You are on page 1of 10

Automation in Construction 48 (2014) 64–73

Contents lists available at ScienceDirect

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

Supplier selection process in an integrated construction materials


management model
Mahdi Safa a,⁎, Arash Shahi a, Carl T. Haas a, Keith W. Hipel b
a
Department of Civil and Environmental Engineering, University of Waterloo, 200 University Avenue West, Waterloo, Ontario N2L 3G1, Canada
b
Department of Systems Design Engineering, University of Waterloo, 200 University Avenue West, Waterloo, Ontario N2L 3G1, Canada

a r t i c l e i n f o a b s t r a c t

Article history: The procurement and management of construction materials involve challenges related to reducing inventory,
Received 23 October 2013 speeding delivery, and increasing the control of materials, thus decreasing the overall project cost. The objective
Received in revised form 2 July 2014 of this research was to define and develop an integrated construction materials management (ICMM) model to
Accepted 20 August 2014
address these challenges by deploying principles of virtual inventory management, feasible materials manage-
Available online 16 September 2014
ment networks, and a supplier selection process. Contributions include the development of the ICMM model
Keywords:
and the demonstrated potential of the selection supplier process for improving procurement for construction
Construction material management projects. The use of the supplier selection process has been demonstrated through implementation on an indus-
Supplier selection process trial project and using the TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) method. This
BOM process optimizes and validates purchasing at each stage of fabrication for each construction package. A variety of
Decision making project-specific criteria are involved in the supplier selection process, including price, lead-time, cash rebate, and
Optimization supplier performance.
© 2014 Elsevier B.V. All rights reserved.

1. Introduction These elements of materials management systems are associated pri-


marily with the purchasing process. Purchasing usually involves tasks
Construction materials management can best be defined as “the such as selecting and awarding suppliers, analyzing proposals,
planning and controlling of all activities necessary for ensuring and delivering purchases, inspecting goods supplied, and maintaining a
confirming that the correct and accurate quantity and quality of variety of purchasing records [9]. This paper introduces a supplier selec-
materials and equipment are appropriately specified in a timely tion procedure for facilitating the purchasing process. The TOPSIS
manner, are obtained at a reasonable cost, and are available when method has been used to demonstrate the new procedure for an
needed” [1]. In typical construction projects, the cost of materials can industrial project.
constitute 50% to 60% of the total cost of the project [2–5], and their Two factors associated with construction materials management
management affects 80% of the project schedule [6]. Strong evidence should be noted. The first is that the specification of an integrated
shows that an effective and optimized materials management system construction materials management system should be conducted over
can ensure the availability of sufficient quantities of materials and a reasonable period and as a strategic decision during the Front End
equipment for construction needs and the minimization of surpluses Planning (FEP) phase [9]. The second is related to the fact that materials
at the end of the project [7,8]. Because of their role as important and management is affected by the “temporary multi-organization” [10]
complex functions of a project, construction materials management nature of uniquely defined and executed projects [11,12] that are
systems and their elements require substantial improvement [4], characteristic of the construction industry. As with other construction
creating significant demand for both their enhancement or amendment systems, any supplier selection process developed should therefore be
and the development of new applications. flexible enough to meet the materials management requirements for a
The CII Materials Management Handbook lists critical elements that wide variety of projects. The process presented in this paper was
can cause shortages of materials: lost or damaged materials, materials created with this goal in mind.
required but not purchased, materials purchased but not received, Section 2 provides a review and summary of the scholarly works
errors in material takeoff, variances in additional material requirements, whose concepts informed the research presented in this paper. The
and materials that are issued to crafts and are not used or installed [8]. section is arranged thematically so that it corresponds to the steps
involved in the developed supplier selection process. The discussion
covers a synthesis of current knowledge as well as the gaps between
⁎ Corresponding author. existing construction materials management practices and the desired
E-mail address: msmahdisafa@gmail.com (M. Safa). model, which were subsequently addressed through the development

http://dx.doi.org/10.1016/j.autcon.2014.08.008
0926-5805/© 2014 Elsevier B.V. All rights reserved.
M. Safa et al. / Automation in Construction 48 (2014) 64–73 65

