Knowledge-based Decision Support System Quality Function Deployment (KBDSS-QFD) tool for assessment of building envelopes

Automation in Construction 35 (2013) 314–328

Contents lists available at ScienceDirect

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

Knowledge-based Decision Support System Quality Function Deployment (KBDSS-QFD) tool for assessment of building envelopes Natee Singhaputtangkul ⁎, Sui Pheng Low, Ai Lin Teo, Bon-Gang Hwang Department of Building, School of Design and Environment, National University of Singapore, 4 Architecture Drive, 117566 Singapore, Singapore

a r t i c l e

i n f o

Article history: Accepted 18 May 2013 Available online 22 June 2013 Keywords: Decision support system Knowledge-based system Quality Function Deployment Building envelopes Design team Design management

a b s t r a c t A building design team has faced several decision-making problems when assessing building envelope materials and designs for a private high-rise residential building in the early design stage. This study developed an automated fuzzy Knowledge-based Decision Support System Quality Function Deployment (KBDSS-QFD) tool to facilitate the team to mitigate such problems. A case study of the design team comprising an architect, a civil and structural (C&S) engineer and a mechanical and electrical (M&E) engineer was selected as the research design of this study. Results from the qualitative data analysis showed that the tool has the potential to mitigate the decision-making problems. The contributions of using this automated tool include not only achieving better design management but also raising the level of productivity in the construction industry. © 2013 Elsevier B.V. All rights reserved.

1. Introduction Building envelopes, as the interface between interior space and exterior environment, serve the function of weather and pollution exclusion and thermal and sound insulation [1]. Their performance affects occupant comfort and productivity, energy use and running costs, stability, durability, esthetics appeal of a building, etc. [2,3]. A thoughtful building envelope assessment can make a building work more effectively for its builders and occupants as part of stakeholders of a project and indoor environment [4]. Thus, success of the project is tied with the assessment and selection of building envelope materials and designs that can satisfy requirements of the stakeholders [5]. However, assessment of the building envelope materials and designs for high-rise residential buildings in the early design stage is not a simple task. It requires large amount of information and inputs from a design team comprising several building professionals particularly architects and engineers [2,6,7]. As a result, the assessment appears to be simultaneously affected by a number of decision-making problems; for instance, inadequate consideration of requirements, lack of communication and integration between members of a team, subjective and uncertain requirements, etc. [8,9]. These problems can cause significant adverse impacts to a project such as delays, increase in expenses, increase in manpower of a building project, poor professional relationship and poor client satisfaction [8,10].

⁎ Corresponding author. Tel.: +65 9398 6772. E-mail addresses: [email protected] (N. Singhaputtangkul), [email protected] (S.P. Low), [email protected] (A.L. Teo), [email protected] (B.-G. Hwang). 0926-5805/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.autcon.2013.05.017

Hence, there is a need for the design team to mitigate the decision-making problems. It was found that the use of the Quality Function Deployment (QFD) approach, fuzzy set theory and knowledge management system (KMS) shows a potential not only to facilitate decision-making processes of a design team, but also to improve quality of design solutions [7,11]. As such, in an effort to deal with the decision making problems, the main objective of this study is to develop a decision support system (DSS) by applying the QFD approach integrated with the fuzzy set theory and KMS to facilitate the design team to mitigate the decision-making problems in the early design stage. This system is named a Knowledge-Based Decision Support System QFD (KBDSS-QFD) tool. 2. Concepts of building envelope design Building envelope design alternatives can generally be grouped into four major hypothetical types based on basic external wall materials including precast, masonry, fixed-glass and curtain walls as shown in Fig. 1 [12]. As can be seen, the systems typically comprise three fundamental material categories; namely external walls, windows and shading devices. The assessment of these building envelope materials and combinations of their designs for high-rise residential buildings involves a number of decision makers (DMs) as part of the design team from different backgrounds and requires intensive considerations related to design and construction of building envelope systems. These considerations are associated with esthetics, labor's skill sets, availability of manpower and equipment, building performance, durability, costs of a building project and so on [13,14]. In Singapore, architects from an architectural firm lead the entire building design development including building envelope design

N. Singhaputtangkul et al. / Automation in Construction 35 (2013) 314–328

315

Fig. 1. High-rise residential buildings in Singapore.

development. In the early design stage of a private high-rise residential building project, the architects receive relevant information regarding the building envelope design development of the project from the developer/owner. The architects then develop a conceptual building envelope design with the help of the civil and structural (C&S) engineers, and mechanical and electrical (M&E) engineers from engineering consultancy firms to satisfy the requirements of the developer by providing a set of design alternatives. After that, the developer selects and finalizes the conceptual design, and this allows the architects and engineers to move on to develop schematic and detailed building envelope designs. In this regard, the design team in-charge of the assessment of the building envelope materials and designs in the early design stage consists of three main DMs; namely the architect, C&S engineer and M&E engineer. 3. Decision-making problems in the early stage design From a pilot study conducted in Singapore in February 2012 and extensive literature reviews, six major decision-making problems faced by the architects and engineers as a team when assessing the building envelope materials and designs were found to include the following. 3.1. Inadequate consideration of requirements Inadequate consideration of requirements in the early design stage is a major cause of poor performances of construction projects [15]. For instance, because of inadequate consideration of requirements, designers may not be able to develop a comprehensive design, which may lead to numerous adverse impacts during different project phases [16]. Singhaputtangkul et al. [5] suggested that inadequate consideration of building envelope requirements by designers tends to lead to redesigning activities, particularly when new assessment criteria have to be additionally considered. These activities can cause progress delay, project delay, increase in expenses and increase in manpower needed of a building project [10]. 3.2. Inadequate consideration of possible materials and designs The field of building envelope design and engineering is quite established, while new materials and systems are being developed on a continual basis. El-Alfy [16] and Makenya and Soronis [17] pointed out that architects and engineers typically select materials drawn from their personal collection of literature and their knowledge of what is available in the market, and frequently use short cuts based on their experience in order to save time. In addition, architects and engineers prefer to stick to familiar products, have a strong preference for certain materials and components used previously, and typically refuse to use new products unless these are unavoidable. This consequently seems to reduce a number of possible building envelope materials and design alternatives that could satisfy requirements of the stakeholders.

