Using QFD and ANP to analyze the environmental production requirements in linguistic preferences

Expert Systems with Applications 37 (2010) 2186–2196

Contents lists available at ScienceDirect

Expert Systems with Applications journal homepage: www.elsevier.com/locate/eswa

Using QFD and ANP to analyze the environmental production requirements in linguistic preferences YuanHsu Lin a, Hui-Ping Cheng b, Ming-Lang Tseng b,*, Jim C.C. Tsai b a b

Department of Finance MingDao University, Taiwan Department of Business Innovation and Development MingDao University, Taiwan

a r t i c l e

i n f o

Keywords: Fuzzy quality function deployment Analytical network process Sustainable production indicators Environmental production requirements

a b s t r a c t This study is to apply fuzzy quality function deployment (QFD) model with interdependence relations of environmental production requirements (EPRs) aspects and sustainable production indicators (SPIs) criteria for original equipment manufacturing (OEM) firm in Taiwan. At first, to facilitate the main issue of the QFD problem, however, the ‘‘Whats” question of EPRs and ‘‘Hows” problem of the SPIs have to be made, which are two major components and be emphasized on the house of quality matrices. In conjunction with fuzzy sets theory and analytical network process, the systematic analytical procedures are proposed. Subsequently, a case study demonstrated the evaluation process for identifying ‘‘Whats” and ‘‘Hows”. The results of empirical study show that (1) five aspects of EPRs are deemed to have priority to improve the environmental practice; and (2) twenty-two feasible SPIs criteria are practical indicated. In addition, it is suggested that case firm should aware attentively the SPIs and emphasize on exploiting these EPRs effectively. And develop the ‘‘Hows” issues, which should continuously strengthen of the EPRs, respectively. Ó 2009 Elsevier Ltd. All rights reserved.

1. Introduction Leading environmental management issues emphasize the need to understand manufacturing decisions and practices for an original equipment manufacturing (OEM) firm to improve its production process in Taiwan. Various studies argue that manufacturing decisions and choices have to be consistent with the corporate strategy for effective environmental production management (Tseng, Lin, Chiu, & Liao, 2008a, 2008b; Tseng & Lin, 2008). The environmental management has become the most concern of manufacturing firms, which seek for higher levels of green product quality and continuous improvement to keep up with the change throughout the world. However, the environmental practices are dependent on wider aspects to be integrated in order to achieve firm’s goal of waste elimination and lower environmental impact. Hence, firms must integrate environmental aspects to ensure corporate survival and toward sustainable development. A number of studies have explored how to evaluate the environmental production requirements (EPRs) in life cycle assessment (Veleva & Ellenbecker, 2001; Veleva, Hart, Greiner, & Crumbley, 2001). None efforts, have been focused on establishing EPRs evaluation framework for OEM firms, because most OEM firms are based * Corresponding author. Tel.: +886 920 309400. E-mail address: [email protected] (M.-L. Tseng). 0957-4174/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.eswa.2009.07.065

on ISO 14000 manual to reduce the impact of environment (Balzarova & Castka, 2008; Nishitani, 2009; Tsai & Chou, 2009). However, to differentiate from the ISO 14000, this study follows the nature of operations process of OEM firm urges the necessary to justify the EPRs evaluation aspects and criteria. Therefore, an effective and structured EPRs aspects and criteria for OEM firms need to be developed. A challenge of this approach is that EPRs evaluation is always in uncertainty due to environment information are rapidly changes and the aspects are measured in linguistic terms. In many practical cases, the human preference model is uncertain and might be reluctant or unable to assign exact numerical values to describe the preferences. As this result more desirable for the researchers to use fuzzy set theory evaluation. The OEM firms are aim the satisfaction of the customer at the very beginning, namely the product design phase, the approach evaluation bases on total quality management, which offers a vast techniques to ensure the improvement of quality and productivity. Especially, quality function deployment (QFD) is one of these techniques to design the needs of customer and into practical measures. This approach enables the firms to become proactive to quality problems rather than taking a reactive position by acting on customer complaints. In addition, QFD is applied to plan and design new or improved products or services. It employs a crossfunctional team to determine customer needs and translate them into product designs through a structured and well-documented

Y. Lin et al. / Expert Systems with Applications 37 (2010) 2186–2196

framework. This systematic nature evaluates the necessary decisions for change and development at the beginning of the design process, reducing and avoiding the project changes and corrections. Hence, in real situation, numerous criteria are interdependence when evaluating an OEM firm in EPRs. As a typical study to understand the interdependence relations, for instance, a firm that has outstanding energy and material for natural environment to performance in controlling the waste elimination and hazardous for natural environment such as controlling process for producing and improving green products. The analytical network process (ANP) is most suitable tool for solving interdependence relations among the criteria and attributes (Saaty, 1996). The merits of ANP are as follows: (i) both tangibles and intangibles, individual values, and shared values can be included in the decision process; (ii) the discussion can be structured so that every criteria relevant to the decision is considered; and (iii) in a structured analysis, the discussion continues until relevant information from each individual member in the group is considered and a consensus is achieved. In addition to these merits of ANP provides a more generalized model in decision-making without making assumptions about the independence of the aspects and criteria (Dyer & Forman, 1992). The aim of this study was to get a better understanding of a particular, strategic operating decision area in improvement of production process toward sustainability. This study attempts to develop a EPRs hierarchical framework that is sufficiently general that it can be applied. To date, few studies have adopted such a rigorous methodology. This study presents a combined of fuzzy set theory, ANP and QFD that is sufficiently general and it can be applied under various study settings. This firm evaluation in this EPRs assessment can help firms to develop the measures of optimal sustainable production indicators (SPIs). Consequently, resolving problems in evaluating firm is fundamentally important to both researchers and practitioners. The unique point of this study is involved in qualitative measures in linguistic terms presented by triangular fuzzy numbers (TFNs) and defuzzified into a crisp value for analyze in interdependence relations among aspects and criteria, and apply ANP and QFD to acquire the optimal decision-making. This study begins with a brief introduction of the EPRs and the study objectives. Section 2 follows a literature discussion of the EPRs and SPIs and related literatures. Section 3 presents the proposed method. Section 4 presents the empirical study case. A study framework suggests providing a context for applied the proposed method. Section 5 concludes with a summary of the study finding of the method as well as recommendations for its further development and practical application.

2. Literature review Due to the increased green competitive pressures are also forcing firms to continuously develop to enhance production competitiveness for OEM firms. For this reason, a firm must develop and evaluate the EPRs more rapidly than other firms, and must facilitate SPIs within its organization to strengthen its green competitive advantage. Lowell Center for Sustainable Production (LCSP) (1998) defined sustainable production as the creation of goods and services using processes and systems that are non-polluting; conserving of energy and natural resources; economically viable; safe and healthful for employees, communities and consumers; and socially and creatively rewarding for all working people. This definition is consistent with current understanding of sustainable development, since it emphasizes environmental, social and economic aspects of firms’ activities. Only recently have a growing number of firms begun to use environmental, health and safety, and social indicators (Tseng, 2009; Tseng & Lin, 2008).

