Department of Mechanical Engineering 1. Hindustan College of Science & Technology (HCST), Mathura, UP, India, INDIA

Transcription

1 Enhancing the leanness of supply chain by integrated Fuzzy-QFD approach Rajiv K Upadhyay 1, Ajay Bangar 2, Pawan Kumar Singh 3, Ashish Shastri 4 1,2,3,4 Department of Mechanical Engineering 1 Hindustan College of Science & Technology (HCST), Mathura, UP, India, INDIA drrajeevkrupadhyay@gmail.com, abangar @gmail.com pawanme48@gmail.com, ashishshastri2000@yahoo.co.in Abstract Lean attributes and lean enablers, i.e. the capabilities which allow promptly responding to reduce the waste in the business environment and the available leverages to achieve the leanness. Few studies, however, provide empirical evidence about the main characteristics of lean enterprises and tools practically exploited by companies to achieve leanness. This study attempts to improve the existing knowledge on leanness. On the other hand, many researchers have focused on the supply chain in recent years because of the high importance of this issue in the supply chain management context. By linking Lean Attributes (LAs) and Lean Enablers (LEs), this study used Quality Function Deployment (QFD) to identify viable LEs to be practically implemented in order to increase the leanness of the supply chain. FQFD (First QFD)) are used to rank lean attributes Furthermore, fuzzy logic is used to deal with linguistic judgments expressing relationships and correlations required in QFD. To illustrate the practical implications of the methodology, the approach is exemplified with the help of a case study in the some processing industry. Keyword:-supply chain, Lean supply chain, Quality Function Deployment (QFD), House Of Quality (HOQ), Fuzzy logic. Introduction The need for holistic modeling efforts that capture the extended supply chain at a strategic level has been clearly recognized first by industry and recently by academia. In addition, in today s world of global markets and stiff competition for every product along with increasing consumer demand, it becomes imperative for companies to explore ways to improve their productivity in terms of implementing flexible and standardized technology and adopting proven management principles. Lean production is one of the initiatives that many major businesses all around the world have been trying to adopt in order to remain competitive in the increasingly global market. Lean production promises significant benefits in terms of increased organizational and supply chain communication and integration. The core thrust of lean production is the capability of working synergistically to create a high-quality system that produces finished products at the pace of customer demand with little or no waste. A supply chain consists of all parties involved, directly or indirectly, in fulfilling a customer request. The supply chain includes not only the manufacturer and suppliers, but also transporter, ware house, retailer, and even customers themselves. Leanness as a concept has been firstly stated by John Krafcif as a term for the new production system applied by Toyota. Herron and Hicks argue that the reason behind Toyota implementation for such concept has been the fact that it couldn t afford the huge capital- based mass production systems applied by the US companies. As a result, it has searched for means to reduce waste in all its operational activities, and hence, Lean production has been born. Quality Function Deployment (QFD) is an effective tool for planning attributes of new products based on customer demands and involves all members of the producer or supplier organization. QFD can be used to integrate an organization s diverse sources of information during product and process development, so that the goal of Total Quality Management (TQM) and Concurrent Engineering (CE) inside the organization can be facilitated. The aim of this study is to develop an integrated approach to increase the leanness of the supply chain. Furthermore, a practical tool is introduced that can be easily adopted to implement lean strategies. By linking Lean Attributes

2 (LAs) and Lean Enablers (LEs), this study, which is based on the House of Quality (HOQ) of Quality Function Deployment (QFD) methodology, aims to identify the most appropriate LEs for implementation by supply chain management. LAs represent company requirements and appear as what s in the HOQ, while LEs are listed as how s, since they are considered as practical tools that the company can use to achieve leanness. Moreover, in the proposed approach, fuzzy logic is used to deal with linguistic judgments expressing the relative importance of LAs as well as the relationships and correlations required in the HOQ. Contribution In Study A first outcome of the literature analysis is that none of the approach proposed in the literature ground on the QFD methodology. This study, and specially the HOQ represents a particular tool, which allows directly assessing the impact of lean attributes on competitive bases and of lean enablers on lean attributes, through the relationship matrixes. Clearly in practical cases, it would also be possible that a company directly identifies a set of suitable lean enablers to be implemented, without linking them with lean attributes and competitive priorities. However, in the case, the risk is that the selected strategic leverages do not marketing objectives. Moreover, QFD allows identifying correlations between attributes or enablers, which are not examined in the methodologies available in the literature. As a further contributions, existing methodologies ground on crisp assessments, and in this regards, we believe that fuzzy approach could provide useful advantages. A main reason why our approach is based on fuzzy set theory is that, as a result of the literature analysis, it emerged that leanness assessment is often dealt with through fuzzy logic, due to the imprecise and vague definition of leanness indicators consequently, linguistic expressions are often used to estimate both companies performance against lean attributes and enablers, and the relative importance of such parameters. When vague or ill-defined issues should be examined, the adoption of fuzzy logic is recognized to substantially improve the capabilities of traditional crisp approaches. As a matter of fact, the main strengths of fuzzy logic have to be found in the opportunity to take into account the different meanings that decision makers (DMs) may give to the same linguistic expression, as well as to translate linguistics judgments into numerical values. Methodology The framework for achieving a lean supply chain by Fuzzy- QFD comprises four main parts. It has a stepwise description as shown in Fig.1 and below. The fuzzy HOQ whose specific structure is detailed in Fig. 2 is adopted here. Start Identify LAs and LEs of the supply chain Prioritize LAs by FQFD to obtain LAs priority weights (Wi) Determine the relationships between LAs and LEs and the correlation between LEs Calculate the relative importance (RIj) and priority weights of LEs (RIj) Conclusions Fig.1.Schematic representation of the methodology.