3.3. Lack of efficiency and consistency Lack of efficiency and consistency is another major problem in making decisions particularly in the early design stage of the design team. This problem leads to delay, lack of confidence and participation among members of the team, and eventually affect a client's satisfaction [18,19]. There are numerous sources of this problem. One of these is an absence of an organized KMS. This issue is significant because, for instance, in cases where professionals leave the organization or the design team while the project experience continues to reside within the individual professionals, in the absence of an established and organized KMS, the professional team would face problems in designing and planning [15,20]. 3.4. Lack of communication and integration between members of the team In building design, communication and integration play a vital role in combining ideas of designers from different parties together during design processes. However, communication and integration among designers are often fraught with difficulties and are seldom linked to design outcomes. In fact, lack of communication and integration has been recognized as a crucial problem not only during the design development stage but also during the entire project development cycle. Poor communication and integration render the achievement of an optimal design difficult as well as a time-consuming process [9,21]. This problem tends to lead to unclear instructions, additional works, progress delay, project delay and poor quality of design solutions [22,23]. 3.5. Subjective and uncertain requirements Practical building design depends heavily on intuitive thinking and professional expertise that usually have a large variation of shades of gray as opposed to black and white colors [24]. Significantly, Brock [13], Lu et al. [25] and Pedrycz et al. [26] suggested that designers have faced problems in interpreting this type of requirements. In particular, under vague and uncertain circumstances especially in the early design stage, designers seem to be unable to estimate their preferences with an exact numerical input. This makes finding a balance between the subjective criteria one of the major decisionmaking problems encountered by most architects and engineers when assessing the building envelope materials and designs. 3.6. Disagreement among members of the team Nutt [27] defined “decision making” as a process made up of stages carried out to set directions, identify solutions, evaluate courses of action and implement a preferred plan. The effectiveness of the group decision process has become an increasingly important organizational concern [28]. A common organizational response to this concern is to design cross-functional teams [29]. Nevertheless, these heterogeneous groups exhibit additional problems. As the

316

N. Singhaputtangkul et al. / Automation in Construction 35 (2013) 314–328

assessment of the building envelope materials and designs involves many complex and conflicting aspects intrinsic to human individuality and human nature, a design team inevitably encounters different levels of disagreement between opinions of architects and engineers. It is noted that unsettled disagreement can possibly cause disputes within a party, disputes among parties, poor professional relations and ambiguous design details [30,31].

4. Applying QFD approach to mitigate the decision-making problems identified In making decisions of organizations in any industry, one of the most privileged DMs is the customers. Satisfying their needs and expectations appears to be of utmost importance for the organizations. Many companies have adopted approaches to improve quality of their products to satisfy their customers. Among these approaches, QFD is regarded as a highly effective and structured planning tool to systematically deal with customer demands and to precisely define their requirements [32,33]. In the building industry, as a building project is relatively unique in the sense that each building is tailor-made to meet the requirements and needs of the customers which refer to all stakeholders of a project, using a QFD approach seems to make good sense [7,32,34]. By linking the customers' requirements to engineering characteristics, QFD controls quality of the project through its House of Quality (HOQ) which contains rooms to achieve targets of the project as shown in Fig. 2. The first room refers to a list of customer requirements. It contains customer needs or expectations from a particular task. The second room presents a list of engineering characteristics that can affect one or more of the customer requirements. The third room keeps an interrelationship matrix between pairs of engineering characteristics, while the fourth room records a relationship matrix between the customer requirements and engineering characteristics. The fifth and sixth rooms contain calculation algorithms for prioritizing the customer requirements and engineering characteristics, respectively, coupled with the realization that the computing algorithms should reflect the context of both the customer requirements and engineering characteristics [33,34]. A conventional QFD tool promotes identifying the requirements of the stakeholders and design alternatives, reducing disagreement between members of a design team, and making decisions as a team to a certain level. Nevertheless, the conventional QFD tool has some drawbacks. These include the amount of time to implement the tool [35], lack of knowledge-base decision making [36], lack of the techniques to deal with qualitative and subjective requirements [32], and the difficulty in manually recording the QFD matrix in a paper form [35] and in dealing with complex product and conflicting

requirements [36,37]. Importantly, to a great extent, the drawbacks seem to limit the capability of the conventional QFD tool to mitigate the decision-making problems identified, so much so that there is a need to improve this conventional tool. In response to this, from the literature reviews, the study developed the KBDSS-QFD tool by integrating a modified QFD tool with a KMS to store knowledge of key criteria and building envelope materials and design alternatives that aims to mitigate the problems related to inadequate consideration of requirements and inadequate consideration of possible materials and designs, respectively [7]. In parallel, using such a KMS is expected to reduce the problem related to lack of efficiency and consistency in making decisions of the design team [15,20]. Mean time, the tool incorporates fuzzy set theory and automated user interface to mitigate the problems related to subjective and uncertain requirements and lack of communication and integration between members of a design team, respectively [38,39]. Apart from this, a fuzzy consensus scheme is also embedded in the tool to mitigate disagreement between opinions of members of the design team [26]. With these concepts in mind, it is hypothesized that the KBDSS-QFD tool proposed can be applied to facilitate the design team to mitigate the decision making problems identified as a whole.

5. Research methodology Fig. 3 illustrates the research methodology of this study. The study conducted a series of semi-structured interviews with 15 architects and engineers in parallel with thorough literature reviews to build the automated KBDSS-QFD tool and to acquire the knowledge for the KMS database. The tool and its database were developed using Microsoft Visual Studio and Microsoft Access for Windows, respectively. Case study was selected as the research design, and interview was adopted as the method of data collection. The study engaged the design team consisting of a senior architect, C&S engineer and M&E engineer as the case study to test the KBDSS-QFD developed. Subsequently, a qualitative data analysis approach was used to assess the perspectives of these three DMs with respect to the potential of applying the tool to mitigate the decision-making problems identified.

6. Development of the KBDSS-QFD tool Fig. 4 presents the architecture of the KBDSS-QFD tool. It comprises three major elements which are the HOQ, KMS and fuzzy inference engine. These three elements are managed through a user interface of the tool developed in consideration of feedbacks obtained from the semi-structured interviews.

Fig. 2. A conventional House of Quality (HOQ).

N. Singhaputtangkul et al. / Automation in Construction 35 (2013) 314–328

317

6.2.1. KM-C Sustainability and buildability in building envelope design have gained more importance in recent years. This can be seen for example in Singapore where there are a number of regulations implemented to promote sustainability and buildability in building designs [40]. However, compliance with these regulations does not guarantee satisfactions of the stakeholders of a building project [5]. Singhaputtangkul et al. [40] suggested major criteria applied by the architects and engineers in Singapore for achieving sustainability and buildability in the assessment of the building envelope materials and designs as shown in Table 1. These criteria were grouped into four categories which are environmental, economic, social and buildability criteria and then stored in the KM-C. This system covers the “Criteria for sustainability and buildability” and “Criteria for individual material assessment” tables in Fig. 5. The knowledge in these tables includes description, relevant laws and regulations, importance weights of the criteria and contribution weights of the building envelope materials. In addition, the system also allows the DMs to manage the knowledge, keep it current and add new criteria.

Fig. 3. Overall research methodology of this study.