of an integrated construction materials management model (ICMM) for 2.2. Site materials management
the industrial construction sector. While some articles and studies were
drawn on more heavily than others, and some provided a direction for In recent years, there has been an increased research interest in the
reflection rather than actual concepts, all of the work cited provided in- materials management theories and methods. These researches recom-
sights that proved valuable in the development of the model. Section 3 mended an integrated management of the interface between site activ-
presents the vision of this research with respect to the development of ities, the computer systems that are used for design and scheduling, and
the ICMM based on the needs and knowledge gaps identified through other components of the integrated materials management model
the literature review. Individual elements of the developed ICMM are [21–23]. The potential benefits to the integrated materials management
also described. In Section 4, the approach taken during the creation of include: high accuracy, low cost, and robustness [24]. The attributes of
the supplier selection process is explained, and Section 5 discusses the materials management systems that contribute to the success of a
implementation of the new process in an organization for a specific project were identified by some scholars and experts [8,25,26]. They
project, based on a real case study. Section 6 presents the conclusions associated their findings related to the attributes of materials manage-
that can be drawn from our findings. ment systems for mega and complex industrial construction projects
with the control of the functions related to quantity takeoff, vendor
evaluation, purchasing, expediting, receiving, warehousing, and distri-
2. Important concepts and characteristics of materials management
bution. These findings showed that lack of proper management of
these functions leads to shortages and surpluses of materials and to
Construction materials management is a multi-disciplinary field that
cash flow problems.
is appropriate for and can build on the correspondence between con-
Bell and Stukhart (1995) quantified the savings in the areas of
struction processes and some aspects of the manufacturing industry
improved labor productivity, reduced surpluses of bulk materials,
that entail a variety of activities for the large-scale conversion of raw
fewer manpower requirements for materials management, and
materials into finished products (fabricated products). Thus, some
enhanced cash flow. Surpluses of bulk materials were reduced from a
manufacturing principles may be appropriately applied to construction
range of 5% to 10% of the amount purchased to about 1% to 3% of the
as staged-fabrication.
amount purchased. This study also showed that an efficient materials
Because EPC (Engineering, Procurement and Construction) entities
management system can reduce the man-hours needed for materials
and construction companies often cannot perform all required project
management; in the absence of a materials management system, craft
tasks on-site in an efficient and timely manner, construction managers
foremen spent up to 20% of their time searching for materials and
(experts) have tended to use staged-fabrication technologies whereby
another 10% tracking purchase orders (POs) and expediting delivery
operations traditionally performed on-site have been moved to a
[25].
manufacturing environment. Staged fabrication offers an opportunity
to overcome some of the challenges associated with construction mate-
2.3. Planning and control for materials management
rials management, such as the absence of site inventories, demanding
schedules, permit requirements, and adverse site conditions [13,14].
One of the most important elements of any effective construction
The staged-fabrication method is addressed in the next section, follow-
materials management system is planning [3]. Materials planning inte-
ed by a discussion of site materials management and other relevant
grates and combines the functions of vendor enquiry and evaluation,
concepts.
project planning, purchasing, material takeoff, transportation, field ma-
terials control, and warehousing [3]. Significant cost overruns can be
2.1. Materials management for staged fabrication (off-site) caused by shortages of materials resulting from lack of proper materials
planning and control. Improper materials planning and control result
Staged fabrication is a method of working with contractors so in: the absence of adequate statistics related to the availability of the
that value is added at the appropriate “stage” in the supply chain, materials, a shortage of funds for procurement (budgeting), inadequate
and the resulting products such as pipe spools and precast concrete transportation capacities, excessively long average waiting times and
elements are efficiently delivered to a construction site. The value uncertainty about the delivery of the materials ordered and about the
of such techniques includes the possibility of enhanced quality availability of some materials [8].
control, improved design details, compressed schedules, and Transparency among the suppliers, subcontractors and owners is
reduced reliance on site labor. These benefits come at an increased important for materials planning [27]. The subcontracting concept is
cost for some projects, but the additional cost could decrease over crucial in the materials management process, which should be consid-
time, as the construction industry becomes familiar with the ered over developing and implementing the ICMM. The scope of the
staged-fabrication process [15]. Already, the application of subcontracts and lists of equipment and materials to be furnished by
staged-fabrication principles to construction materials manage- subcontractors are defined during the planning stage [28,29]. The criti-
ment is beginning to demonstrate benefits with respect to quality, cal materials are usually procured by the owner (or an EPC), while the
time, and cost savings [16–18]. Materials management for staged minor equipment and bulk materials may be furnished by the subcon-
fabrication is concerned with both the input and output of the tractors. In general, subcontracting requires considerable effort on the
fabrication process. tasks of materials procurement and scheduling but relatively less
Implementation of staged-fabrication requires the collaboration of emphasis on tasks related to the on-site materials management [30,31].
clients, developers, designers, contractors, manufacturers, suppliers, Transportation planning, materials specification and a coding sys-
government agencies, advisors, and researchers. A variety of staged- tem are important functions of construction materials planning.
fabrication strategies are commonly used in the construction industry, Transportation planning should include consideration of cost, safety,
including semi-prefabrication, prefabrication, comprehensive, and timely delivery, and hazardous materials content. A materials specifica-
modular [19]. With respect to transportation considerations, tion and coding system must first be established for handling bulk ma-
prefabricated elements have a higher added value than bulk materials. terials. A materials takeoff establishes the quantity of bulk and
The materials to be delivered are, in fact, manufactured products with engineered items needed for the project, based on the design. However,
value added by a separate workforce, and the percentage of the total the creation of composite items requires that they be quantified and
costs represented by the amount required for direct labor and materials arranged based on their relationships, therefore a Bill of Materials
for a project may thus change if additional fabricated products are (BOM) must be created from the materials takeoff in order to define
incorporated into the construction project [20]. the requirements with respect to project materials.
66 M. Safa et al. / Automation in Construction 48 (2014) 64–73

2.4. Labor productivity and cost impact of materials management based on their specific requirements, the research presented in this
paper shows that an automated/integrated supplier selection process
The review of construction materials management on major capital with an appropriate degree of integration has yet been developed or
projects demonstrates that the absence of materials on-site is one of suggested in the literature for the construction industry. The developed
the most commonly experienced causes of delays and a significant driv- integrated construction materials management (ICMM) model as the
er of overrun costs [4,32–36]. On the other hand, costs associated with foundation of the developed process, as well as its supplier selection
developing and implementing materials management systems are and automation processes, is further discussed in the following section.
substantial. Growing evidence shows that an efficient materials man-
agement system can ensure that the quantities of materials available 3. Integrated construction materials management (ICMM) model
are sufficient for construction needs and can minimize surpluses at
the end of the project, thus leading to large cost savings [8]. Such a The above discussion clearly establishes the need for an integrated
system creates an interface between the most important contractors materials management model and provides an understanding of the
and the suppliers, which can decrease the average cost of materials by fundamental concepts related to construction materials management
up to 10% [35,36]. The appropriate supervision of the flow of materials systems, including their benefits and deficiencies. This section explains
may also play a substantial role in enhancing the effectiveness of a how this background informed the development of an ICMM model
contractor [37–39]. that can play a critical role in the success of a construction project. The
Labor cost is one of the key factors that impacts the overall project core function or process of such a system model is the supplier selection
cost. In one study 27.7% of the time that craft labor was idle was because process, which is the focus of the research presented in this paper.
of the unavailability of the required materials, resulting in costly delays An ICMM model has the potential to improve construction supply
[40]. The unavailability of materials and resources has reportedly had a management processes and thereby lead to enhanced project perfor-
negative impact on craft labor productivity by increasing the related mance. Such a model would help ensure that selected materials are
amount of time spent on craft labor activities [41]. delivered to the construction site, in a manner that adheres to project
requirements such as price, lead-time, cash rebates, supplier perfor-
2.5. Waste of materials mance, and preferred supplier early payments. The goal of developing
this type of model is to reduce inventory and waste with respect to
The minimization of waste (surplus) with respect to construction on-site materials, speed up delivery, and provide increased control of
materials is a significant issue in many projects worldwide. Experts materials, all of which will decrease the cost of a construction project.
recommend that when designs are being prepared, designers employ In the ICMM model defined here, project managers can view and
a system that has less potential to produce site waste [42]. Construction forecast the anticipated arrival dates of goods and are then presented
experts have established a set of waste-prevention strategies for con- with consistent and clear shipping documentation so that they can suc-
struction projects. These strategies involve the effective coordination cessfully complete the receiving process.
of materials management processes such as: the well-organized A critical aspect of this model is based on the concept of construction
purchase and proficient ordering of materials, an effective delivery packaging. Construction packaging is an emerging paradigm suited to
schedule, effectual storage, the maximization of reuse, and the preven- the complex and dynamic processes inherent in a construction project
tion of rework [43]. The implementation of an integrated materials [50,51]. Construction packages can be defined for specific stages of fab-
management model would provide a valuable opportunity to reduce rication according to the nature of a project. Such staged-fabrication
the amount of materials wasted and to increase the recovery of those processes require early collaboration between the construction project
that would be otherwise be wasted. team and other project stakeholders because the knowledge provided
by specialists and experts and the information stored in the historical
2.6. BOM database are crucial for these processes. Fig. 1 illustrates the progression
of stages in a typical staged-fabrication process. In staged-fabrication
A BOM is one of the principal elements of any materials manage- projects, the output of a construction package can be a prefabricated
ment system. In general, The BOM is a technical document that clarifies unit, a subassembly, or a module. The term “stage” refers primarily to
the composition of a module (or a facility). BOMs determine the struc- the idea of progressive assembly into interim products that together
ture and relationship between the module, sub-assembly, prefabricated form the final facility, but a “stage” can also occur at different locations
units, as well as the corresponding quantities of the subordinate items and at varying points in the supply chain or in the project schedule.
and materials of each assembly [44]. Between the primary products The following are three contrasting examples:
and the top products is a category termed subassemblies [45]. In
1- Prefabricated unit: prefabricated wood frame, pre-stressed rein-
construction, the concept of subassemblies is related to staged-
forced concrete element, precast elements, pipe spools, and
fabrication.
prefabricated structural steel frame.
In manufacturing, a generic BOM describes all variants of all compo-
2- Subassembly: purified water storage system, cantilevered deck of an
nents and makes it possible to construct a specific BOM for a specific
elevated bridge, decks for a long-span bridge, and platform for a
product variant [46]. Similar components can be grouped as one generic
railway station.
component, so that each component description would only exist once
3- Module: pipe module, modular building, and electrical modular box.
in the generic structure, but the variant set of a component would be
described independently. A generic BOM system requires the selection In this paper, these types of output are referred to generally as
of the specifications for the generation of the required variant [47]. “fabricated products,” regardless of the stage of fabrication, and in
BOMs have been used in other industries to address the challenges order to avoid confusion due to a multitude of terms. It should be
associated with materials management. For example, Zhang et al. noted that any fabricated product of a construction package might con-
(2011) designed an auto-parts data management system based on the tain more than one BOM; however, assigning only one BOM to a fabri-
BOM [48]. Lee et al. (2012) integrated the BOM with the design of cated product is more desirable and easier to manage. All BOMs
ship-outfitting equipment and showed that the architecture of a BOM contain design “fabricated components” that are listed in the schematic
can be used to define the relationships among parts, assemblies, and source documents and also include supplementary items that could be
an overall view of a BOM that evolves with ship design [49]. added manually at any stage of fabrication.
While some construction companies may use in-house or commer- The ICMM model defined as part of this research was developed
cial systems to meet their day-to-day materials management needs with the goal of overcoming the manual and inefficient nature of the
M. Safa et al. / Automation in Construction 48 (2014) 64–73 67