2187

Based on differing EPRs perspectives, some studies have presented the indicators or constrains for the sustainable production measures. Existing business-related sustainability indicators tend to emphasize the environmental aspects of production. However, Veleva et al. (2001) argue that SPIs should include not only production measures but also measures of the economic and social. Therefore, they proposed a framework that consists of five levels for categorizing existing indicators relative to the basic principles of sustainability. The study purpose provided a method to evaluate a set of indicators focusing on environmental, health and safety aspects of production and work is underway to expand it to include social and economic aspects, to inform decision-making and measure progress toward more sustainable production. Veleva and Ellenbecker (2001) presented a set of indicators of sustainable production for promoting business sustainability. It first introduces the concept of sustainable production as defined by the LCSP, including six dimensions and desirable qualities. Based on the framework, the study suggests five stages of core and supplemental indicators for raising firms’ awareness and measuring their progress toward sustainable production systems. The six dimension: (1) energy and material use; (2) natural environment; (3) economic performance; (4) community development and social justice; (5) workers, and (6) products. Ra˘dulescu, Ra˘dulescu and Ra˘dulescu (2009) formulate and study a multi-objective approach for production processes, which implements suitable constraints on pollutant emissions with considered two alternative optimization problems. In addition, Su, Chiueh, Hung and Ma (2007) studied many modern decision-making support systems which already partially consider social factor analysis in addition to expenses and benefits, environmental effects, technical issues, and management aspects. In conclusion, the SPIs of a firm bases on multiple aspects and criteria approach with qualitative measures. Morrissey and Browne (2004) proposed that a sustainable management model should not be only environmentally effective and economically affordable but also socially acceptable. Therefore, overhauling production process to achieve firm’s goal of waste elimination and reduce the impacts of EPRs are necessary for composing a set of SPIs. Based on literature findings, activities, processes and characteristics associated with SPIs applications are adopted as SPIs dimensions. Fresner (1998) resulted on Austrian Preventive environmental protection approaches in Europe project that the process water could be saved by reusing cooling water as process water: (1) water usage could be avoided by optimizing the use of water through better process control. (2) Operational sequences have been changed to avoid waste. (3) The operators were trained to calculate the exact demand of chemicals to avoid bath rests. (4) Wasted dyeing baths are reused. De Bruijn and Hofman (2000) analyzed the contribution of pollution prevention to the transformation of industry by evaluating the results of various pollution prevention projects. As Resulted that pollution prevention has proven to be a valuable concept, because prime focus on material flows and the emphasis on minimization of environmental effects. It leads to improvements in efficiency and reductions in waste and emissions. The product orientation is undervalued in pollution prevention. Pollution prevention can be an important path of companies towards a more sustainability-based strategy. Shriberg (2002) described ‘‘. . ... The Michigan housing sustainability study was development thirty-eight recommendations to move the organization toward sustainability and assessed in terms of their importance, cost and implementation time to produce priority”. Grutter and Egler (2004) have described cleaner production is a preventative integrated continuous strategy for modifying products, processes or services, has been considered as the best technological strategy and good housekeeping toward sustainable development. Kjaerheim (2004) showed that the description of success cleaner production can give often both environmental

2188

Y. Lin et al. / Expert Systems with Applications 37 (2010) 2186–2196

and economic benefits; it promotes facility efficiency, reduces the need for expensive end-of-pipe treatment and disposal technologies, improved material, energy efficiency, quality system and reduces the long-term liabilities associated with releases into the environment. Firms wish to become more sustainable in their daily production practices should aim to address the aspects and criteria. The decision-maker should consider the interdependence between different aspects and criteria of sustainable production while the number of sustainability indicators in the literature is growing, none of them is evaluating of firm’s SPIs in uncertainty (Tseng, Wu, Lin, & Liao, 2008b). With regard to uncertainty, the natural language effectively expresses human thought and subjective preferences. Fuzzy set theory utilizes linguistic variables to accommodate natural language (Zadeh, 1965). The EPRs evaluation models permitting intuitive judgment have garnered acceptance by various experts, such as scholars and professionals’ representatives. Many other studies have employed fuzzy set theory to fuzzy problems with substantial uncertainty (Chen & Chiou, 1999; Chiou, Tzeng, & Cheng, 2005; Tseng et al., 2008a). Most evaluators cannot give exact numerical values to represent opinions, based on human perception, on MSW criteria; more realistic evaluation uses linguistic assessments rather than numerical values. After Zadeh (1965) introduced fuzzy set theory to deal with vague problems, linguistic terms have been used for approximate reasoning within the framework of fuzzy set theory to handle the ambiguity of evaluating data and the vagueness of linguistic expression (Zadeh, 1975). A linguistic information is a variable whose values (namely linguistic values) have the form of phrases or sentences in a natural language (Von Altrock, 1996). The linguistic variable is useful method in dealing with situations which are described in quantitative expressions (Asan, Erhan Bozdag, & Polat, 2004; Wang & Chuu, 2004). Especially, linguistic variables are used as variables whose values are not numbers but linguistic term. In practice, linguistic values can be represented by fuzzy numbers, and the TFN is commonly used. From interdependence relations point of view, ANP is a flexible analytical program that enables decision makers to find the best possible solution to complex problems by breaking down a problem into a systematic and hierarchical network of inter-relationships among the various levels and criteria. Mesey (2008) studies on multi-objective resource allocation of shared resources by group decision- making can combine analytic and qualitative modeling, the subsequent phases of the qualitative and the analytic solution of a multi-objective cooperative resource allocation problem can be applied within the group decision-making framework of defense requirements capability-based planning. Tseng et al. (2008a) used fuzzy ANP approach to incorporate various criteria, included organizing, systems and technologies, assessment and feedback, training and people into an interdependence framework, to align with their managerial competitive priority. Ayag˘ and Özdemir (2009) evaluated conceptual design alternatives in a new product development used ANP to accommodate the variety of interactions, dependencies and feedback between higher and lower level elements in uncertainty. This approach proposed to evaluate a set of conceptual design alternatives developed in a new product development environment in order to reach to the best one satisfying both the needs and expectations of customers, and the engineering specifications of the case company. GarcíaCascales and Lamata (2009) studied environmental impact assessment is an intrinsically complex multi-dimensional process which involves multiple criteria and this approach used the analytic hierarchy process, as a potential decision-making method for use in management maintenance processes. A hierarchical structure is built for the prequalification of the criteria and the alternatives within the system. The criteria can be prioritized and the alternatives can be organized in descending order so that the best parts

cleaning system may be selected. A more general form of AHP, instead of AHP due to the fact that AHP cannot accommodate the variety of interactions, dependencies and feedback between higher and lower level elements. The traditional approach is always assumed that the criteria are independent. In reality, within production process, the criteria are inter-related. With regard to QFD analysis approach, QFD determines product design specifications (how) based on customer needs (what) and competitive analysis (why), which represents a customer-driven and market-oriented process for decision-making. It is quite natural to use QFD in decision making for such purposes as determining customer needs (Stratton, 1989) and development priorities (Crowe & Cheng, 1996; Han, Kim, Choi, & Kim, 1998; Jugulum & Sefik, 1998). Karsak, Sozer, and Alptekin (2002) presented a QFD systematic decision procedure in product planning, which has been traditionally based on expert opinions and consider the interdependence between the customer needs and product technical requirements, and the inner dependence within themselves, along with resource limitations, and design metrics such as extendibility and manufacturability. Tsai, Lo, and Chang (2003) used fuzzy QFD approach with an optimistic index in the priority ranking procedures. This fuzzy index can correct bias problems in a consistent way for prioritizing strategic functions. A priority change display in the priority ranking according to different scenarios can provide ‘‘what-if” analysis in a decision-making environment. Bevilacqua, Ciarapicab, and Giacchettab (2006) suggested a new method that transfers the house of quality (HOQ) approach typical of QFD problems to the supplier selection process. The study starts by identifying the features that the purchased product should have (internal variables ‘‘what”) in order to satisfy the firm’s needs, then it seeks to establish the relevant supplier assessment criteria (external variables ‘‘how”) in order to come up with a final ranking based on the fuzzy suitability index. In conclude, the EPRs are multi-dimensional, complex, and interdependency activities. Several interdependency aspects and criteria should be considered and evaluated in terms of numerous aspects and criteria, resulting substantial amounts of qualitative data that are commonly inaccurate or uncertain. The EPRs are difficult to quantify and outcome is highly uncertain when data and information is inaccurate. Thus, EPRs are difficult to quantify and outcome is highly uncertain when data and information is lacking. This preliminary literature reviewed illustrates the fact that the criteria and decision-making are critically important for the success of measures. Hence, this study attempts to apply a combined method of fuzzy set theory, ANP and QFD approach by explicitly describing the decision structure of SPIs by utilizing subjective judgments of evaluators based on this decision structure. 3. Research method To determine the evaluation aspects and criteria, the measures are multiple and frequently structured into study framework, with qualitative assessment. Numerous aspects and criteria must be considered in structuring the hierarchical framework. This proposed hierarchy allows experts to identify options using linguistic expressions. To effectively solve the study problems with a hierarchical structure, this study presents a set of fuzzy numbers in a straightforward method. The triangular fuzzy membership function (Table 1) can accommodate the qualitative data while the evaluators process the evaluation in linguistic information. The following sections are presented the application method for this study. 3.1. Fuzzy set theory Fuzzy set theory expresses and handles vague or imprecise judgments mathematically. The theory indicates that each number