6.1. HOQ The HOQ is the central element of this tool functioning to structure the decision making process and to integrate the other elements of the tool together. As a result of modifying the conventional HOQ, a House of Quality for Sustainability and Buildability (HOQSB) was built to facilitate the assessment of the building envelope materials and designs. The HOQSB has five major rooms which are Criteria for sustainability and buildability room (CR), Building envelope materials and designs room (MR), Relationships between the criteria and materials and designs room (RR), Fuzzy inference engine for prioritizing design alternatives room (FR) and Preference list room (PR). The CR and MR are applied to identify relevant criteria and building envelope design alternatives, respectively. The RR contains the relationships between the criteria and design alternatives. These relationships include a matrix to indicate the parameters affecting each criterion and rules to guide the DMs when assessing the building envelope materials and designs. The FR stores fuzzy calculation techniques operated by a fuzzy inference engine for prioritizing the design alternatives. The PR records outputs of the FR in the form of a preference list of the design alternatives ranked by a Sustainability and Buildability Index (SBI). In brief, this index is a function of importance weights of the criteria, contribution weights of the building envelope materials, and performance satisfactions of the materials and design alternatives with respect to the criteria. 6.2. KMS The KMS is made of Knowledge management of the criteria system (KM-C), Knowledge management of the materials and designs system (KM-M) and Knowledge management of relationships between the criteria and design alternatives system (KM-R). As shown in Fig. 4, the KM-C, KM-M and KM-R of the KMS serve as the database of the CR, MR and RR of the HOQSB, respectively. The knowledge stored in these systems was acquired from the semi-structured interviews and literature reviews. Fig. 5 illustrates a simplified relational diagram of the KMS.

6.2.2. KM-M As there are many possible materials and designs, the scope of the KM-M was limited in the first instance to only basic building envelope materials and designs. Table 2 illustrates the basic building envelope materials and their corresponding design alternatives considered in this study. For example, alternative “1” is made of “PC1” Precast wall, “WG1” Single layer window glazing and “SD1” Horizontal concrete shading device. The KM-M covers the “Alternative properties” and “Material properties” tables in Fig. 5. The knowledge stored in the KM-M includes thermal transmittance (U-value), dimensions, costs, construction techniques, relevant technical standards and specifications of the materials, etc. New materials and hybrid designs can also be added into this system. 6.2.3. KM-R The KM-R was developed to manage the relationships between the criteria and the building envelope materials and designs. This system covers the “Performance of individual material”, “Performance of design alternative” and “Relationship matrix” tables in Fig. 5. As the names suggest, the first two tables store the performance satisfactions of the individual building envelope materials and design alternatives, respectively. The last table records important parameters of the materials and designs affecting each criterion, and collectively presents these relationships in the form of IF-THEN rules. An example of these rules for the assessment of the performance satisfaction of a design alternative with respect to the “SC2” Appearance demands is “If the design alternative supports esthetics, trend and image of a building project, then the performance satisfaction of the design increases”. 6.3. Fuzzy inference engine In a real-world decision situation, making decisions has to process not only large amount of information but also subjective and uncertain requirements [41]. Particularly, DMs may encounter practical constraints from several criteria perhaps containing imprecision, subjective or vagueness inherent in the information [38]. It was suggested that the fuzzy set theory introduced by Zadeh [42] can mitigate this problem by translating subjective information, incomplete information and partially ignorant facts into the decision model. Hence, the study adopted the fuzzy set theory for evaluating the preferences of the DMs for the assessment of the building envelope materials and designs [25,26]. By taking into consideration the knowledge provided by the KM-C, KM-M and KM-R, the DMs of the design team rate the importance weights of the criteria and performance satisfactions of the materials and design alternatives through the use of fuzzy linguistic terms. The tool then prioritizes

318

N. Singhaputtangkul et al. / Automation in Construction 35 (2013) 314–328

Fig. 4. Architecture of the KBDSS-QFD tool.

the building envelope design alternatives in the form of the SBI as shown in Fig. 6 [40]. Under fuzzy operations, the SBI of each design alternative is a sum of products of the performance satisfactions of the design alternative with respect to the criteria and the importance weights of such criteria. The tool allows two types of the criteria for assessment of the performance satisfaction of the design alternative; namely criteria for overall design assessment and criteria for individual material assessment. The performance satisfaction of the design alternative

1

with respect to the criteria for overall design assessment is determined by the performance satisfaction of that alternative as a whole. In contrast, the performance satisfaction of the design alternative with respect to the criteria for individual material assessment is modeled by a sum of products of the performance satisfactions of the materials that assemble such alternative and contribution weights of the corresponding materials. This structure is provided as an option for the DMs if there is a need to break down the performance satisfaction of the design alternative into the performance

Performance of individual material Material ID (PK) Performance with respect to criteria

1 Alternative properties 1 Project name Alt ID (PK) Material ID (PK) WWR Orientation Green Mark Score Initial cost Floor-to-floor Height Location Client info etc

1

Material properties

MaterialID (PK) Material type External finishes 1..* Thickness Height Width Length Uw STC SC VT Initial cost etc

1

1

Performance of design alternative Alt ID (PK) Performance with respect to criteria

1

1

Fig. 5. Relational diagram of the KMS.

1

Criteria ID (PK) Criteria name Description Compliance Importance weight

Relationship matrix Criteria ID (PK) 1 Criteria name Material ID (PK) External finishes Thickness Height Width Length Uw STC SC VT Initial cost IF-THEN rule etc

Criteria for sustainability and buildability

1

1

Criteria for Individual material assessment Criteria ID (PK) Criteria name Wall contruibution weight Window contribution weight Shading contribution weight

0..1

N. Singhaputtangkul et al. / Automation in Construction 35 (2013) 314–328 Table 1 Criteria for sustainability and buildability. Criteria for achieving sustainability and buildability Sustainability

Environmental

Economic

Social

Buildability

EN1: Energy consumption EN2: Resource consumption EN3: Waste generation EC1: Initial costs EC2: Long-term burdens EC3: Durability SC1: Energy efficiency SC2: Appearance demands SC3: Health, safety and security of occupants and society SC4: Weather protection performance SC5: Acoustic protection performance SC6: Visual performance BC1: Health and safety of workers BC2: Simplicity of design details BC3: Material deliveries from suppliers BC4: Material handling BC5: Ease in construction with respect to time BC6: Community disturbance

satisfactions of the materials individually for achieving a better estimation.

6.3.1. Fuzzy linguistic terms This study adopted the triangular fuzzy numbers, as shown in Table 3, to define the fuzzy linguistic terms, as shown in Fig. 7, for assessing the importance weights of the criteria, contribution weights of the building envelope materials and performance satisfactions of the materials and designs [38].