reliable suppliers [53]. The ICMM model eliminates the necessity of


manually updating a BOM and takes into account both cost and time
constraints along with any other criteria that must be evaluated. The
supplier selection process described in the remainder of this paper facil-
itates the critical integration of all of these considerations and thus plays
a key role in achieving shorter lead times, reduced costs, and higher
quality [52].

4. Supplier selection process

Today's complex global market and the advancement of enabling


information technologies require a robust framework covering all facets
of material management. Herein, the configuration of the developed
model reflects the expertise gained from extensive surveys, discussion
with experts, and studies of several construction projects in the North
America. The ICMM model illustrated in Fig. 2 consists of six main
components: a warehouse, the selection process, work package compo-
nent, purchase order, supplier's information, and decision component.
The inputs of the ICMM model include updated supplier information,
generic BOM structures, and criteria weighting factors. A simplified
architecture of the ICMM model is depicted in Fig. 2.
Even though there was substantial effort devoted to completing the
programming and implementing the logics of the developed model, the
most critical challenges in implementing the model was defining the
integration architecture and interfaces. All distributed data sources
along the ICMM model must be properly connected and data is
Fig. 1. Flow of staged-fabrication and terms for stages.
retrieved from one database (e.g. warehouse) should be able to easily
integrate with another database (e.g. project planning). Hence, the
MATLAB programming was employed in this study as it can be integrat-
ed with Excel (purchase department's information and suppliers' data
source), Oracle (used for work packages), and SAP (the warehouse
order process by enabling the identification of an appropriate BOM that database).
results in the total materials cost and primary product availability that Most construction capital projects face uncertainty in the estimates
will lead to the most overall cost-effective alternative for a particular of the performance ratings and criteria weights due to the subjectivity
product. The ICMM model is intended to be suitable for internal parties of procurement experts' judgments [54,55]. Selection of appropriate
related to a project, such as the site management team, warehouses, the criteria influences the supplier selection process, when, conflicting
financial department, and the purchasing department as well as relationships exist between the criteria and qualitative criteria are
external parties, such as suppliers, fabricators, and transportation com- included [56]. Hence, the criteria for evaluating potential suppliers and
panies; however, the following discussion relates to the personnel also the weight of each individual criterion should be provided by the
responsible for the supply chain, who normally determine require- project management team prior to procurement process inception. A
ments from a number of sources. Materials may be identified from a further requirement is the purchasing department order (initiated by
request, a planning inquiry, a schedule review, or a restocking demand. site manager/supervisor), which indicates the required quantity of a
The purchasing (supply chain) department can employ the defined target product (output(s) of a work package). This number is multiplied
model as a means of ensuring that timely and cost-effective procure- by the amount listed for that product in the generic requisition order so
ment practices are followed and also that appropriate tasks are that a final requisition order can be produced for the acquisition of
addressed proactively prior to procurement. These tasks include deter- materials for a specific project. A detailed discussion of this process is
mining shipping requirements and costs, creating and maintaining provided in the subsequent sections, including decision support.
accurate BOMs, ensuring that warehouse procedures are established at Several selection methods and multi-criteria decision making
the facility, ascertaining procurement requirements, determining approaches have been developed for supplier selection, such as multi-
whether materials are stocked, establishing a process for evaluating criteria utility theory models [57], the simple multi-attribute rating
critical items, and setting up a procedure for verifying and coordinating technique (SMART), the Compromise Ranking method (VIKOR) [58],
the availability and cost of materials through monthly system updates. evidential reasoning [59], fuzzy set theory [60], the analytic hierarchy
An important function implemented in the defined ICMM model is a process (AHP), the analytic network process (ANP), case-based reason-
software algorithm that generates the best possible purchasing strategy ing (CBR), data envelopment analysis (DEA), genetic algorithms (GA),
for each BOM item through a supplier selection process. For any large and mathematical programming. All these methods can handle multiple
BOM, obtaining the most cost-effective and timely arrangement of quantitative and qualitative factors [61].
primary products typically requires extensive manual work. Both the The multi-attribute decision making methods in the construction
number of potential suppliers and the prices of materials fluctuate management context has been investigated by many scholars and
over time and require constant updating. Manually updating these experts in this field [62–64]. The construction industry needs simple
records on all corresponding BOMs can become very time-consuming. but effective decision making methods; hence, the TOPSIS approach
Time constraints related to delivery can add to or subtract from the [65] is considered and applied in this study. TOPSIS uses the multi-
value to be derived from obtaining a BOM item from a particular suppli- attribute suppliers ranking method to order the feasible alternative
er, and experience with a specific product source and the possibility of solutions (best supplier for each item of the BOMs). The MATLAB
bulk discounts could also impact project cost. An additional factor is programming language was used to apply the TOPSIS method to allow
that modern construction material management processes require a simple interaction with Microsoft Excel TM spreadsheets, because it
contractors to maintain long-term partnerships with fewer and more can extract and manipulate the data as well as maintain the indexing
68 M. Safa et al. / Automation in Construction 48 (2014) 64–73