2189

Y. Lin et al. / Expert Systems with Applications 37 (2010) 2186–2196 Table 1 Linguistic scales/model for the importance weight of aspects for ANP. Linguistic terms Extreme importance Demonstrated importance Strong importance Moderate importance Equal importance

Linguistic values (0.75, 1.0, 1.0) (0.5, 0.75, 1.0) (0.25, 0.5, 0.75) (0, 0.25, 0.5) (0, 0, 0.25)

between 0 and 1 indicates a partial truth, whereas crisp sets correspond to binary logic [0, 1] (Al-Najjar & Alsyouf, 2003). In particular, to tackle the ambiguities involved in the process of linguistic estimation, it is a beneficial way to convert these linguistic terms into TFNs. This study builds on some important definitions and notations of fuzzy set theory from Cheng and Lin, 2002. X ¼ fx1 ; x2 ; x3 ; . . . ; xn g designates the universe of discourse. Fuzzy e of X is a set of ordered pairs fðx1 ; f ðx1 ÞÞ; ðx2 ; f ðx2 ÞÞ; . . . ; set A eA eA ðxn ; feðxn ÞÞg, where fe : X ! ½0; 1 represents the membership funcA A e The e and f ðxi Þstands for the membership degree of xi in A. tion of A, eA following definitions apply in this study: Definition 1. When X is continuous rather than a countable or e is denoted as A e ¼ R f ðx Þ=ðxÞ, where finite set, the fuzzy set A i X e A x 2 X. e is Definition 2. When X is a countable or finite set, the fuzzy set A represented as

e¼ A

X i

fe ðxi Þ=ðxi Þ; A

where xi 2 X:

e of the universe of discourse X is normal Definition 3. A fuzzy set A when its membership function fe ðxÞ satisfies max fe ðxÞ ¼ 1. A

A

Definition 4. A fuzzy number is a fuzzy subset in the universe of discourse X that is not convex but is normal.

e and B e to be two TFN parameterized by Definition 8. Assigning A the triplet ða1 ; a2; a3 Þ and ðb1 ; b2; b3 Þ respectively, the operational laws of these two TFN are as follows:

e e ¼ ða1 ; a2 ; a3 ÞðþÞðb1 ; b2 ; b3 Þ ¼ ða1 þ b1 ; a2 þ b2 ; a3 þ b3 Þ AðþÞ B e e ¼ ða1 ; a2 ; a3 ÞðÞðb1 ; b2 ; b3 Þ ¼ ða1  b1 ; a2  b2 ; a3  b3 Þ AðÞ B e e ¼ ða1 ; a2 ; a3 ÞðÞðb1 ; b2 ; b3 Þ ¼ ða1 b1 ; a2 b2 ; a3 b3 Þ AðÞ B e e ¼ ða1 ; a2 ; a3 ÞðÞðb1 ; b2 ; b3 Þ ¼ ða1 =b3 ; a2 =b2 ; a3 =b1 Þ AðÞ B However, these TFNs must be transformed into crisp numbers. Many methods can achieve this transformation (e.g., means of maxima, center of sum, center of gravity, and the a-cut method). Any fuzzy aggregation method always needs to contain a defuzzification method because the results of human judgments with fuzzy linguistic variables are TFNs. The term defuzzification refers to the selection of a specific crisp element based on the output fuzzy set, which convert fuzzy numbers into crisp may score. This study is applying the converting fuzzy data into crisp scores developed by Opricovic and Tzeng (2003), the main procedure of determining the left and right scores by fuzzy minimum and maximum, the total score is determined as a weighted average according to the ~ kij ¼ ðak1ij ; ak2ij ; ak3ij Þ, suppose that a membership functions. Let us w e kij to present the fuzzy weight decision group has k members; take w of ith criteria affects the jth criteria assessed by kth evaluators Normalization:

e a and strong a-cut A e aþ of the fuzzy Definition 5. The fuzzy a-cut A e in the universe of discourse X is defined by set A

xak1ij

e a ¼ fxi jf ðxi Þ P a; xi 2 Xg; A e

xak3ij ¼ ðak3ij  min ak1ij Þ=Dmax min

A

e aþ ¼ fxi jf ðxi Þ P a; xi 2 Xg; A e A

where a 2 ½0; 1 where a 2 ½0; 1

e of the universe of discourse X is convex Definition 6. A fuzzy set A e a is a close interval of R. e a is convex, that is, A if and only if every A This definition can be written as

e a ¼ ½PðaÞ ; PðaÞ ; A 1 2

where a 2 ½0; 1

¼ ðak1ij  min ak1ij Þ=Dmax min

xak2ij ¼ ðak2ij  min ak1ij Þ=Dmax min k k where Dmax min ¼ max a3ij  min a1ij Compute left (ls) and right (rs) normalized value:

k

xlsij ¼ xak2ij =ð1 þ xak2ij  xak1ij Þ

9 8 0; x  a1 > > > > > > > = < ðx  a1 Þ=ða2  a1 Þ; a1 6 x 6 a2 >

feðxÞ ¼ A > ða3  xÞ=ða3  a2 Þ; a2 6 x 6 a3 > > > > > > > ; : 0; x  a3

ð2Þ

xrskij ¼ xak3ij =ð1 þ xak3ij  xak2ij Þ Compute total normalized crisp value k

Definition 7. A TFN can be defined as a triplet ða1 ; a2; a3 Þ; the meme as defined, is shown in bership function of the fuzzy number A, Fig. 1.

ð1Þ

k

k

xkij ¼ ½xlsij ð1  xlsij Þ þ xrskij xrskij =½1  xlsij þ xrskij 

ð3Þ

Compute crisp values:

wkij ¼ min ak1ij þ xkij Dmax min

ð4Þ

To integrate the different opinions of evaluators, this research adopted the synthetic value notation to aggregate the subjective judgment for k evaluators, given by

~j ¼ w

1 1 ~ þw ~ 2ij þ w ~ 3ij þ    þ w ~ kij Þ ðw k ij

ð5Þ

2190

Y. Lin et al. / Expert Systems with Applications 37 (2010) 2186–2196

Part Component Characteristics

Product design Requirements

Customer Needs

Part Component Characteristics

Product design Requirements

(I) Product Planning

Manufacturing Operations

(II) Part Planning

(III) Process Planning

Production Requirements

Manufacturing Operations

(IV) Production Planning

Fig. 1. Deploying a series of QFDs.

This study uses TFNs and defuzziffied into a crisp value. In order to satisfy the decision-making criteria with the interdependence relations, the ANP and QFD are proposed to solve this particular issue.