Building envelope material Window glazing

It is assumed that there are n DMs in the design team who assess the importance weights of k criteria and performance satisfactions of g materials and f design alternatives. A linguistic set of both the importance and contribution weights is; W = (very unimportant (VU), unimportant (U), medium (M), important (I), very important (VI)). The fuzzy num
 ~ tj ¼ p ; q ; r tj bers of the importance and contribution weights are W tj tj
 ~ atj ¼ datj; eatj; f and W atj , respectively, where t = (1, 2, ⋯ , k), a = (external wall, window glazing, shading device,…, g) and j = (1, 2, ⋯ , n). A linguistic set for the performance satisfactions of both the materials and design alternatives is; A = (very unsatisfied (VU), unsatisfied (U), fair (F), satisfied (S), very satisfied (VS)). Assigned by the j DM to the g material and f design alternative with respect to the k criteria, the fuzzy numbers of the performance satisfactions of the materials and design al
 ~ ait ¼ g haijt; laijt and A ~ it ¼ aijt; bijt; cijt , respectively, ternatives are A aijt;

where i = (1, 2,…, f). 6.3.2. Fuzzy operations Based on the extension principle [44], the fuzzy operations for calculating the SBI in this study consist of the following six major steps [7,43]: ~ C , and contriStep 1: Assess the importance weights of the criteria, W t C ~ bution weights of the materials, W at , through a fuzzy aggregation operation based on Eqs. (1) and (2), respectively. ~C¼ W t

  n ptj n qtj n r tj ∑j¼1 ;∑j¼1 ;∑j¼1 n n n

ð1Þ

~C ¼ W at

  n dtj n etj n f tj ∑j¼1 ;∑j¼1 ;∑j¼1 n n n

ð2Þ

Step 2: Determine the performance satisfactions of the design alternatives with respect to the criteria for overall design assess~ C , and performance satisfactions of the materials ment, A it with respect to the criteria for individual material assessC ~ , through a fuzzy aggregation operation based on ment, A ait Eqs. (3) and (4), respectively.

Table 2 Building envelope materials and alternatives.

External wall

Shading device

~C ¼ A it PC1

CB1

WG1 SD1

BL1

WG2

a

b SD2

CI1

WG3

FG1

WG4

SD3

CW1

WG1: Single layer

SD1: Horizontal concrete

CB1: Brick

WG2: Low-e single layer

SD2: Horizontal aluminum

BL1: Block CS1: Cast in-situ

WG3: Double layer

SD3: None

WG4: Low-e double layer

FG1: Fixed-glass

  n g aitj n haitj n laitj ; ∑j¼1 ;∑j¼1 ∑j¼1 n n n

PC1

WG1


 ~C ¼ ∑ W ~ C =∑ W ~ C ⊗A ~C A a a it at at ait

ð4Þ

  ∑a ðg  dÞ ∑a ðh  eÞ ∑a ðl  f Þ ; ; ∑a d ∑a e ∑a f

ð5Þ ð6Þ

Step 4: Determine fuzzy preference index of the design alternative, F~ i , through a fuzzy multiplication operation based on Eqs. (7) and (8).

SD1

For the precast concrete wall, only the concrete shading device prefabricated as part of the panel by the manufacturer is considered. For the brickwall, concrete blockwall, and cast in-situ RC wall, only the concrete shading device installed on site is considered. b For the fixed-glass and glass curtain wall, only the aluminum shading device installed on site is considered.

ð3Þ

Step 3: Determine the performance satisfactions of the design alternative based on the performance satisfactions of the mate~ C , through a fuzzy multiplication operation based on rials, A it Eqs. (5) and (6).


 t ~ C =∑t W ~ C ⊗A ~C F~ i ¼ ∑1 W t it 1 t

CW1: Curtain wall

a

  n aitj n bitj n citj ;∑j¼1 ;∑j¼1 ∑j¼1 n n n

~C ¼ A ait

~C ¼ A it

PC1: Precast

Alternative 1

319

F~ i ¼

  ∑t ða  pÞ ∑t ðb  qÞ ∑t ðc  r Þ ; ; ∑t p ∑t q ∑t r

ð7Þ ð8Þ

Step 5: Convert the fuzzy preference index, F~ i , into a crisp number by ~ ¼ ðd ; d ; d Þ, could be assuming that fuzzy number, D 1 2 3

320

N. Singhaputtangkul et al. / Automation in Construction 35 (2013) 314–328

Fig. 6. A flowchart to calculate the SBI.

converted into the crisp number through a defuzzification operation based on Eq. (9). Si ¼ ðd1 þ d2 þ d3 Þ=3

ð9Þ

where Si is the SBI. Step 6: Translate the fuzzy number into fuzzy linguistic term based ~ is “approximateon the assumption that the fuzzy number D ly the linguistic term A”, when it has the membership function as shown in Eq. (10). For this study, (b–a) and (c–b) for each of the linguistic terms is equal to 1. As a result, Eq. (11) shows the μA(x) representing the possibility that ~ is “approximately the linguistic term A”. the fuzzy number D 8 0; xba; or x > c > > < x−a ; a≤x≤b b−a > > : c−x ; bbx≤c c−b

μ A ðxÞ ¼

μ A ðxÞ ¼

8 < :

0;

ð10Þ

xba; or x > c x−a; a≤x≤b c−x; bbx≤c

ð11Þ

where x is the crisp number converted by Eq. (9). Table 3 Fuzzy numbers of the weights and performance satisfactions. Importance/contribution weight

Performance satisfaction

Fuzzy number (a, b, c)

Very unimportant (VU) Unimportant (U) Medium (M) Important (I) Very important (VI)

Very unsatisfied (VU) Unsatisfied (U) Fair (F) Satisfied (S) Very satisfied (VS)

(0,0,0.25) (0,0.25,0.5) (0.25,0.50,0.75) (0.50,0.75,1.0) (0.75,1.0,1.0)

  μ ðxÞ ∑yu¼1 Au Au ~ which is could represent the possibility that the fuzzy number D, “approximately the linguistic terms A1, A2, ⋯ , Ay”, the triangular fuzzy ~ can be transformed into the linguistic terms, Az, where number D Furthermore, if it is assumed that the fuzzy set; A ¼

1 b z b y, as shown in Eq. (12). μ Az ðxÞ Az

  μ A ðxÞ y ¼ max ∑u¼1 u Au

ð12Þ

Calculation example of these steps can be found in the studies of Yang et al. [7] and Bayrak [43]. 6.3.3. Fuzzy consensus scheme When dealing with multicriteria decision problems under group settings, conflicting opinions among group members are very likely to occur, even in a cooperative environment. A fuzzy consensus scheme is found useful in helping the team to achieve consensus solutions, thereby mitigating potential disagreement between opinions of DMs [45]. A consensus level is an essential tool in this scheme applied to measure the degree of compatibility between the fuzzy linguistic terms of DMs. The consensus level, ranging from 0 to 1, is a function of fuzzy distance and fuzzy similarity, and is modeled to quantify how far a group of DMs is from perfect agreement [25,26]. In theory, the value of 1 corresponds to full and unanimous concordance, whereas 0 refers to nonexistent concordance [46]. This scheme was adopted in this study; however, to keep the scope of the tool manageable in the first instance, the consensus level for making the decisions by the design team comprising three DMs was divided into three levels which are “High”, “Medium” and “Low” consensus levels. The decision receives the “High” consensus level if all the three DMs give the same linguistic term, or if any pairs of the DMs share the same linguistic term, while the other DM gives the linguistic term next to it. The decision obtains the “Medium” consensus level if all the three linguistic terms assigned by the DMs