Fig. 2. Simplified architecture of an ICMM model.

established within the spreadsheet. The first element which is involved decision makers have to choose for a specific item, and the remaining
in this process is the generic BOM. A generic BOM contains the general columns contain criteria with which alternative solutions are measured.
description and quantity of each item in a fabricated product of a The evaluation attributes are selected by taking into consideration the
work package. A purchasing strategy for those items should then be op- objectives and interests of the stakeholders.
timized based on some criteria such as availability, lead time, and cost. The TOPSIS method allows simultaneously dealing with qualitative
The main steps of the TOPSIS decision making approach, for this optimi- and quantitative (different dimensions) criteria. Hence, the attribute di-
zation using Shannon's entropy weighting method, are explained in the mensions must be transformed into non-dimensional attributes, when
following paragraphs [66]. the decision matrix is established. In other words, the raw data are nor-
TOPSIS considers that there are m alternatives (potential suppliers) malized to eliminate inconsistencies with respect to differing measure-
and n attributes/criteria and that there exists a unique score for each op- ment units and scales. The columns containing each criterion are then
tion with respect to each criterion. If xij is the score of option i with re- normalized. Each row indicates that an item originates from a specific
spect to criterion j, then we have a matrix X = (xij), which is an m × n supplier. The scores or data are normalized with the following equation:
decision matrix. The decision matrix containing all the criteria for a par-
ticular item is taken from the supplier information database. The deci- xij xij
nij ¼ vffiffiffiffiffiffiffiffiffiffiffiffiffiffi ; or nij ¼ ; j ¼ 1; …; m; i ¼ 1; …; n: ð1Þ
sion criteria are important elements of the developed process and can uX X
m
u m 2 xij
be changed from one project to another in order to accommodate the t xij ;
j¼1 j¼1
one-off nature of the construction projects and unique needs of each
project. However, criteria for selecting the best potential suppliers are
previously investigated and are employed in most construction projects The nij are normalized values of decision matrix elements, where i is
[67–69]. The selected criteria by the project management team for the the supplier and j is the criterion. The normalization is corrected, when
development of the decision matrix developed in this study were: 0 ≤ nij ≤ 1.
Accordingly, the normalized decision matrix Pij, shown in Fig. 4, pro-
• Price,
vides the relative performance of the alternatives through utilizing
• Lead time,
Eq. (1). The criteria weight vector needs to be determined in the next
• Supplier performance, and
step to present the relative ranking of the alternatives. Shannon's entro-
• Preferred supplier
py was used in this study to calculate the weight vector. Shannon's en-
The decision matrix, Fig. 3, shows how potential suppliers are evalu- tropy is one of the various methods for obtaining weights for each
ated against these criteria for a specific item of the BOM. The first col- criterion when obtaining a suitable weight based on the decision mak-
umn consists of the possible suppliers' list (alternatives) from which ing preferences is challenging. The use of objective weights based on

Fig. 3. Structure of the alternatives performance matrix used for each item.
M. Safa et al. / Automation in Construction 48 (2014) 64–73 69

Fig. 4. Normalized alternatives performance matrix.

Shannon's entropy ensures that the evaluation result is not affected by be calculated through Eq. (8), and the separation from the negative ideal
the inter-dependency of criteria and inconsistency of subjective alternative is given as Eq. (9).
weights. The original procedure of Shannon's entropy can be expressed
in a series of steps [70,71]: X   1
þ n þ 2 2
dj ¼ i¼1
vij −v i ; j ¼ 1; …; m ð8Þ
Step 1) The amount of normalized decision matrix (Fig. 3) should be
measured by the entropy value as:
m X   1
hi ¼ −K∑i¼1 pij  lnpij ð2Þ þ n − 2 2
dj ¼ i¼1
vij −v i ; j ¼ 1; …; m ð9Þ
where K = (ln m)−1 is the entropy constant and pij is the
amount of decision information (normalized decision matrix)
Then, the closeness coefficient should be defined to determine the
emitted from each criterion Ci.
ranking order of all suppliers (alternatives). The relative closeness of
Step 2) The degree of divergence di should be calculated as di = 1 − hi,
the alternative Aj with respect to A+ is described as:
i = 1,…, n. The di represents the inherent contrast intensity of
the criterion Ci.  
− þ −
Step 3) The degree of importance of criterion Ci (objective weight for R j ¼ d j = d j þ d j ; j ¼ 1; …; m ð10Þ
each criterion) can thus be calculated through:
Since d+ +
j ≥ 0, dj ≥ 0, Rj ∈ [0, 1]
di
wi ¼ ; i ¼ 1; …; n ð3Þ In this step, the preference order of all alternatives must be “ranked”
∑ni¼1 ds according to the closeness coefficient. Then, the best option (supplier) is
selected. The selected supplier for each item (alternative) must have the
Step 4) The comprehensive weight matched with the ith index factor shortest distance from the positive ideal solution, and the farthest dis-
can be obtained through experts' presumption (A =a1, …, am) tance from the negative ideal solution [72]. In other words, the ranking
and the entropy weight vector (3), as follows: order of all suppliers (according to the closeness coefficient) can be de-
a i  wi termined. Then, the best supplier for each specific item from among a
0
wi ¼ ; i ¼ 1; …; m ð4Þ set of potential suppliers can be selected.
∑m
i¼1 ai  wi

5. Supplier selection process implementation


Thus, the comprehensive weight vector is: w′ = (w1′, w2′, …, wi′).
Based on the nature of the project (entropy method) and the opinion The supplier selection process described in this paper, as part of the
of the experts who are all from industry, the weight vector calculated ICMM model, was programmed in MATLAB environment in order to
for this study is [(1) price = 23%; (2) lead time = 31%; (3) supplier per- provide a simple implementation platform that could effectively
formance = 37%; and (4) preferred supplier = 9%]. After calculating a communicate with Excel spreadsheets, the most common data source
set of weights for each criterion (wi′ for i = 1…n), each column of the on construction projects. The program, upon initialization, checks for
normalized decision matrix must be multiplied by its associated weight. an existing generic BOM database file to load. If no database file is
The weighted normalized value vij (performance ratings of the alterna- detected, it prompts the user to create a database by selecting a folder
tives and objective weights of the criteria) is calculated as: for each individual fabricated product one by one. For each folder
selected, the program processes every BOM within that folder and
0
vij ¼ wi nij ð5Þ stores the order lists for all primary fabricated products, along with
the total cost of all fabricated products, taking into account the varying
The next step is producing an overall performance index for each quantities for each separate BOM. After the database has been created
supplier (alternative). Now, the positive ideal solution (Eq. (6)) and and detected, it then creates a directory where all the BOM selection
negative ideal solution (Eq. (7)) can be identified as: files are stored.
To obtain a selection of a supplier for a particular item from the
n o n   o
þ þ þ construction package, the user selects the fabricated product(s) of a
A ¼ v1 ; …; vn ¼ maxj v ji∈I ; mini vij ji∈ J ð6Þ
construction package from the drop-down menu that lists every fabri-
cated products within the database, and initiates the selection process
n   o which then requires the user to choose the quantity of the specific
− − −
A ¼ fv1 ; …; vn g ¼ minj v ji∈I ; maxi vij ji∈J ð7Þ fabricated product desired. The program detects the best selection of
suppliers for the items of a BOM with the minimum cost and lead
where I is associated with supplier performance, and preferred supplier time from the database and outputs the order of primary fabricated
criteria; and J is associated with price and lead time criteria. At this step, products with the quantities adjusted based on the desired amount of
the separation of each alternative (supplier) from the ideal solution can the fabricated product input by the user.
70 M. Safa et al. / Automation in Construction 48 (2014) 64–73