CI ¼

kmax  n n1

3. The consistency ratio (CR) of CI to the mean random consistency index (RI) is expressed as CR (Less than 0.1)

3.2. ANP

CR ¼ Analytic hierarchy process (AHP) was first proposed by Saaty (1980). AHP must satisfy the characteristic of independence among the criteria before it can proceed to decision-making. However, given the problems encountered in reality, a dependent and feedback relationship is usually generated among the evaluation criteria and such an interdependent relationship usually becomes more complex with the change in scope and depth of the decision-making problems. Therefore, Saaty (1996) developed a new analysis method that simultaneously takes into account both the relationships of feedback and dependence. Mesey (2008) studies on multi objective resource allocation of shared resources by group decision-making can combine analytic and qualitative modeling, the subsequent phases of the qualitative and the analytic solution of a multi-objective cooperative resource allocation problem can be applied within the group decision-making framework of defense requirements capability-based planning. Tseng and Lin (2008) applied ANP to selecting of competitive priority in cleaner production implementation. The merits of ANP in group decision-making are as follows (Dyer & Forman, 1992): (i) both tangibles and intangibles, individual values, and shared values can be included in the decision process; (ii) the discussion in a group can be focused on objectives rather than on alternatives; (iii) the discussion can be structured so that every factor relevant to the decision is considered; and (iv) in a structured analysis, the discussion continues until relevant information from each individual member in the group is considered and a consensus is achieved. 1. Assuming there are n number of criteria, denoted as (C 1 , Cn), its pairwise comparison matrix would be A ¼ ðaij Þ, in which aij represents the relative significance of C i to C j . Then, by using the row vector average normalization, the approximate weight Wi of Ci is calculated as

Pn Wi ¼

j¼1 ðaij

Pn

i¼1 aij Þ

n

;

8i; j ¼ 1; 2; . . . ; n

ð6Þ

2. The consistency test of ANP is designed to ensure the consistency of judgments by decision makers throughout the decision making process. When inconsistencies exist in the pairwise comparison matrix A, Saaty (1980) proved that for consistent reciprocal matrix, the kmax value is equal to the number of comparisons, or kmax ¼ n. Then gave a measure of consistency, called Consistency Index as deviation or degree of consistency using the following formula

ð7Þ

CI RI

ð8Þ

The CR ratio should be less than 0.1; it indicates that the consistency level of the pairwise comparison matrix is acceptable. When CR is greater than 0.1, it indicates that the results of the decision process are not consistent and suggests that the decision maker performs the pairwise comparison again. ANP uses supermatrix to deal with the relationship of feedback and interdependence among the criteria. If no interdependent relationship exists among the criteria, then the pairwise comparison value would be 0. If an interdependent and feedback relationship exists among the criteria, then such value would no longer be 0 and an unweighted supermatrix M will be obtained. If the matrix does not conform to the principle of column stochastic, the decision maker can provide the weights to adjust it into a supermatrix that conforms to the principle of column stochastic, and it will become a weighted supermatrix M. 3.3. QFD The QFD is an implement to translate customer needs into product technical requirements of new products and services that have been developed from Japan in the late 1960s to early1970s (Chan & Wu, 2002). The main concept of traditional QFD considered four relationship matrices that included product planning, parts planning, process planning, and production planning matrices, respectively (Akao, 1997; Karsak et al., 2002). Each translation used a matrix, also called house of quality (HOQ), as shown in Fig. 1. In the first place, the product planning matrix is established. The customer needs translated to the second QFD as inputs for the development of product design requirements. Secondly, in the part planning matrix, important design requirements are linked to part component characteristics deployment. Furthermore, the part component characteristics are similarly linked to manufacturing operations. In the production planning matrix, the process parameters and control limits are determined in the same way. This study is to consider the construction of production planning matrix (the last of the four matrices). That is, the sustainable production indicators are defined by the QFD model. In order to establish these interdependence relations, the production natures are translated into EPRs aspects and controls of the EPRs aspects are depending on the SPIs criteria. Several of the critical notions can be expressed as follows. (1) EPRs – the first

2191

Y. Lin et al. / Expert Systems with Applications 37 (2010) 2186–2196

step is to identify the ‘‘Whats”. In sum, there are five aspects for defining the EPRs that suggested by Veleva and Ellenbecker (2001), namely, energy and material for natural environment, economic performance, community development and social justice, workers, and products. The EPRs help manufacturing process toward continuous improvement and sustainable management. Therefore, the EPRs tell the company ‘‘what to do” the aspects for sustainable production. (2) The SPIs are specified as the ‘‘Hows” of QFD. As the rule of QFD, the SPIs indicate the firm ‘‘how to do”. The twenty-two SPIs are used to establish how well the firm satisfies the EPRs. Veleva and Ellenbecker (2001) indicated that SPIs are to measure common issues for all production facilities, such as chemical releases, energy use, water use, hazardous and non-hazardous waste, work-related accident and injuries and charitable contributions. (3) The interdependence relation matrix of EPRs and SPIs are constructed. This study assumed that there is interdependence relations existed. This study employed ANP to represent the interdependence relations between ‘‘Whats” and ‘‘Hows”. It is noteworthy that the occurrences of interdependence between pairs of SPIs. In general, the conventional QFD approach lack of consideration in interdependence between SPIs. This approach supposes interdependence relations of SPIs, EPRs and HOQ. (4) The main outputs of this study obtained from preceding steps of this approach. The notations are the proposed QFD model of this study. Hence, the structure of QFD model is shown in Fig. 2.

may score, the fuzzy assessments applying in Eqs. (1)–(5) are ~ j Þ. defuzziffied and aggregated as a crisp value ðw 3. Analyze the interdependence relations from ANP pairwise comparison matrix using Eqs. (6)–(8). And compare the aspects and criteria between the degree of importance. The crisp value is composed the weighted matrix. 4. After obtaining, the crisp value of importance and interdependence levels, the priorities or relative importance of EPRs and SPIs has to calculate to evaluate the perceptions of expert evaluation. This is because that the higher the importance levels, the higher the aspects and criteria should be improved. 4. Empirical study To operationalize evaluation methodology of the EPRs aspects and criteria to the case firm, there are reasons, first, the case firm continues to improve manufacturing processes and face challenge to how they manage the environmental practices in the competitive environment. Second, case firm is willing to practice the EPRs in the industrial sector in order to deal with market competition. The expert team is formed with two professors, one vice president and seven management professions with extensive experience consulting in this study. 4.1. Case information

3.4. Proposed research procedures In sum, to combine the proposed methods in this study, the proposed research procedures are as follows: 1. Developing evaluation criteria and survey instrument – identify technical solutions – the design requirements or technical specifications of QFD involves in identifying technical solutions are expressed by literatures and experts. Those correspond to ‘‘how” and ‘‘what” in the proposed EPRs and SPIs of HOQ matrix. The direction of this ‘‘how” issue will be thought from the critical environmental practice aspects. The direction of this ‘‘what” issue will be thought from the critical SPIs criteria. The criteria have the nature of complicated relationships within the cluster of aspects and criteria. The ANP and QFD is an appropriatenesscombined tool to be applied. 2. Interpret the linguistic information into fuzzy linguistic scale using linguistic information to convert fuzzy numbers into crisp

The XYZ Group is an integrated healthcare services provider and OEM manufacturer of a wide range of medical consumable products for large multinational healthcare distributors, pharmaceutical companies and hospitals in Taiwan. Major product lines manufactured include linen, hospital apparel, hospitality apparel and bandages. The group strives to stay on top of industry trends and technological advances and take pride to quality and efficiency toward sustainable organization. XYZ group takes promoting EPRs as core competence source to survive furious green market competitions. The Chief executive officer (CEO) is thinking to understand the role of environmental practices following EPRs, especially emphasis on green production development. Therefore, the researchers present this assessment to the CEO and vice president to develop the indicators of criteria for understanding the interdependencies of SPIs criteria. The SPIs are relatively important for XYZ group to sustain in such competitive green market. The benefit of this evaluation is to acquire the purchasing order from USA and

Interdependence between SPIs

Importance of EPRs Interdependence between EPRs

SPI (HOWs) EPRs (WHATs)

Sustainable production indicators (SPIs)

Environmental production requirements (EPRs) Weights of SPIs Relationships between EPRs and SPIs

Overall Priorities of SPIs Fig. 2. QFD model for EPRs.