N. Singhaputtangkul et al. / Automation in Construction 35 (2013) 314–328

321

Fig. 7. Triangular fuzzy linguistic terms applied in this study.

can be arranged in relative order and right next to each other, regardless of which DM is responsible for each linguistic term. The rest of the combinations receive the “Low” consensus level. Table 4 presents example of decision results and their corresponding consensus levels for assessment of the importance weights. Fig. 8 illustrates how the fuzzy consensus scheme is operated. After setting the fuzzy linguistic terms and numbers, the DMs establish freezing conditions for the assessment. These conditions include a minimum consensus level, maximum assessment cycle of the individual DM and maximum assessment cycle of the team. In the first assessment cycle of the team on any decision, if the consensus level of the team for that decision meets the minimum consensus level agreed, the team moves on to make the next decision. However, if the consensus level of that decision is lower than the minimum consensus level, a team facilitator invites the least concordant DM to explain his/her reason to the group for the group discussion and to reassess that particular decision. It is noted that if there is more than one least concordance DM, the reassessment takes place on a voluntary basis. The least concordant DM may or may not change his/her decision, but this increases both the number of the assessment cycle of that DM and the team by one. This loop goes on until one of the freezing conditions is met. In addition, to maintain a conducive atmosphere for the team, in the event where the least concordant DM does not change the decision, the second least concordance DM is invited to reassess his/her decision and so on [26,46]. Based on the HOQSB and fuzzy inference engine developed, this study formed seven steps for the DMs to provide their inputs through the user interface of the tool to determine the SBI as follow:

Step 6: Assess the performance satisfactions of the design alternatives with respect to the criteria selected for overall design assessment. Step 7: Assess the performance satisfactions of the materials with respect to the criteria selected for individual material assessment. As can be seen, these seven steps correspond with the rooms in the HOQSB. The CR governs Step 2 to Step 4, whereas the MR administers Step 5. The RR then enables the DMs to make the decisions regarding Step 6 and Step 7. Subsequently, the FR takes the inputs from Step 1 to Step 7 to calculate the importance weights, performance satisfactions, consensus levels and SBI, and then records a summary of these in the PR. 7. A case study The design team comprising an architect, C&S engineer and M&E engineer was engaged to develop a conceptual building envelope design of a representative private high-rise residential building project in Singapore. There were a series of individual meetings between the researcher and each DM prior to holding a team meeting to allow the DMs to be familiar with using the tool and with the project information and its criteria as shown in Tables 5 and 6, respectively. During the team meeting, the researcher acted as a facilitator to operate the tool by presenting the project information, its criteria and then following through the seven steps for determining the SBI. Step 1: The team entered relevant information of the project and set up the fuzzy linguistic terms as shown in the actual screenshots in

Step 1: Input project relevant information, and set up membership functions of the triangular fuzzy linguistic terms and freezing conditions of the fuzzy consensus scheme. Step 2: Select the criteria for the assessment and decide whether the criteria are for overall design or individual material assessment. Step 3: Assess the importance weights of all the criteria selected. Step 4: Assess the contribution weights of the materials with respect to the criteria chosen for individual material assessment, if any. Step 5: Select the building envelope materials to form the design alternatives.

Table 4 Examples of the consensus levels with respect to different decisions. Decision example

1 2 3 4 5 6

Importance weight DM1

DM2

DM3

VU M VU I U VI

VU U M VI I VI

VU M U M I M

Consensus level

Least concordance DM

High High Medium Medium Low Low

None DM2 DM1 or DM2 DM2 or DM3 DM1 DM3

Fig. 8. Fuzzy consensus scheme applied in this study.

322

N. Singhaputtangkul et al. / Automation in Construction 35 (2013) 314–328

Table 5 General project information for the case study. Developer Project title Contract type Project location Preferred external wall material Orientation/plan configuration WWR Height Floor-to-floor Area per floor Design and construction period

Condominium developer High-rise residential building Design–bid–build Central area of the city Curtain wall vs. fixed-glass North–south/square 0.3 75 m 3m 400 m2 33 months

Fig. 9. The team adopted the minimum consensus level of “Medium”, maximum assessment cycle of an individual DM of two cycles, and maximum assessment cycle of the group of three cycles as the freezing conditions of the fuzzy consensus scheme. It is noted that the numbers of these cycles were manually recorded by the facilitator. Step 2: The team selected the criteria as given in Table 6. Apart from these criteria, since the project would be located in a central area of the city, after having gone through the full list of the criteria provided by the KM-C, the team agreed that access to site, transportation of materials and community disturbance were also major concerns of this project and should be taken into account. As a result, the “BC3” Material deliveries from suppliers and “BC6” Community disturbance were added into the assessment, contributing to a total of twelve criteria selected for the assessment. The “EC1” Initial costs, “SC1” Energy efficiency, “SC2” Appearance demands, “SC4” Weather protection performance, “SC5” Acoustic protection performance and “SC6” Visual performance were selected as the criteria for overall design assessment. By default, the rest of the criteria were automatically recorded as the criteria for individual material assessment to provide a systematic evaluation. Step 3: The DMs rated the importance weights of the criteria selected in consideration of the guided importance weights and relevant knowledge stored in the KM-C. Fig. 10 shows the screenshot for rating the importance weights of the “BC3” Material deliveries from suppliers, “BC5” Ease in construction with respect to time and “BC6” Community disturbance. The tool employs Eq. (1) and the fuzzy consensus scheme to calculate the collective importance weights and consensus levels, respectively. The weights were then converted back to the linguistic terms by applying Eq. (12). In this step, out of the twelve criteria selected, nine criteria received the same weights as suggested by the KM-C, while the other three criteria which are the “EN3” Waste generation, “BC3” Material deliveries from suppliers and “BC6” Community disturbance