The developed supplier selection process is intended for use with a prefabricated modules on hand, whether on other sites, suppliers' stores
large BOM database so that any specific company can generate an or at the shipping process. Internet connectivity and cloud computing
order quickly and efficiently, without the requirement for manual have provided the infrastructure for virtual inventories. Virtual invento-
browsing of BOMs and a lengthy selection process. With the ICMM ry lists do not include the materials and equipment currently at the
model, most of the detailed record keeping and updates are automated, construction site. Virtual inventory provides the ability to achieve a
and project managers are notified about purchasing requirements. The faster purchasing process, to avoid the surplus (left over in other
code related to the MATLAB functions that were programmed for the projects), and to achieve flexibility, and higher valuations than the
selection process has been previously published by the author [73]. more traditional “stock and reactive purchasing process”.
The BOM selection process requires the presence and processing of The warehouse component of the ICMM model, which is specified to
the components necessary for the selection of a best possible BOM so be integrated with the other elements of the model, is intended to
that a database can be created. The first element is an updated supplier record whether the item is a new or replacement material, a safety- or
information sheet, which should include the cost of a particular (or min- augmented-quality material, measuring and test equipment (M&TE), a
imum) quantity of an item, the lead time, and a company rating, or rotatable spare, or a lead-time item and whether it is covered by the
experience factor for that specific supplier. The full fabricated product maintenance rule. The ICMM model requires a robust warehouse
data are then extracted and stored in a variable in the same structure system for the control and management of all activities related to the re-
where the pre-arranged Excel sheet is located. The next step incorpo- ceipt, inspection, storage, counting, distribution, issuance, and shipping
rates all generic BOM structures, which serve as a “skeleton” for of equipment and materials. Following this comparison, the draft of the
constructing a selected BOM. All BOM structures required for a particu- purchase order is ready, and the model is run in order to obtain the
lar project are extracted in a system-specified format from each optimal solution, which is the final purchase order. The running of the
construction work package. program and support for this part of the model requires accurate infor-
For the ICMM model, every generic BOM is placed in its own unique mation about suppliers, which must then be entered into the model. The
Excel sheet. A generic BOM file of each fabricated product contains each study with 20 construction packages is described next.
item name, item code, units, hierarchy, and quantity. The supplier selec- Suppliers S1, S2, …, S8 have almost the same characteristics, and their
tion process filters all items of a BOM for the completion of a requisition location is the same locale as the project site and therefore would entail
order for a particular fabricated product. This process is run for all BOMs no delivery fees. These suppliers also have knowledge of the policies
imported into the model, and a database of fabricated product requisi- and delivery procedures used by the Canadian group of construction
tion order templates, complete with item quantities for each unit of companies and are on the group's Approved Suppliers List (ASL)
that specific BOM, is generated and stored. because of years of proven reliability in supplying several projects.
The last step involves the extraction of the user-specified Excel sheet Other characteristics that were considered with respect to each individ-
containing the weighting factor for each criterion in the updated suppli- ual supplier included quick order turnaround time, consistently correct
er information sheet. Once the processing algorithm is initialized, the documentation accompanying shipments, immediate correction of
supplier selection process checks for a materials list database. If none order problems by the inside sales contact, immediate correction of in-
is detected, the user is prompted to load every generic BOM structure voicing issues and prompt issuing of credits, early payment discounts
sequentially for every fabricated product, following which a fabricated for payments made within the specified terms, and status as a preferred
product database is created. The user next selects the materials manage- supplier of the construction group companies involved, which entails
ment spreadsheet, which the program reads and stores within a new valuable year-end sales rebates.
materials list database. Finally, the user is prompted for the weighting The database information about suppliers S9, S10 and S11 showed
factor spreadsheet, which is then stored. that none of these suppliers are local, so additional delivery fees
would be charged. Suppliers S9, S10 and S11 are also not on the ASL for
6. Validation and data preparation the group of companies, which means that these suppliers would be
required to complete an evaluation form and submit it to quality control
The following case study provides an understanding of the way the for review and possible audit. The quality control department would
process can be developed and implemented in an organization for a also have to evaluate and then approve or deny all materials, a process
specific construction project. Since the names of suppliers and all details that can be quick or lengthy, depending on the supplier's certifications,
associated with the contracts must be kept confidential and may not be qualifications, and internal quality programs.
made public, capital letters are used as anonymous designations for After the supplier selection process was completed, using all of the
suppliers. In this case, the project included 680 work packages, all of processes explained previously, the potential awarded suppliers against
which had to be procured using a BOM per work package. For the each item can be determined. Table 1 shows the sample results of the
purposes of this demonstration study, only the fabrication of a complex procurement exercise. It depicts a portion of one of the BOM structures
piping module consisting of 20 of the construction packages is for this project, showing the corresponding work package, item name,
addressed. Its fabrication is discussed with respect to the implementa- item code, awarded supplier for each item, hierarchy, and quantity for
tion of the ICMM model as a means of illustrating the production of an all fabricated products. The cost column in this figure represents only
automated final order. For these 20 construction packages used in this the best cost option after the selection process was run. It should be
validation, 14 different potential suppliers of the required materials noted that even after selecting best items and initiating the purchase
were available. After the pre-evaluation process, three suppliers were order, the purchasing department would still need to negotiate and fi-
removed by the project management team. It should be noted that if a nalize the contract details with the successful supplier.
supplier provides two or more prices with different lead times or condi- The implementation of the selection process was discussed with
tions, the model recognize this suppliers as the two or more suppliers experts, throughout this research. The results were then compared
with distinctive IDs. with those obtained through the purchasing department. The results
As mentioned previously, when a request is issued by an authorized showed that the process is accurate and can be performed project-
person (e.g., the project manager, superintendent, or project engineer), wide. Our study showed that the most challenging part of this process
the first draft is prepared through appropriate amounts listed in the is assigning the proper weight to each criterion, based on experts'
generic BOM. This draft order must then be compared with the ware- opinion. This process is critical as the results of the selection process
house inventory stock. This process is shown in Fig. 2. The warehouse may change by small changes in weighting vector (Table 2). In this re-
inventory can be categorized into virtual and actual inventories. A search, approximately 350 man-hours was spent to extract all necessar-
virtual inventory includes all of the materials, equipment, and ily information from PDF and Excel files provided by contractor, only for
M. Safa et al. / Automation in Construction 48 (2014) 64–73 71

Table 1
Sample results of the supplier selection process for a work package.