2192

Y. Lin et al. / Expert Systems with Applications 37 (2010) 2186–2196

European Union due to the requirement of social responsibility approach. The expert group strived to recommend the EPRs aspects and SPIs criteria, and expected it to remain long-term green product competition in intensive market. The expert group reviewed the aspects and criteria because firm evaluation is one of the most prioritized issues of the management team. It intends to evaluate the firms more logical and persuasive as there is a growing need for an analytical and systematic way of evaluation of firms in management procedures. For better handling of this study problem, the ten experts’ management group should adopt possible evaluation from the aspects and criteria. This study would provide firm recommendations, and would be useful in the efficient and effective implementation.

employee well-being and job satisfaction (C18). The product aspect (AS5) signifies a firm’s aspect to promote and sell green products on the basis of understanding customer demand. The aspect is primarily influenced by design all products can be disassembled, reused or recycled, free hazardous materials (C19), use 100% biodegradable packaging (C20), reduce amount of hazardous waste generated (C21), and increase percent of products with take-back policies (C22) (Farrell & Hart, 1998; Veleva et al., 2001; Veleva & Ellenbecker, 2001). Based on the literatures and reality status an evaluation framework presented with five primary interdependence aspects in the evaluation criteria. Table 2 presented the study structure description encountered when evaluating the EPRs measures.

4.2. Hierarchical structure of firm’s EPRs

4.3. Analytical result

This section presents the composition of a set of valid aspects and criteria. In conclusion, the EPRs of a firm are based on multiple criteria, comprising of qualitative approach. Successful EPRs depends on both economics performance and other critical aspects, such as energy and material use and natural environment, community development and social justice, workers and products. Similar to study of Tseng et al. (2008b) presented that the organizational design is being a form of proactive key factor environmental practices to economic performance. The evaluation framework consists of five aspects and twenty-two measurement criteria. The proposed model integrates and finds the relevant literature mentioned above, activities or components or characteristics that are found to be associated EPRs are put forward as aspects. Evaluating EPRs creates typical multi-criteria problems based on varying aspects and criteria. This study discusses these criteria and their associated in EPRs below. However, many aspects are fuzzy and related with others. If a firm only focuses on Energy and material for natural environment (AS1), there are only lesser resources to support Community development or social justice (AS3) and vise versa. The assessment of EPRs in the context of firm history must be collected by a literature review and expert management staffs. Particularly, to what aspects have enabled the firm to sustain in the long run. Past studies have offered valuable structure based on some indicators. If one organization spends a vast investment on Energy and material for natural environment (AS1) and takes a hard endeavor on production approach, the SPIs are reduce the use of fresh water (C1), reduce material use (C2), reduce energy use (C3), and increase the use of energy from renewable sources (C4). It presents the daily operation process criteria. On the other hand, if an aspect of EPRs without the supporting of economic performance and products, then the exertion of operation process will be restricted. The criteria of economic performance (AS2) are reduce the amount of waste generated before recycling (air, water and land) (C5), reduce environment health and safety compliance costs (C6), zero customer complaints or returns (C7), and percent age suppliers participating in raw material or packaging Life Cycle Assessment (C8). And somehow the firm must interact with community development or social justice (AS3) such as reduce greenhouse gas emissions (C9), reduce of waste generated by contracted service/ material provides (C10), percentage of suppliers from the local area (C11), percentage of products consumed locally (C12), and increase employment opportunities for local community (C13). In view of worker aspects denotes a firm’s ability to execute safety environment for workers (AS4). This aspect comprises of 5 criteria, those are achieve zero lost workdays as result of work-related injuries and illness (C14), increase the rate of employee suggested improvement in quality, social and environment health and safety performance (C15), reduce employee turnover rate (C16), increase employee training on green knowledge (C17), and increase

This empirical study presents to illustrate the application of the solution for evaluating a favorable criterion in SPIs. This study attempts to apply the ANP and QFD to the evaluation of five aspects and twenty-two criteria in uncertainty. The expert group followed the application solution with the four-phase procedures. 1. Developing ‘‘Whats” and ‘‘Hows” for measurement instrumentit is important to establish a set of EPRs ‘‘Whats” aspects and SPIs ‘‘Hows” criteria for evaluation. The criteria have the nature of complicated relationships within the cluster of aspects and criteria. For instance, a firm that has outstanding energy and material for natural environment (AS1) to performance in controlling the waste elimination and hazardous for natural environment and Reduce amount of hazardous waste generated (C10), which somehow related to Design all products can be disassembled, reused or recycled, free hazardous materials (C19). Therefore, the ANP and QFD is an appropriateness-combined tool to be applied. 2. Using linguistic information to convert fuzzy numbers into crisp may score, the fuzzy assessments applying in Eqs. ~ j Þ. (1)–(5) are defuzziffied and aggregated as a crisp value ðw For instance, the first column versus second row (AS1 vs AS2), k k k where Dmax min ¼ 1, the normalization xa1ij ¼ 0; xa2ij ¼ 0; xa3ij ¼ ð0:25  0Þ=1 ¼ 0:25, using Eq. (1). Using Eq. (2) to compute left and right normalized value, the computation result are as folk lows xlsij ¼ 0=ð1 þ 0  0Þ ¼ 0; xrskij ¼ 0:25=ð1 þ 0:25  0Þ ¼ 0:2. To compute the total normalized crisp value using Eq. (3) The crisp xkij ¼ 0xð1  0Þ þ ð0:2x0:2Þ=ð1  0 þ 0:2Þ ¼ 0:033. value uses Eq. (4) is wkij ¼ 0 þ 0:033  1 ¼ 0:033. Using Eq. (5) ~ ij ¼ 0:033, the result to average ten experts’ evaluation w showed in Table 3. 3. Analyze the interdependence relations from ANP pairwise comparison matrix using Eqs. (6)–(8). Table 4 presents the crisp values in a pairwise comparison matrix and decomposes the matrix with MATLAB 6.5 to acquire the eigenvector, the eigenvector are 0.252, 0.467, 0.466, 0.532, 0.470. Moreover, the eigenvector normalized into local priority (0.115, 0.214, 0.212, 0.243, 0.215) for composing the unweighted supermatrix. In testing for the consistency of judgment matrix, the CI is 0.099, applied Eq. (6). If the CI ratio is greater than 10%, we need to revise the subjective judgment. The CR is resulted as 0.089, which is also lower than 0.1. 4. The crisp value of importance and interdependence levels, the priorities or relative importance of EPRs and SPIs has to calculate to evaluate the perceptions of expert evaluation. Table 5 presented the interdependence of EPRs and Table 6 shown the interdependence of SPIs, those are the results from ANP. The HOQ is presented the relationships between EPRs and SPIs, The detailed construction of HOQ in Table 7 is described as follows.

2193

Y. Lin et al. / Expert Systems with Applications 37 (2010) 2186–2196 Table 2 Hierarchical structural of SPI of firm’s evaluation. Goal

EPRs (aspects)

SPIs (criteria)

Evaluation of EPRs

Energy and material for natural environment(AS1)

Economic and environmental performance (AS2)

Community development or social justice (AS3)

Workers (AS4)

C1 C2 C3 C4 C5 C6 C7 C8 C9 C10 C11 C12 C13 C14 C15 C16 C17 C18 C19 C20 C21 C22

Products (AS5)

Reduce the use of fresh water Reduce material use Reduce energy use Increase the use of energy from renewable sources Reduce of waste generated by contracted service/material provides Reduce environment health and safety compliance costs Zero customer complaints or returns Percent suppliers participating in raw material or packaging life cycle assessment Reduce greenhouse gas emissions Reduce amount of hazardous waste generated Percent of suppliers from the local area Percent of products consumed locally Increase employment opportunities for local community Achieve zero lost workdays as result of work-related injuries and illness Increase the rate of employee suggested improvement in quality, social and environment health and safety performance Reduce employee turnover rate Increase employee training on green knowledge Increase employee well-being and job satisfaction Design all products can be disassembled, reused or recycled, free hazardous materials Use 100% biodegradable packaging Reduce the amount of waste generated before recycling (air, water and land) Increase percent of products with take-back policies