received a higher weight due to increasing concerns over the impacts of the project on the surrounding environments during the construction period. Considering the consensus level, a majority of the decisions received the “High” consensus level in the first assessment cycle of the team. There were two decisions for rating the importance weights of the “EC2” Long-term burdens and “SC1” Energy efficient that obtained the “Medium” consensus level in the second assessment cycle, and one decision for rating the importance weight of the “EN3” Waste generation that received the “Medium” consensus level in the third assessment cycle of the team. This seemed to suggest that the perspectives among the DMs on the importance weights of these three criteria appeared to be more divergent than the others. Step 4: The DMs as a team rated the contribution weights of the external wall, window glazing and shading device with respect to the criteria assigned for individual material assessment. Fig. 11 presents the screenshot for rating such contribution weights regarding the “BC3” Material deliveries from suppliers, “BC5” Ease in construction with respect to time and “BC6” Community disturbance. Step 5: As Table 5 suggests that the preferred external wall materials include curtain wall and fixed-glass wall, the DMs selected “CW1” Glass curtain and “FG1” Fixed-glass as the external wall material options, “WG3” Double layer glazing and “WG4” Low-E double layer glazing as the window glazing material options, and “SD2” Horizontal aluminum shading as the shading device material option. According to this selection and Table 2, four design alternatives which are the alternative “47” CW1WG3SD2, “48” CW1WG4SD2, “39” FG1WG3SD2 and “40” FG1WG4SD2 were extracted from the KM-M as shown in the screenshot given in Fig. 12. Step 6: The DMs rated the performance satisfactions of these four design alternatives with respect to the criteria for overall design assessment. The screenshot as shown in Fig. 13 reflects rating the performance satisfactions of the design alternatives with respect to the “SC2” Appearance demands in consideration of the guided performance satisfactions, relationship matrix and IF-THEN rule. Eqs. (3) and (12) were applied to determine the collective performance satisfactions of the alternatives in the forms of the fuzzy numbers and linguistic terms, respectively. Although a majority of the decisions in this step received the same performance satisfactions as suggested by the KM-R, as can be seen in Fig. 13, the performance satisfactions of the alternative “39” FG1WG3SD2 and “40” FG1WG4SD2 with respect to the “SC2” Appearance demands appeared to be lower than the guided performance satisfaction as there were two DMs viewing that the fixed-glass wall design-based alternatives do not capture the appearance demands of the project well. In

Table 6 Project criteria for the case study. Criteria

Criteria name

Detail

Environmental

EN3: Waste generation

Economic

EC1: Initial costs EC2: Long-term burdens SC1: Energy efficiency

Waste generation especially air pollution and wastewater should be contained to reduce their impacts on the surrounding environments. The project budget must be minimized The building envelope design must decrease long-term burdens particularly repairing and replacing costs. Energy efficiency of the design must be maximized to achieve high Green Mark (GM) Scorea and occupant comfort. Appearance demands of the design must be maximized and the design must be modern and represent positive image. Health, safety and security of the occupants and society must be enhanced. The design should reduce negative influence from adverse weather during occupation phase. The design should reduce adverse acoustical impacts from both indoor and outdoor activities. Visual performance of the design should be maximized to achieve high occupant comfort. The material, design and construction techniques should be labor efficient while promoting high buildability.

Social

SC2: Appearance demands

Buildability a

SC3: Health, safety and security of occupants SC4: Weather protection performance SC5: Acoustic protection performance SC6: Visual performance BC5: Ease in construction with respect to time

Calculation of the GM Score can be found in Singhaputtangkul et al. [5].

N. Singhaputtangkul et al. / Automation in Construction 35 (2013) 314–328

323

Fig. 9. Project information and fuzzy linguistic terms.

addition, all the decisions in this step received either the “High” or “Medium” consensus levels within the second assessment cycle of the team. Step 7: The DMs assessed the performance satisfactions of the materials with respect to the criteria chosen for individual material

assessment. Fig. 14 presents the screenshot for rating the performance satisfactions of the individual materials of each alternative with respect to the “BC3” Material deliveries from suppliers. Eqs. (4) and (12) were applied to determine the performance satisfactions in the form of the fuzzy numbers and

Fig. 10. Assessment of the importance weights for all the criteria.

324

N. Singhaputtangkul et al. / Automation in Construction 35 (2013) 314–328

Fig. 11. Assessment of the contribution weights for the criteria chosen for individual material assessment.

linguistic terms, respectively. It is noted that, in this step, a majority of the decisions obtained the “High” consensus level in the first assessment cycle. Fig. 15 presents the screenshot of the tool showing a summary of the importance weights of the criteria, performance satisfactions of the design alternatives and their corresponding SBI calculated by Eqs. (8) and (9). As can be seen, although the alternative “39” FG1WG3SD2 and “40” FG1WG4SD2 received a higher performance satisfaction with respect to the “EC1” Initial costs as compared to that of the alternative “47” CW1WG3SD2 and “48” CW1WG4SD2, the latter pair obtained higher performance satisfactions with respect to the “SC2” Appearance demands, “BC3” Material deliveries from suppliers and “BC6” Community disturbance. This contributed to their higher SBI overall. Furthermore, comparing between the alternative “47” CW1WG3SD2 and “48” CW1WG4SD2, the latter posed a higher performance satisfaction with respect to the “SC1” Energy efficiency due to energy-saving applications of the low-E window glazing. For this reason, its SBI appeared to be slightly higher. In conclusion, the design team members mutually agreed to adopt the alternative “48” CW1WG4SD2 as a base case for the conceptual design of the project. The team took approximately 3 h in this case study to reach a consensus through clear and step-by-step deliberations. 8. Findings and discussions Qualitative data analysis was applied to analyze findings from the interviews conducted with the DMs after completing the case study. From this analysis, the DMs agreed that using the tool coupled with the knowledge provided by the KMS helped the design team at the early design stage to consider key criteria required for the assessment. This also reminded the DMs about relevant regulations, reasons for compliance, description and importance of each criterion. Additionally, the tool facilitated the team to collectively consider the criteria altogether at once. The literature review supports that, instead of redesigning a product, when design parameters are changed, or when

new assessment criteria have to be additionally considered, the design would be more comprehensive if a full set of the criteria can be identified before conducting such design deliberations [40]. Regarding mitigation of the decision-making problem related to inadequate consideration of possible building envelope materials and designs, the DMs agreed that the tool assisted the team to consider various basic building envelope materials and designs in the first instance. In particular, before making the decisions, the KBDSS-QFD tool provided the DMs with the building envelope materials and designs for consideration coupled with their relevant design-, construction- and procurement-related information with respect to all the criteria. This gave the DMs an instant access to information relating to important properties of such materials and alternatives enabling the designers to evaluate a wider range of possible design alternatives. Findings from the interviews also suggested that the KMS containing a wealth of useful knowledge helped the DMs to overcome cognitive limitation of knowledge, to increase consensus and confidence of the team, to reduce bias when dealing with similar decisions and to make a prompt response. Hence, the KMS has played an important role in enhancing the efficiency and consistency in making the decisions of the design team. Similarly, Vat [47] and Wegner [48] reported benefits of applying a well-established KMS such as improvement of organizational learning, business resilient, business competitiveness, etc. Furthermore, Arain and Low [20] supported that an established KMS storing relevant knowledge and creating several situational decisions can assist building professionals in learning from similar situational decisions. In addition, from the interviews, the DMs held similar views that the user interface of the tool supported participation and decision-making of the team members. It was further highlighted that using the tool as a team through the structured decision-making steps strengthened the communication and integration process among the team members as compared to traditional way to assessing the building envelope materials and designs. Supporting this, Holsapple and Whinstone [49] pointed out that a computerized tool provides a smoother decisionmaking process and saves significant time in processing inputs, thereby mitigating lack of communication and integration among the DMs.