Work package # Item # Item Code Awarded supplier(S) Quantity Cost ($)

3 1 FT 12″ XH 1333 GR6 SMLS DR 1154788 S10 170 21,590


3 2 FT 8″ XH A333 GR6 SMLS 1158486 S10 20 900
3 3 FT. 15″ XH A106B SMLS DR 1166788 S4 80 392
3 4 FLANGE 12″ PN100 RFWN XH A350LF2 CL1 1165879 S7 10 5650
3 5 ELBOW 90 LR 12″ XH A420WPL6 1768688 S7 13 9360
3 6 ELBOW 90 1.5″ 3M THREADED A105 1768329 S7 11 136
3 7 CONIC REDUCER 12″ × 8″ XH × STD A420WPL6 1662187 S7 3 2250
3 8 SWAGE CONC THRD 1.5″ × 0.5″ XH A234WPB 1157777 S7 6 102
3 9 UNION THRD 1.5″ 3M A105 1323255 S10 8 112

3 117 BULL PLUG HEX SOLID BODY 1.5″ 6M A234WPB 1766978 S10 8 98
3 118 TEE STRAIGHT THRD 1687878 S7 10 158

a small portion of the project (20 out of 680 work packages). Even relationship between managers (decision makers) and suppliers can
though this seems to be a very significant effort, it should be noted become too personal over the course of years. While this kind of person-
that once an ICMM framework is fully implemented as described in al selection practice has associated benefits, such as stability, mutual
this paper, all required information is captured and automatically trust, and reduced transaction and search costs, it also entails problems.
utilized through the system. Therefore, the supplier selection process Trust between an owner and a supplier which is created through
for the entire project can be performed in a matter of minutes, hence credibility and benevolence is such an important factor [75] that should
utilizing the true advantages of automation in construction. Most of be always considered for the development of the ICMM. However, a
the information captured in an ICMM model is transferrable from critical recommendation is the avoidance of any unsystematic tenden-
project to project, with minor modification, thus increasing the return cies and reliance, instead, on a systematic method for designating a
on investment in an ICMM system in the long-term. supplier.
For this case study, partial sensitivity analysis was performed for the
weight-vector through changing parameter values and rerunning the 7. Conclusion
selection process. In other words, managers changed the weight of
criteria and examined the impact that the change had on the model's The major contribution of this paper lies in the development of the
results. In practice, the factors affecting the purchasing decision process ICMM model for the efficient procurement of construction materials,
are changing from project to project. They may also change over the primarily through the use of the TOPSIS (Technique for Order
course of a project. Thus, project managers can control the process Preference by Similarity to Ideal Solution) method as an effective frame-
through alterations in the weight of attributes. This approach could pro- work for the supplier selection process. The developed model is based
vide valuable insights into the robustness of purchase-decisions. Differ- on consideration of the majority of the activities associated with the
ent scenarios were considered for this study (Table 2). For example, procurement of materials: (1) processing and administration related
Scenario 6 occurs when project execution is deemed to be far behind to purchase requisitions, (2) purchase orders, (3) internal supply
the project schedule and therefore the weight of lead-time is increased. requests, (4) shipping, (5) expediting of materials, and (6) supplier
Scenario 5 could be considered when the cost needs to be kept down. selection. The general-purpose framework for the selection of the
Table 2 shows the different scenarios' results. Any changes in the most suitable supplier is formulated as a multi-criteria decision-
weights of criteria may change the results; however, the outcomes of making problem that incorporates a variety of material management
some scenarios are identical for this specific study. parameters. The TOPSIS method, an effective multi-criteria decision
It should be noted that some of procurement experts prefer to analysis method, was therefore employed as the framework for ranking
eliminate the weighting criteria to avoid the challenges such as conflict the suppliers for each BOM item. A MATLAB programming language
opinions, and lack of time and precise information due to the project environment was used for developing and implementing the supplier
complexity. A few methods, such as the FPSI method, have been devel- selection process developed in this research. The new process was
oped to eliminate the weighting criteria through assigning relative im- applied for an industrial construction project and was functionally
portance of each (based on decision maker's opinions) [74]. demonstrated to be effective for facilitating the decision-making pro-
By implementing the ICMM model, the procurement process can cess by providing optimal supplier rankings for each BOM item based
also be integrated and automated. Furthermore, the user can mitigate on consideration of a number of constraints and interdependencies. It
the significant risk of excessive partnership strength in which the can thus be concluded that materials supplier selection for construction

Table 2
The consequence table in sensitivity analysis for the total of items purchased (20 selected work package).

Price Lead time Performance Early payment Awarded suppliers

Base case 0.2 0.3 0.4 0.1 S1 & S2 & S4 & S5 & S6 & S7 & S9 & S10 & S11
Scenario 1 0.25 0.25 0.4 0.1 S1 & S2 & S3& S4 & S5 & S6 & S7 & S9 & S10 & S11
Scenario 2 0.3 0.2 0.4 0.1 S1 & S2 & S3& S4 & S5 & S6 & S7 & S9 & S10 & S11
Scenario 3 0.4 0.4 0.1 0.1 S1 & S2 & S3& S4 & S5 & S9 & S11
Scenario 4 0.3 0.3 0.3 0.1 S3 & S4 & S5 & S6 & S7 & S9 & S11
Scenario 5 0.5 0.2 0.2 0.1 S1 & S2 & S4 & S7 & S8 & S9 & S11
Scenario 6 0.2 0.5 0.2 0.1 S1 & S2 & S4 & S5 & S6 & S7 & S8 & S11
72 M. Safa et al. / Automation in Construction 48 (2014) 64–73