Table 3 Pair comparison in TFNs interpreted from linguistic information (respondent no. 1). Goal

AS1

AS1 AS2 AS3 AS4 AS5

½ 0:00 ½ 0:00 ½ 0:00 ½ 0:00 ½ 0:25

0:00 0:00  0:00 0:25  0:00 0:25  0:00 0:25  0:50 0:75 

AS 2

AS 3

½ 0:00 0:00 0:25  ½ 0:00 0:00 0:00  ½ 0:00 0:25 0:50  ½ 0:25 0:50 0:75  ½ 0:00 0:25 0:50 

½ 0:00 ½ 0:00 ½ 0:00 ½ 0:25 ½ 0:00

0:00 0:25  0:25 0:50  0:00 0:00  0:50 0:75  0:25 0:50 

AS 4

AS 5

½ 0:00 0:00 0:25  ½ 0:25 0:50 0:75  ½ 0:25 0:50 0:75  ½ 0:00 0:00 0:00  ½ 0:00 0:25 0:50 

½ 0:25 ½ 0:00 ½ 0:00 ½ 0:00 ½ 0:00

0:50 0:75  0:25 0:50  0:25 0:50  0:25 0:50  0:00 0:00 

5. Discussions

Table 4 Defuzzification into crisp value. AS 1

AS1

AS2

AS3

AS4

AS5

E-vector

Weights

AS1 AS2 AS3 AS4 AS5

0.000 0.033 0.033 0.033 0.540

0.033 0.000 0.293 0.540 0.293

0.038 0.291 0.000 0.525 0.291

0.038 0.525 0.525 0.000 0.291

0.540 0.293 0.293 0.293 0.000

0.252 0.467 0.466 0.532 0.470

0.115 0.214 0.212 0.243 0.215

kmax ¼ 7:304; CI: 0.099; CR: 0.089.

Table 8 showed the completed result of QFD ranking on SPIs. This is because that the higher the importance levels, the higher the aspects and criteria could be improved in the OEM case firm. The top five ranking SPIs are as follows: (1) reduce of waste generated by contracted service/material provides (C5); (2) reduce material use (C2); (3) reduce energy use (C3); (4) reduce employee turnover rate (C16); and (5) percentage of supplier participated in raw material or packaging life cycle assessment (C8). The ranking sequences are C5 > C2 > C3 > C16 > C8 > C1 > C13 > C6 > C10 > C14 > C21 > C12 > C19 > C4 > C7 > C15 > C17 > C11 > C20 > C18 > C9 > C22.

Table 5 Interdependence of EPRs. EPRs

AS 1

AS 2

AS 3

AS 4

AS 5

AS1 AS 2 AS 3 AS 4 AS 5

0.115 0.214 0.212 0.243 0.215

0.262 0.157 0.168 0.196 0.217

0.256 0.076 0.246 0.165 0.257

0.171 0.264 0.123 0.294 0.148

0.345 0.146 0.256 0.112 0.141

In summary, the empirical results show that the OEM case firm should listen attentively the voice of EPRs on improving the functional activities of SPIs, especially on reduce of waste generated by contracted service/material provides; reduce material use; reduce energy use; reduce employee turnover rate; and percentage of supplier participated in raw material or packaging life cycle assessment. It is suggested that more attentions should be paid to exploit these SPIs effectively and then develop the ‘how’ issues of profiles of solutions. This study suggests case firm should continuously strengthen the perspectives of EPRs and internal manufacturing process respectively. Following discussions describe how these key solutions, are influencing for case firm. Reduce of waste generated by contracted service/material provides – the material suppliers are critical for the OEM firms. The perfect material design/ supply plans of supplier relationship to create a long-term plan are the main target for reducing waste generated. This is because enhancing trustiness of supplier relationship can collaborate on the product designs. The retention and reused desire or repeated purchases for OEM firm have positive impacts on their new product development plan. Reduce material use – a well-designed process-based production system can reduce the material waste degree for OEM firms, and finally can increase the total efficiency and effectiveness of material usage. Reduce energy use – OEM firm devotes to EPRs as the main spindle for environment practices. All the production plans, activities, and strategies made from the EPRs aspects can achieve the expectations of external requirements, and further to understand the trend of green market requirements.

2194

Table 6 Interdependence of SPIs. C1

C2

C3

C4

C5

C6

C7

C8

C9

C10

C11

C12

C13

C14

C15

C16

C17

C18

C19

C20

C21

C22

C1 C2 C3 C4 C5 C6 C7 C8 C9 C10 C11 C12 C13 C14 C15 C16 C17 C18 C19 C20 C21 C22

1.000 0.346 0.467 0.313 0.316 0.404 0.405 0.378 0.346 0.317 0.440 0.346 0.556 0.407 0.376 0.285 0.285 0.313 0.375 0.255 0.345 0.311

0.378 1.000 0.375 0.347 0.498 0.407 0.376 0.407 0.376 0.468 0.347 0.374 0.407 0.433 0.434 0.345 0.314 0.314 0.314 0.318 0.375 0.374

0.467 0.405 1.000 0.467 0.407 0.344 0.282 0.375 0.405 0.282 0.311 0.347 0.376 0.525 0.436 0.373 0.345 0.222 0.314 0.376 0.405 0.313

0.283 0.284 0.348 1.000 0.404 0.406 0.408 0.346 0.436 0.283 0.403 0.312 0.346 0.221 0.317 0.312 0.317 0.314 0.314 0.317 0.407 0.403

0.345 0.379 0.439 0.467 1.000 0.405 0.407 0.409 0.525 0.407 0.438 0.345 0.407 0.407 0.496 0.374 0.379 0.254 0.374 0.410 0.379 0.345

0.311 0.283 0.347 0.348 0.409 1.000 0.494 0.407 0.374 0.339 0.438 0.378 0.314 0.313 0.436 0.285 0.316 0.345 0.438 0.407 0.283 0.285

0.221 0.344 0.253 0.465 0.403 0.374 1.000 0.374 0.402 0.250 0.402 0.314 0.436 0.376 0.376 0.405 0.282 0.283 0.309 0.347 0.374 0.220

0.315 0.404 0.314 0.344 0.407 0.584 0.344 1.000 0.339 0.347 0.345 0.402 0.345 0.342 0.254 0.378 0.465 0.253 0.374 0.407 0.407 0.282

0.373 0.376 0.343 0.343 0.494 0.190 0.251 0.252 1.000 0.311 0.314 0.283 0.345 0.463 0.405 0.254 0.338 0.313 0.313 0.252 0.370 0.309

0.346 0.379 0.317 0.465 0.407 0.282 0.405 0.465 0.251 1.000 0.376 0.405 0.377 0.283 0.314 0.405 0.377 0.285 0.254 0.496 0.223 0.316

0.348 0.435 0.406 0.252 0.374 0.432 0.496 0.256 0.283 0.284 1.000 0.403 0.437 0.346 0.466 0.317 0.191 0.254 0.436 0.286 0.316 0.223

0.314 0.345 0.252 0.248 0.342 0.284 0.376 0.409 0.465 0.339 0.369 1.000 0.376 0.376 0.314 0.347 0.465 0.404 0.376 0.376 0.433 0.340

0.405 0.434 0.287 0.318 0.347 0.374 0.318 0.438 0.439 0.318 0.407 0.347 1.000 0.463 0.403 0.403 0.407 0.316 0.405 0.348 0.254 0.308

0.407 0.438 0.376 0.405 0.408 0.374 0.285 0.254 0.347 0.253 0.342 0.379 0.467 1.000 0.434 0.405 0.313 0.347 0.376 0.254 0.316 0.278

0.407 0.316 0.432 0.252 0.436 0.256 0.376 0.407 0.406 0.372 0.222 0.317 0.345 0.376 1.000 0.467 0.403 0.254 0.374 0.406 0.317 0.223