N. Singhaputtangkul et al. / Automation in Construction 35 (2013) 314–328

325

Fig. 12. Building envelope design alternatives of the case study.

Regarding mitigation of the decision-making problem concerning subjective and uncertain requirements, the DMs showed the positive attitude that the KBDSS-QFD tool provided a systematic and structured approach that can enhance the team to analyze design information, to generate the design alternatives, and to deliver the optimal design solutions through the use of the fuzzy inference engine. It was observed that the tool equipped with the fuzzy techniques

sufficiently captured complex and imprecise perspectives of the designers, and it can translate these into a more simple form, the SBI. Chou and Chang [50] and Juan et al. [51] agreed that applying the fuzzy set theory helped professionals to determine a meaningful set of solutions. Additionally, as the study applied the fuzzy consensus scheme to mitigate disagreement between opinions of the DMs, findings from

Fig. 13. Assessment of the performance satisfactions of the design alternatives.

326

N. Singhaputtangkul et al. / Automation in Construction 35 (2013) 314–328

Fig. 14. Assessment of the performance satisfactions of the individual materials.

the interviews suggested that the consensus level reminded the team to discuss and clarify related issues prior to making the decisions. Furthermore, the DMs also felt that they had an equal opportunity to influence the decision and would continue to support the group. In accordance with this, Ekel [45] found that the fuzzy consensus scheme can enhance discussion and communication between members of a team. This could be because the concept of the scheme lies upon continuous discussion and negotiation in the group until everyone affected through understanding, agree with what will be done [26].

With these findings in mind, the hypothesis that the tool can facilitate the design team to mitigate the decision-making problems as a whole was supported. Nevertheless, a few comments were obtained from the interviews for future improvement of the tool. The KBDSSQFD tool was perceived to be a bit complicated due to its many functions. This makes the assessment quite dependent on the team facilitator and preliminary discussions between the facilitator and individual DMs. In addition, as the tool contains a wealth of rich knowledge from different designers, updating the KMS could be a time-consuming

Fig. 15. Summary of the design solutions for the case study.

N. Singhaputtangkul et al. / Automation in Construction 35 (2013) 314–328

process. It was also suggested that selecting the criteria and materials for the assessment still relatively relies on experience of the design team and availability of project information.

9. Conclusions Success of the assessment of the building envelope materials and designs for private high-rise residential buildings is affected by several decision-making problems faced by the architects and engineers as part of the design team. These problems include inadequate consideration of requirements, inadequate consideration of possible materials and designs, lack of efficiency and consistency, lack of communication and integration between members of the team, subjective and uncertain requirements, and disagreement between members of the team. The objective of this study is to develop the KBDSS-QFD tool comprising three main elements which are the HOQSB, KMS and fuzzy inference engine to mitigate these problems as a whole. The case study of the design team consisting three DMs is selected as the research design to fulfill this objective. Results from the qualitative data analysis show that the KBDSSQFD tool can be used to remind the DMs about key criteria and possible building envelope materials and designs for the assessment of the building envelope materials and designs. It also improves efficiency as well as consistency in making the decisions for the assessment by facilitating the DMs to make a prompt decision and to learn from past experience. In addition, through the structured decision process offered by the tool, communication and integration among the DMs are enhanced. Mean time, the fuzzy inference engine embedded with the fuzzy techniques assists the team in translating subjective and uncertain requirements into a more useful format, and the consensus scheme helps the team to reduce disagreement among opinions of its members. The main contributions of this study can be found in both the project and industry level. At the project level, the tool can assist the design team to overcome the decision-making problems identified thereby delivering better design management, and to determine the SBI that can offer a good balance between the environmental, economic, social and buildability criteria. This would be helpful for the team in presenting more informed decision results to a developer and other stakeholders of a project. At the industry level, the tool contributes to an integrative way of raising the existing level of productivity in the construction industry by promoting the use of more sustainable and buildable materials and designs. Importantly, the findings of this study not only broaden a body of academic knowledge related to the assessment of the building envelope materials and designs, but also guide future studies in developing a practical decision support tool for dealing with decision-making problems in other industrial contexts.

References [1] C.J. Kibert, Sustainable Construction: Green Building and Delivery, second ed. John Wiley and Son, Inc., New Jersey, 2008. [2] Y.L. Chew, Construction Technology for Tall Buildings, World Scientific Publishing, Singapore, 2009. [3] K.J. Chua, S.K. Chou, Evaluating the performance of shading devices and glazing types to promote energy efficiency of residential buildings, Building Simulation (2010) 1–14. [4] J. Boecker, S. Horst, T. Keiter, A. Lau, M. Sheffer, B. Toevs, The Integrative Design Guide to Green Building: Redefining the Practice of Sustainability, John Wiley and Sons, Inc., New Jersey, 2009. [5] N. Singhaputtangkul, S.P. Low, A.L. Teo, Integrating sustainability and buildability requirements in building envelopes, Facilities 26 (5/6) (2011) 255–267. [6] J. Carmody, S. Selkowitz, D. Arasteh, L. Heschong, Residential Windows: A Guide to New Technologies and Energy Performance, third ed. W.W. Norton and Company, New York, 2007. [7] Y.Q. Yang, S.Q. Wang, M. Dulaimi, S.P. Low, A fuzzy quality function deployment system for buildable design decision-makings, Automation in Construction 12 (4) (2003) 381–393.