enterprises can be optimized using the TOPSIS method and the frame- [31] R. Sacks, M. Radosavljevic, R. Barak, Requirements for building information
modeling based lean production management systems for construction, Autom.
work presented here. Further development and enhancements could Constr. (ISSN: 0926-5805) 19 (5) (2010) 641–655.
include connecting the ICMM model to the materials management sys- [32] M.C. Forde, High performance construction materials and systems: An essential
tems of trusted suppliers and implementing additional automation and program for America and its infrastructure technical report 93–5011, American
Society of Civil Engineers, New York, 1993. (ISBN 0-87262-938-2, 212 pp.).
integration within their models. Future research will provide further [33] M.R. Manavazhia, D.K. Adhikarib, Material and equipment procurement delays in
comparisons of the proposed and current selection processes and also highway projects in Nepal, Int. J. Proj. Manag. 20 (2002) 627–632.
investigate alternative optimization approaches. [34] G. Sweis, R. Sweis, Hammad A. Abu, A. Shboul, Delays in construction projects: the
case of Jordan, Int. J. Proj. Manag. 26 (6) (2008) 665–674.
[35] H. Abdul-Rahman, M.N. Alidrisyi, A perspective of materials management practices
in a fast developing economy: the case of Malaysia, Constr. Manag. Econ. 12 (1994)
References 413–422.
[36] P.N. Christensen, G.A. Sparks, K.J. Kostuk, A method-based survey of life cycle
[1] Business Roundtable, Modern management systems, Construction Industry Cost costing literature pertinent to infrastructure design and renewal, Can. J. Civ.
Effectiveness Report A-61982. Eng. 32 (1) (2005) 250–259.
[2] G. Stukhart, L.C. Bell, Attributes of materials management systems, phase 1 [37] E. Asplund, U. Danielson, Rata ut Byggsvangen (Straightening the Building
research-material management performance, Report prepared for the Construction Roundabout), SBUF, Stockholm, 1991.
Industry Institute (CII), Source Document 1, USA1987. [38] J. Jarnbring, Construction Site's Material Flow Costs, Lund University of Technology,
[3] Construction Industry Institute (CII), Project Materials Management Handbook, Lund, 1994. (in Swedish).
University of Texas at Austin, Materials Management Task Force, CII, 1988. [39] G. Polat, D. Arditi, The JIT materials management system in developing countries,
[4] N.T. Ibn-Homaid, A comparative evaluation of construction and manufacturing Constr. Manag. Econ. 23 (7) (2005) 697–712.
materials management, Int. J. Proj. Manag. 20 (4) (2002) 263–270. [40] J.D. Borcherding, S.J. Sebastian, Major factors influencing craft productivity in nucle-
[5] C.H. Caldas, D.G. Torrent, C.T. Haas, Integration of automated data collection technol- ar power plant construction, Trans. Am. Assoc. Cost Eng. 1 (1980) (Paper I).
ogies for real-time field materials management, Proceedings of the 21st International [41] S.A. Assaf, S. Al-Hejji, Causes of delay in large construction projects, Int. J. Proj.
Symposium on Automation and Robotics in Construction, Jeju, Korea, 2004. Manag. 24 (4) (2006) 349–357.
[6] A.E. Kerridge, Manage materials effectively, Hydrocarbon Processing, Part 1 and 2, [42] L.L. Ekanayake, G. Ofori, Building waste assessment score: design-based tool, J. Build.
661987. (63-71 and 67-70). Environ. (2004) 851–861.
[7] H.R. Thomas, D.R. Riley, John I. Messner, Fundamental principles of site material [43] B.K. Fishbein, Building for the Future: Strategies to Reduce Construction and Demolition
management, J. Constr. Eng. Manag. 131 (7) (2005) 808–815. Waste in Municipal Projects, http://www.informinc.org/cdreport.html1998.
[8] Construction Industry Institute (CII), Procurement and Materials Management: A [44] J. Guoli, G. Daxin, F. Tsui, Analysis and implementation of the BOM of a tree-type
Guide to Effective Project Execution, University of Texas at Austin, Texas, 1999. structure in MRPII, J. Mater. Process. Technol. (ISSN: 0924-0136) 139 (2003)
[9] M. Safa, C.T. Haas, J. Gray, K. Hipel, Electronic Process Management System based 535–538 (20).
Front End Planning Tool (FEPT), J. Constr. Eng. Proj. Manag. (ISSN: 2233-9582) 3 [45] Y. Chung, G.W. Fischer, A conceptual structure and issues for an object-oriented bill
(2) (2013). of materials (BOM) data model, Comput. Ind. Eng. 26 (2) (1994) 321–339.
[10] I. Chapman, P. Olomolaiye, F. Harris, Automation problems in materials manage- [46] H.M.H. Hegge, J.C. Wortmann, Generic bill-of-material: a new product model, Int. J.
ment on large construction projects, Proceedings of the 7th ISARC, Bristol, United Prod. Econ. 23 (1–3) (1991) 117–128.
Kingdom, 1990, pp. 499–504. [47] J. Jiao, M.M. Tseng, Q. Ma, Y. Zou, Generic bill-of-materials-and-operations for high-
[11] R. Vrijhoef, L. Koskela, The four roles of supply chain management in construction, variety production management, Concurr. Eng. Res. Appl. 8 (4) (2000) 297–321.
Eur. J. Purch. Supply Manag. 6 (2000) 169–178. [48] Y. Zhang, Y. Yang, M. Jiang, P. Xia, B. Li, Design and implementation of auto parts
[12] D.M. Gann, A.J. Salter, Innovation in project-based, service enhanced Ž rms: the data management system based on BOM, Appl. Mech. Mater. 127 (2012) 42–47.
construction of complex products and systems, Res. Policy 29 (2000) 955–972. [49] J.H. Lee, S.H. Kim, K. Lee, Integration of evolutional BOMs for design of ship outfitting
[13] C.T. Haas, W.R. Fagerlund, Preliminary research on prefabrication, pre-assembly, equipment, CAD Comput. Aided Des. 44 (3) (2012) 253–273.
modularization and off-site fabrication in construction, Construction Industry [50] Construction Owners Association of Alberta (COAA), Workface Planning Model and
Institute (CII), University of Texas at Austin, Austin, Texas, 2002. Implementation Guide, 2007. (Edmonton. Alberta, Canada).
[14] S.H. Han, M. Al-Hussein, S. Al-Jibouri, H. Yu, Automated post-simulation visualiza- [51] Construction Industry Institute (CII), Work Packaging for Project Control, CII Infor-
tion of modular building production assembly line, Autom. Constr. 21 (2012) mation Publication, Univ. of Texas at Austin, Texas, 1988.
229–236. [52] P. Bhutia, R. Phipon, Application of AHP and TOPSIS method for supplier selection
[15] N. Yeung, A. Chan, D. Chan, Application of prefabrication in construction — A new problem, IOSR J. Eng. 2 (2012) 43–50.
research agenda for reform by CII-HK, Conference on Precast Concrete Building [53] A.A. Costa, L.V. Tavares, Advanced multicriteria models to promote quality and rep-
Systems, Hong Kong, 2003. utation in public construction e-marketplaces, Autom. Constr. 30 (2013) 205–215.
[16] R. Sacks, C.M. Eastman, G. Lee, Process model perspectives on management and en- [54] P. Jaskowski, S. Biruk, R. Bucon, Assessing contractor selection criteria weights with
gineering procedures in the precast/pre-stressed concrete industry, J. Constr. Eng. fuzzy AHP method application in group decision environment, Autom. Constr. 19
Manag. 130 (2) (2004) 206–215. (2) (2010) 120–126.
[17] C.T. Haas, J.T. O'Connor, R.T. Tucker, J.A. Eickmann, W.R. Fagerlund, Prefabrication [55] A. Nieto-Morote, F. Ruz-Vila, A fuzzy multi-criteria decision-making model for con-
and preassembly trends and effects on the construction workforce, Report No. 14, struction contractor prequalification, Autom. Constr. 25 (2012) 8–19.
Center for Construction Industry Studies, Austin, TX, 2000. [56] P.O. Akadiri, P.O. Olomolaiye, E.A. Chinyio, Multi-criteria evaluation model for the
[18] H. Voordijk, B. Meijboom, J.D. Haan, Modularity in supply chains: a multiple selection of sustainable materials for building projects, Autom. Constr. 30 (2013)
case study in the construction industry, Int. J. Oper. Prod. Manag. 26 (6) (2006) 113–125.
600–618. [57] G. Arslan, S. Kivrak, M. Birgonul, I. Dikmen, Improving sub-contractor selection pro-
[19] V. Tam, C. Tam, W. Ng, On prefabrication implementation for different project types cess in construction projects: Web-based sub-contractor evaluation system
and procurement methods in Hong Kong, J. Eng. Des. Technol. 5 (1) (2007) 68–80. (WEBSES), Autom. Constr. 17 (2008) 480–488.
[20] C. Hendrickson, A. Tung, Project Management for Construction, Fundamental [58] Cristóbal J. San, Contractor selection using multicriteria decision-making methods, J.
Concepts for Owners, Engineers, Architects and Builders, Prentice Hall, 1998, ISBN Constr. Eng. Manag. 138 (2011) 751–758.
0-13-731266-0. [59] M. Sonmez, G. Holt, J.B. Yang, G. Graham, Applying evidential reasoning to
[21] A. Segerstedt, T. Olofsson, Supply chains in the construction industry, Supply Chain prequalifying construction contractors, ASCE J. Manag. Eng. 18 (3) (2002) 111–119.
Manag. Int. J. 15 (5) (2010) 347–353. [60] D. Singh, R. Tiong, A fuzzy decision framework for contractor selection, ASCE J.
[22] R. Vrijhoef, L. Koskela, The four roles of supply chain management in construction, Constr. Eng. Manag. 131 (1) (2005) 62–70.
Eur. J. Purch. Supply Manag. 6 (3-4) (2000) 169–178. [61] W. Ho, X. Xu, P.K. Dey, Multi-criteria decision making approaches for supplier
[23] A. Dubois, L.E. Gadde, Supply strategy and network effects — purchasing behavior in evaluation and selection: a literature review, Eur. J. Oper. Res. 202 (2010) 16–24.
the construction industry, Eur. J. Purch. Supply Manag. 6 (3-4) (2000) 207–215. [62] E.K. Zavadskas, T. Vilutiene, Z. Turskis, J. Tamosaitiene, Contractor selection for
[24] W. Jang, M. Skibniewski, Cost–benefit analysis of embedded sensor system for con- construction works by applying saw‐g and topsis grey techniques, J. Bus. Econ.
struction materials tracking, J. Constr. Eng. Manag. 135 (5) (2009) 378–386. Manag. 11 (2010) 34–55.
[25] H.M. Elzarka, L.C. Bell, Object-oriented methodology for materials-management sys- [63] Y.H. Lin, P.C. Lee, T.P. Chang, H.I. Ting, Multi-attribute group decision making model
tems, J. Constr. Eng. Manag. 121 (4) (1995) 438–445. under the condition of uncertain information, Autom. Constr. (ISSN: 0926-5805) 17
[26] H.R. Thomas, V. Sanvido, Impact of material management on productivity—a case (6) (2008) 792–797.
study, J. Constr. Eng. Manag. ASCE l (i 15) (1989) 370–384 (1(i 15). [64] X. Xue, X. Li, Q. Shen, Y. Wang, An agent-based framework for supply chain coordi-
[27] J.C.P. Cheng, K.H. Law, H. Bjornsson, A. Jones, R. Sriram, A service oriented frame- nation in construction, Autom. Constr. (ISSN: 0926-5805) 14 (3) (2005) 413–430.
work for construction supply chain integration, Autom. Constr. (ISSN: 0926-5805) [65] C.L. Hwang, K. Yoon, Multiple Attribute Decision making: Methods and Applications,
19 (2) (2010) 245–260. Springer, New York, 1981.
[28] A. Segerstedt, T. Olofsson, Supply chains in the construction industry, Supply Chain [66] S.J. Chen, C.L. Hwang, Fuzzy Multiple Attribute Decision Making: Methods and
Manag. Int. J. 15 (5) (2010) 347–353. Applications, Springer-Verlag, Berlin, 1992.
[29] H.P. Tserng, P.H. Lin, An accelerated subcontracting and procuring model for [67] I.M. Mahdi, M.J. Riley, S.M. Fereig, A.P. Alex, A multicriteria approach to contractor
construction projects, Autom. Constr. (ISSN: 0926-5805) 11 (1) (2002) 105–125. selection, Eng. Constr. Archit. Manag. 9 (1) (2002) 29–37.
[30] S.T. Ng, C.D.T. Luu, Modeling subcontractor registration decisions through case- [68] J.T. O'Connor, G.E. Gibson, G.C. Migliaccio, R1–2004 Annual Interim Report, Report
based reasoning approach, Autom. Constr. 17 (7) (2008) 873–881 ISSN 0926–5805. 0–4661–1 to Texas Department of Transportation, CTR, Austin, TX, 2005. 32.
M. Safa et al. / Automation in Construction 48 (2014) 64–73 73