0.407 0.438 0.407 0.465 0.433 0.373 0.253 0.436 0.316 0.374 0.347 0.407 0.345 0.373 0.465 1.000 0.405 0.284 0.313 0.316 0.284 0.314

0.286 0.376 0.437 0.346 0.407 0.407 0.316 0.408 0.314 0.286 0.253 0.316 0.314 0.373 0.316 0.347 1.000 0.343 0.283 0.377 0.283 0.314

0.254 0.407 0.348 0.347 0.314 0.343 0.223 0.379 0.222 0.345 0.345 0.345 0.287 0.282 0.254 0.348 0.342 1.000 0.407 0.467 0.374 0.316

0.374 0.434 0.285 0.314 0.374 0.434 0.343 0.467 0.314 0.249 0.344 0.314 0.285 0.374 0.251 0.311 0.282 0.373 1.000 0.347 0.433 0.316

0.316 0.285 0.378 0.316 0.378 0.430 0.347 0.405 0.313 0.465 0.376 0.345 0.376 0.316 0.285 0.405 0.283 0.405 0.345 1.000 0.253 0.314

0.345 0.401 0.345 0.316 0.344 0.310 0.314 0.436 0.341 0.403 0.315 0.371 0.345 0.403 0.222 0.376 0.434 0.521 0.405 0.345 1.000 0.372

0.185 0.186 0.283 0.431 0.401 0.391 0.312 0.432 0.216 0.282 0.250 0.247 0.215 0.215 0.250 0.280 0.185 0.220 0.216 0.251 0.342 1.000

Table 7 Relationships between EPRs and SPIs. HOQ

C1

C2

C3

C4

C5

C6

C7

C8

C9

C10

C11

C12

C13

C14

C15

C16

C17

C18

C19

C20

C21

C22

AS1 AS2 AS3 AS4 AS5

0.046 0.033 0.028 0.059 0.061

0.054 0.036 0.045 0.069 0.026

0.046 0.059 0.050 0.020 0.036

0.021 0.038 0.044 0.019 0.044

0.033 0.041 0.045 0.026 0.033

0.048 0.026 0.053 0.086 0.035

0.053 0.045 0.049 0.021 0.031

0.034 0.055 0.020 0.034 0.068

0.031 0.076 0.047 0.094 0.016

0.050 0.026 0.065 0.043 0.031

0.086 0.065 0.054 0.031 0.120

0.059 0.030 0.076 0.095 0.042

0.086 0.078 0.059 0.045 0.054

0.022 0.059 0.032 0.027 0.071

0.044 0.033 0.095 0.032 0.036

0.037 0.036 0.019 0.043 0.015

0.036 0.044 0.029 0.029 0.035

0.031 0.031 0.012 0.035 0.052

0.019 0.024 0.012 0.029 0.029

0.064 0.087 0.086 0.121 0.059

0.070 0.036 0.041 0.018 0.060

0.033 0.044 0.040 0.028 0.048

Y. Lin et al. / Expert Systems with Applications 37 (2010) 2186–2196

SPIs

1.499 3.56 22 1.935 4.60 11 2.018 4.80 19 1.774 4.22 20 1.974 4.69 6

2.042 4.85 2

1.990 4.73 3

1.829 4.35 14

2.126 5.05 1

1.971 4.69 8

1.866 4.44 15

1.972 4.69 18

1.997 4.75 12

2.050 4.87 7

1.902 4.52 10

1.894 4.50 16

1.931 4.59 4 1.938 4.61 9 1.799 4.28 21 1.967 4.68 5

1.798 4.27 17

1.790 4.25 13

0.347 0.255 0.303 0.300 0.294

Sum. Importance (%) Rank

C21

0.445 0.332 0.391 0.389 0.378 0.462 0.348 0.402 0.411 0.394

C20 C19

0.412 0.309 0.359 0.362 0.348 0.409 0.306 0.356 0.358 0.345 0.413 0.308 0.361 0.364 0.352 0.455 0.336 0.400 0.395 0.386

0.456 0.345 0.400 0.407 0.390

0.470 0.352 0.412 0.415 0.401

0.441 0.325 0.383 0.382 0.372

0.435 0.322 0.381 0.383 0.374

0.442 0.332 0.387 0.392 0.378 0.444 0.331 0.390 0.392 0.380 0.413 0.309 0.359 0.366 0.352 0.453 0.339 0.394 0.398 0.383 0.429 0.318 0.376 0.376 0.367 0.451 0.339 0.395 0.401 0.385 0.489 0.362 0.428 0.429 0.418 0.423 0.311 0.369 0.367 0.359 0.459 0.339 0.401 0.401 0.391 0.466 0.351 0.410 0.415 0.400

C3 C2

Reduce employee turnover rate – this solution involved the staff, in each department of OEM firm, can learn and train the environment requirement information and skills to strengthen the capabilities for business operations. Enhancing the EPRs degree of professional activities may down the operational cost and to contribute to economic performance. Percentage of supplier participated in raw material or packaging life cycle assessment – an efficiency supplier participated in raw material or packaging life cycle assessment services are a main purpose for EPRs. Especially, the viewpoint and concept of participation has changed for OEM firms nowadays. With the issue of EPRs rises and gradually attaches importance to potential customers. The supplier must provide the needs of quick response (QR) and efficiency supply plans to satisfy the diverse demands for OEM firms.

6. Concluding remarks

0.454 0.339 0.398 0.398 0.385

AS5

C18 C17 C16 C15 C14

AS4

C13 C12 C11 C10 C9

AS3

C8 C7 C5

C6 AS2

C1

C4 AS1 QFD

Table 8 Ranking result of QFD.

2195

In addition, this study has several implications for the case firm who intend to evaluate OEM firms in term of EPRs. The main contribution of the paper is the hierarchical model presented in Table 2. This model provides a useful guideline as a structured and logical means of synthesizing judgments for evaluating appropriate SPIs for an OEM firm. It helps structure a difficult and often emotion-burdened decision. The second implication is the aspects of the EPRs listed in the proposed model. Based on a comprehensive review, the features of EPRs and SPIs have been examined and identified. These give an overview structure for the case firm without much knowledge of EPRs. Such firms can better understand the evaluation criteria in term of the aspects and criteria in the EPRs. The fuzzy QFD is particularly useful for decision-making in a multi-criteria interdependence context. Moreover, the EPRs framework may also advantage other firm’s management can take the framework and customize it for use in their own environmental management activities. In this manner, evaluators need to take the EPRs and delete their relevant criteria from it and to add what is missing. Consequently, the EPRs can be used in different aspects/criteria and can be further modified and refined if required.

AS1 AS2 AS3 AS4 AS5

C22

Y. Lin et al. / Expert Systems with Applications 37 (2010) 2186–2196

This paper aims to apply fuzzy QFD to identify the critical criteria of SPIs for OEM firm from the viewpoints of EPRs. To facilitate the main issue, the QFD approach is discussed firstly. Subsequently, in the QFD process, this study must emphasize on ‘‘Whats” EPRs and ‘‘Hows” the SPIs have to be made. For solving the environmental practices problem of ‘‘Whats” and ‘‘Hows”, there are two components and form the cornerstone of this study, the HOQ provides an insightful procedure. However, there are situations in which information is incomplete or imprecise or views that are subjective or endowed with linguistic characteristics creating a ‘‘fuzzy” decision-making environment. And yet, the interdependence relationships from the ‘‘Whats” and ‘‘Hows” should be attend in the particular study. Hence, the fuzzy QFD and ANP approach is used to solve the relationship between ‘‘Whats” and ‘‘Hows” questions. And finally, the systematic procedures using fuzzy QFD were proposed. An empirical analysis is conducted, using XYZ group as a case study, to demonstrate the systematic evaluation process for identifying the weighted SPIs. The empirical results show that the case firm should listen attentively to improve the EPRs. It is suggested that more attentions should be paid to exploit these EPRs effectively and then develop the ‘‘Hows” issues of SPIs. Furthermore, this proposed method can easily and effectively accommodates aspects and criteria that are interdependence. The EPRs evaluation problem addressed by the proposed method is appropriately interdependence. This proposed model with a

