. http://dx.doi.org/10.1016/j.tust.2015.12.016.
[56] J.Y. Liu, S.P. Low, Developing an organizational learning-based model for risk [81] L. Lund Research, Wilcoxon Signed-Rank Test using SPSS Statistics, https://
management in Chinese construction firms: a research agenda, Disaster Prev Manag statistics.laerd.com/spss-tutorials/wilcoxon-signed-rank-test-using-spss-statistics.
18 (2) (2009) 170–186, http://dx.doi.org/10.1108/09653560910953243. php, (2013) , Accessed date: 8 July 2017.
[57] C. Zopounidis, M. Doumpos, N.F. Matsatsinis, On the use of knowledge-based de- [82] P. Wu, S.P. Low, J.Y. Liu, J. Pienaar, B. Xia, Critical success factors in distance
cision support systems in financial management: a survey, Decis. Support. Syst. 20 learning construction programs at Central Queensland University: students' per-
(3) (1997) 259–277, http://dx.doi.org/10.1016/S0167-9236(97)00002-X. spective, J. Prof. Issues Eng. Educ. Pract. 141 (1) (2014) 05014003, http://dx.doi.
[58] F.M. Arain, L.S. Pheng, Knowledge-based decision support system for management org/10.1061/(ASCE)EI.1943-5541.0000217.
of variation orders for institutional building projects, Autom. Constr. 15 (3) (2006) [83] C.H. Caldas, J.-Y. Kim, C.T. Haas, P.M. Goodrum, D. Zhang, Method to assess the
272–291, http://dx.doi.org/10.1016/j.autcon.2005.06.005. level of implementation of productivity practices on industrial projects, J. Constr.
[59] G. Alexouda, A user-friendly marketing decision support system for the product line Eng. Manag. 141 (1) (2014) 04014061, http://dx.doi.org/10.1061/(ASCE)CO.
design using evolutionary algorithms, Decis. Support. Syst. 38 (4) (2005) 495–509, 1943-7862.0000919.
http://dx.doi.org/10.1016/j.dss.2003.09.002. [84] A.S. Hanna, G. Mikhail, K.A. Iskandar, State of prefab practice in the electrical
[60] J.L. Chameau, J.C. Santamarina, Knowledge-based system for soil improvement, J. construction industry: qualitative assessment, J. Constr. Eng. Manag. 143 (2)
Comput. Civ. Eng. 3 (3) (1989) 253–267, http://dx.doi.org/10.1061/(ASCE)0887- (2017), http://dx.doi.org/10.1061/(ASCE)CO.1943-7862.0001236.
3801(1989)3:3(253. [85] H. Said, Prefabrication best practices and improvement opportunities for electrical
[61] C. McIntyre, M.K. Parfitt, Decision support system for residential land development construction, J. Constr. Eng. Manag. 141 (12) (2015), http://dx.doi.org/10.1061/
site selection process, J. Archit. Eng. 4 (4) (1998) 125–131, http://dx.doi.org/10. (ASCE)CO.1943-7862.0001018.
1061/(ASCE)1076-0431(1998)4:4(125. [86] T. Wang, Y. Li, L. Zhang, G. Li, Case study of integrated prefab accommodations
[62] E. Palaneeswaran, M. Kumaraswamy Mohan, Web-based client advisory decision system for migrant on-site construction Workers in China, J. Prof. Issues Eng. Educ.
support system for design–builder prequalification, J. Comput. Civ. Eng. 19 (1) Pract. 142 (4) (2016), http://dx.doi.org/10.1061/(ASCE)EI.1943-5541.0000288.
(2005) 69–82, http://dx.doi.org/10.1061/(ASCE)0887-3801(2005)19:1(69.
178