[69] D. Singh, R. Tiong, Contractor selection criteria: investigation of opinions of [73] M. Safa, A Comprehensive Advanced Construction Supply Nexus Model, University
Singapore construction practitioners, J. Constr. Eng. Manag. 132 (9) (2006) of Waterloo, Canada, 2013. 181–205 (PhD thesis).
998–1008. [74] H. Abbasianjahromi, H. Rajaie, E. Shakeri, A framework for subcontractor selection in
[70] C.E. Shannon, A mathematical theory of communication, Bell Syst. Tech. J. 27 (1948) the construction industry, J. Civ. Eng. Manag. 19 (2013) 158–168.
379–423. [75] I. Rodríguez del Bosque, J. Collado, H. San Martín, Determinants of economic and
[71] H. Deng, C. Yeh, R.J. Willis, Inter-company comparison using modified TOPSIS with social satisfaction in manufacturer–distributor relationships, Ind. Mark. Manag.
objective weights, Comput. Oper. Res. 27 (2000) 963–973. (ISSN: 0019-8501) 35 (6) (2006) 666–675.
[72] G.R. Jahanshahloo, F. Hosseinzadeh Lotfi, M. Izadikhah, An algorithmic method to
extend TOPSIS for decision-making problems with interval data, Appl. Math.
Comput. 175 (2006) 1375–1384.

You might also like