2196

Y. Lin et al. / Expert Systems with Applications 37 (2010) 2186–2196

hierarchical EPRs structure and fuzzy measure establishes a foundation for future research and is appropriate for predicting uncertain EPRs aspects. Furthermore, the management can apply this method to evaluate and determine a firm’s EPRs to improve the firm’s environmental performance and hereby bring the information for manager that will have the great effect in reducing overall sustainable production uncertainty. References Akao, Y. (1997). QFD: Past, present, and future. In Proceedings of the international symposium on QFD’97, Linko¨ ping. Al-Najjar, B., & Alsyouf, I. (2003). Selecting the most efficient maintenance approach using fuzzy multiple criteria decision making. International Journal of Production Economics, 84(1), 85–100. Asan, U., Erhan Bozdag, C., & Polat, S. (2004). A fuzzy approach to qualitative cross impact analysis. Omega – International Journal of Management Science, 32(6), 443–458. Ayag˘, Z., & Özdemir, R. G. (2009). A hybrid approach to concept selection through fuzzy analytic network process. Computers and Industrial Engineering, 56(1), 368–379. Balzarova, M. A., & Castka, P. (2008). Underlying mechanisms in the maintenance of ISO 14001 environmental management system. Journal of Cleaner Production, 16(18), 1949–1957. Bevilacqua, M., Ciarapicab, F. E., & Giacchettab, G. (2006). A fuzzy-QFD approach to supplier selection. Journal of Purchasing & Supply Management, 12, 14–27. Chan, L. K., & Wu, M. L. (2002). Quality function deployment: A literature review. European Journal of Operational Research, 143, 463–497. Chen, L. H., & Chiou, T. W. (1999). A fuzzy credit-rating approach for commercial loans: A Taiwan case. Omega – International Journal of Management Science, 27(4), 407–419. Cheng, C. H., & Lin, Y. (2002). Evaluating the best main battle tank using fuzzy decision theory with linguistic criteria evaluation. European Journal of Operational Research, 142(1), 174–186. Chiou, H. K., Tzeng, G. H., & Cheng, D. C. (2005). Evaluating sustainable fishing development strategies using Fuzzy MCDM approach. Omega – International Journal of Management Science, 33(3), 223–234. Crowe, T. J., & Cheng, C. C. (1996). Using quality function deployment in manufacturing strategic planning. International Journal of Operations and Production Management, 16(4), 35–48. De Bruijn, T. J. N. M., & Hofman, P. S. (2000). Pollution prevention and industrial transformation evoking structural changes within companies. Journal of Cleaner Production, 8, 215–223. Dyer, R. F., & Forman, E. H. (1992). Group decision support with the analytic hierarchy process. Decision Support Systems, 8(2), 99–124. Farrell, A., & Hart, M. (1998). What does sustainability really mean? Environment, 40(9), 5–31. Fresner, J. (1998). Starting continuous improvement with a cleaner production assessment in an Austrian textile mill. Journal of Cleaner Production, 6, 85–91. Garcı´a-Cascales, S. M., & Lamata, M. T. (2009). Selection of a cleaning system for engine maintenance based on the analytic hierarchy process. Computers and Industrial Engineering, 56(4), 1442–1451. Grutter, J. M., & Egler, H. P. (2004). From cleaner production to sustainable industrial production modes. Journal of Cleaner Production, 12(3), 249–256. Han, C. H., Kim, J. K., Choi, S. H., & Kim, S. H. (1998). Determination of information system development priority using quality function deployment. Computers and Industrial Engineering, 35(1–2), 241–244. Jugulum, R., & Sefik, M. (1998). Building a robust manufacturing strategy. Computers and Industrial Engineering, 35(1–2), 225–228.

Karsak, E. E., Sozer, S., & Alptekin, S. E. (2002). Product planning in quality function deployment using combined analytic network process and goal programming approach. Computers & Industrial Engineering, 44, 171–190. Kjaerheim, G. (2004). Cleaner production and sustainability. Journal of Cleaner Production, 13, 329–339. Mesey, G. (2008). A practical prioritization by multi-level group decision Support. Central European Journal of Operations Research, 16, 1–15. Morrissey, A. J., & Browne, J. (2004). Waste management models and their application to sustainable waste management. Waste Management, 24, 297–308. Nishitani, K. (2009). An empirical study of the initial adoption of ISO 14001 in Japanese manufacturing firms. Ecological Economics, 68(3), 669–679. Opricovic, S., & Tzeng, G. H. (2003). Defuzzification within a multi-criteria decision model. International Journal of Uncertainty. Fuzziness and Knowledge-Based Systems, 11(5), 635–652. Ra˘dulescu, M., Ra˘dulescu, S., & Ra˘dulescu, C. Z. (2009). Sustainable production technologies which take into account environmental constraints. European Journal of Operational Research, 193(3), 730–740. Saaty, T. L. (1980). The analytic hierarchy process. New York: McGraw-Hill Book Co. Saaty, T. L. (1996). The analytic network process-decision making with dependence and eedback. Pittsburgh, PA: RWS Publications. Shriberg, M. (2002). Toward sustainable management: The university of Michigan housing division’s approach. Journal of Cleaner Production, 10, 41–45. Stratton, B. (1989). The refined focus of automotive quality. Quality Progress, 22(10), 47–50. Su, J. P., Chiueh, P. T., Hung, M. L., & Ma, H. W. (2007). Analyzing policy impact potential for municipal solid waste management decision making: A case study of Taiwan. Resources Conservation and Recycling, 51, 418–434. Tsai, C. Y., Lo, C. C., & Chang, A. C. (2003). Using fuzzy QFD to enhance manufacturing strategic planning. Journal of the Chinese Institute of Industrial Engineering, 18(3), 33–41. Tsai, W. H., & Chou, W. C. (2009). Selecting management systems for sustainable development in SMEs: A novel hybrid model based on DEMATEL, ANP, and ZOGP. Expert Systems with Applications, 36(2–1), 1444–1458. Tseng, M. L. (2009). Application of ANP and DEMATEL to evaluate the decisionmaking of municipal solid waste management in Metro Manila. Environmental Monitoring and Assessment, 156(1–4), 181–197. Tseng, M. L., & Lin, Y. H. (2008). Selection of competitive advantage in TQM implementation using fuzzy AHP and sensitivity analysis. Asia Pacific Management Review, 13(3), 583–599. Tseng, M. L., Lin, Y. H., Chiu, A. S. F., & Liao, C. H. (2008a). Using FANP approach on selection of competitive priorities based on cleaner production implementation: A case study in PCB manufacturer, Taiwan. Clean Technology and Environmental Policy, 10(1), 17–29. Tseng, M. L., Wu, W. W., Lin, Y. H., & Liao, C. H. (2008b). An exploration of relationships between environmental practice and manufacturing performance using the PLS path modeling. WSEAS Transactions on Environment and Development, 4(6), 487–502. Veleva, V., & Ellenbecker, M. (2001). Indicators of sustainable production: Framework and methodology. Journal of cleaner production, 9, 519– 549. Veleva, V., Hart, M., Greiner, T., & Crumbley, C. (2001). Indicators of sustainable production. Journal of Cleaner Production, 9, 447–452. Von Altrock, C. (1996). Fuzzy logic and neurofuzzy applications in business and finance. New Jersey: Prentice-Hall. Wang, R. C., & Chuu, S. J. (2004). Group decision-making using a fuzzy linguistic approach for evaluating the flexibility in a manufacturing system. European Journal of Operational Research, 154, 563–572. Zadeh, L. A. (1965). Fuzzy sets. Information and Control, 8(3), 338–353. Zadeh, L. A. (1975). The concept of a linguistic variable and its application to approximate reasoning. Information Sciences, 9, 43–80.