327

[8] F.M. Arain, S.P. Low, The nature and frequency of occurrence of variation orders for educational building projects in Singapore, International Journal of Construction Management 5 (2) (2005) 79–91. [9] S.P. Low, L.L. T'ng, Factors influencing design development time of commercial properties in Singapore, Facilities 16 (1/2) (1998) 40–51. [10] B. Fryer, The Practice of Construction Management, fourth ed. Blackwell Publishing, Oxford, 2004. [11] W.W.C. Chung, C.K.S. Tam, M.F.S. Chan, Integrated QFD and knowledge management system for the development of common product platform, in: A. Gunasekaran, O. Khalil, S. Mahbubur Rahman (Eds.), Knowledge and Information Technology Management: Human and Social PerspectiveIdea Group Publishing, PA, 2003. [12] BCA (Building, Construction Authority), Code of practice on buildability, Singapore, http://www.bca.gov.sg/BuildableDesign/bdas2011.html 2011, (Accessed 2011 October). [13] L. Brock, Designing the Exterior Wall: An Architectural Guide to the Vertical Envelope, John Wiley and Sons, Inc., New Jersey, 2005. [14] F.E. Gould, Managing the Construction Process: Estimating, Scheduling, and Project Control, third ed. Pearson Prentice Hall, New Jersey, 2005. [15] C.W. Ibbs, W.E. Allen, Quantitative Impacts of Project Change, Construction Management Technical Report No. 23, University of California at Berkeley, USA, 1995. [16] A.E.D. El-Alfy, Design of sustainable buildings through value engineering, Journal of Building Appraisal 6 (2010) 69–79. [17] A.R. Makenya, G. Soronis, Designing for sustainable buildings in developing countries: problems and priorities, 8DBMC – Proceedings of the Eighth International Conference on Durability of Building Materials and Components, Vancouver, Canada, 30 May – 3 June, 1999. [18] T.H. Davenport, L. Prusak, Working Knowledge: How Organizations Manage What They Know, Harvard Business School Press, Boston, 1998. [19] C. McMahon, A. Lowe, S. Culley, Knowledge management in engineering design: personalization and codification, Journal of Engineering Design 15 (4) (2004) 307–325. [20] F.M. Arain, S.P. Low, Knowledge-based decision support system for management of variation orders for institutional building projects, Automation in Construction 15 (2006) 272–291. [21] J. Marsot, QFD: a methodological tool for integration of ergonomics at the design stage, Applied Ergonomics 36 (2) (2005) 185–192. [22] S. Austina, A. Newtonb, J. Steeleb, P. Waskett, Modeling and managing complex project, International Journal of Project Management 20 (3) (2002) 191–198. [23] M. Kagioglou, R. Cooper, G. Aouad, M. Sexton, Rethinking construction: the generic design and construction process protocol, Engineering Construction and Architectural Management 7 (2) (2000) 141–153. [24] M.K. Malek, Constructability Assessment Using Fuzzy Logic Modeling: Case of Selecting Construction System. , (Unpublished Ph. D. thesis) University of Central Florida, USA, 1996. [25] J. Lu, G. Zhang, D. Ruan, F. Wu, Multi-Objective Group Decision Making: Methods, Software and Applications With Fuzzy Set Techniques, Imperial College Press, London, 2007. [26] W. Pedrycz, P. Ekel, R. Parreiras, Fuzzy Multicriteria Decision-making: Models, Methods and Applications, John Wiley and Sons Ltd., West Sussex, 2011. [27] P.C. Nutt, The identification of solution ideas during organizational decisionmaking, Management Science 39 (9) (1993) 1071–1085. [28] S. Jacksons, Team composition in organizations, in: S. Worchel, W. Wood, J. Simpson (Eds.), Group Process and Productivity, Sage Publications, London, 1992. [29] G. Stasser, W. Titus, Pooling of unshared information in group decision making: biased information sampling during discussion, Journal of Personality and Social Psychology 48 (1985) 1467–1478. [30] K.J. Behfar, R.S. Peterson, E.A. Mannix, W.M.K. Trochim, The critical role of conflict resolution in teams: a close look at the links between conflict type, conflict management strategies, and team outcomes, Journal of Applied Psychology 93 (1) (2008) 170–188. [31] D. Robey, D.L. Farrow, C.R. Franz, Group process and conflict in system development, Management Science 35 (10) (1991) 1172–1191. [32] I. Dikmen, M.T. Birgonul, S. Kiziltas, Strategic use of quality function deployment (QFD) in the construction industry, Building and Environment 40 (2005) 245–255. [33] M. Xie, K.C. Tan, T.N. Goh, Advanced QFD Applications, ASQ Quality Press, Milwaukee, 2003. [34] S.P. Low, L. Yeap, Quality function deployment in design/build projects, Journal of Architectural Engineering 7 (2) (2001) 30–39. [35] L. Cohen, Quality function deployment—how to make QFD work for you, Addison-Wesley Reading, MA (1995). [36] V. Bouchereau, H. Rowlands, Methods and techniques to help quality function deployment (QFD), The International Journal of Quality and Reliability Management 7 (1) (2000) 8–19. [37] B. Prasad, Concurrent function deployment: an emerging alternative to QFD: conceptual framework, in: M. Sobolewski, M. Fox (Eds.), Adv. In Concurrent Engrg. Proc. CE96 Conf, Technomic Publishing Lancaster, 1996, pp. 105–112. [38] K.C. Lam, T. Ran, M.C.K. Lam, A material supplier selection model for property developers using fuzzy principal component analysis, Automation in Construction 19 (2010) 608–618. [39] E. Turban, J.E. Aronson, T.P. Laing, R. Sharda, Decision Support and Business Intelligence Systems, eight edition Prentice Hall, New Jersey, 2007. [40] N. Singhaputtangkul, S.P. Low, A.L. Teo, Assessing building envelope materials for sustainability and buildability criteria: a conceptual framework, The International Construction Business and Management Symposium (ICBMS2011). Malaysia, 21–22 September, 2011, (10 pp. (Conference proceeding in CD-Rom with no running pagination)).

328

N. Singhaputtangkul et al. / Automation in Construction 35 (2013) 314–328

[41] C. Reilly, Making Hard Decisions with Decision Tools, Duxbury Thomson Learning, CA, 2001. [42] L.A. Zadeh, Fuzzy sets, Information and Control 8 (1965) 338–353. [43] M.Y. Bayrak, N. Çelebi, H. Takin, A fuzzy approach method for supplier selection, Production Planning and Control 18 (1) (2007) 54–63. [44] L.A. Zadeh, The concept of a linguistic variable and its application to approximate reasoning, Information Sciences 8 (1975) 301–357. [45] P. Ekel, J. Queiroz, R. Parreiras, R. Palhares, Fuzzy set based models of multicriteria group decision-making, Nonlinear Analysis: Theory, Models and Applications 71 (1) (2009) 409–419. [46] J.L. Garcia-Lapresta, Favoring consensus and penalizing disagreement in group decision making, Journal of Advanced Computational Intelligence and Intelligent Informatics 12 (5) (2008) 416–421.

[47] K.H. Vat, Developing a learning organization model for problem-based learning: the emergent lesson of education from the IT trenchs, Journal of Cases on Information Technology 8 (2) (2006) 82–109. [48] E. Wegner, Communities of Practice: Learning, Meaning, and Identity, Cambridge University Press, New York, 2002. [49] C.W. Holsapple, A.B. Whinstone, Decision Support Systems: A Knowledge-Based Approach, West Pub. Co., Minneapolis, 1997. [50] S.Y. Chou, Y.H. Chang, A decision support system for supplier selection based on a strategy-aligned fuzzy SMART approach, Expert Systems with Applications 34 (4) (2008) 2241–2253. [51] Y.K. Juan, Y.H. Perng, D. Castro-Lacouture, K.S. Lu, Housing refurbishment contractors selection based on a hybrid fuzzy-QFD approach, Automation in Construction 18 (2) (2009) 139–144.