3 Correlation matrix Lean attributes Lean attributes priority weight by FQFD Lean enablers (LEs) Relationship matrix(rij) Relative importance of LE (RIj) Priority weights of LEs(RIj * ) Normalized Rij (NRIj) Crisp Value (ranking of LEs) Fig.2. The proposed integrated Fuzzy-QFD approach A. Identify LAS and LES of the supply chain To be truly lean, a supply chain must possess a number of distinguishing attributes and enablers. LAs here after are defined as elements which constitute the underlying structure of a lean organization. They were originally conceived as core concepts of lean manufacturing. Accordingly, LEs are enabling tools, technologies, and methods critical to successfully accomplish lean supply chain management. LAs enhancing supply chain leanness and LEs to be exploited in order to achieve the required LAs, as accepted by several authors [6], were identified. On the basis of a review of the normative literature some LAs and LEs were defined for the lean supply chain, as shown in Table 1. Furthermore, suggestions to identify viable sets of lean attributes and enablers can be found in literature, and different or additional LAs/LEs may be listed. Table 1 Lean attributes and enablers defined for lean supply chain from related to food industry. Service level improvement Conformance quality Pull production Delivery reliability Eliminate obvious wastes Low buffering cost Cost efficiency Low variability in process time JIT manufacturing Low variability in delivery time Continuous improvement Low variability in demand rates Human resource training Delivery speed Quality improvement Low variability in process time Vendor management inventory

4 Total quality management Supplier management B. Prioritize LAs by FQFD to Obtain LAs Priority Weights (Wi) Due to its wide applicability and ease of use, the FQFD, has been studied extensively for the last twenty years. It has been widely used to address multi-criterion decision- making problems. The FQFD consists of three main operations: Identifiy criteria(competetive base CBi) and Subcriteria Lean Attributes LAs), Find Out relationship and correlationship between CBi and LAs, and Find Out relative importance and crisp value by applying triangular fuzzy logic. Moreover, QFD is one of the five tools that are commonly combined with the Fuzzy Logic. In this study, the FQFD was deployed to prioritize LAs. After defining LAs, their priority weights were computed by using FQFD for this purpose, first, the pair-wise assessment matrices were prepared to evaluate the eight alternatives, i.e. LAs with respect to criteria; the criteria were then evaluated with respect to the goal. Effective management of the supply chain is viewed as the driver of decreasing the cost of material, services, and manufacturing, reducing lead times and improving product quality and responsiveness. Therefore, after evaluating the related project, five criteria Were identified: speed, cost, responsiveness, competency, and quality. To identify the appropriate criteria to rank LAs of supply, those attributes that help meet the rapidly changing needs of the business environment of supply were considered. Thus, LAs were evaluated on the basis of quality as well as the speed of providing products, services, and information to customers, suppliers, and employees, as quality and speed are significant criteria in the leanness assessment of a supply chain. Cost was another criterion identified for the evaluation of LAs because the focus of the lean approach is on cost reduction. The responsiveness ability of a lean supply was also considered. Another notable criterion in pair-wise assessment of LAs was the competency ability obtained in a lean supply by achieving the mentioned LAs. These five criteria are important in evaluating the degree of leanness achieved in a supply chain; therefore, the mentioned alternatives were evaluated with respect to each criterion to determine how much achievement of the LAs may lead to a leaner supply. Output of FQFD represented as Wi, is the input of the Second Fuzzy-QFD (SQFD) component of the proposed model. C. Determine The Relation Between LAS And LES And Correlation Between LES Because of the qualitative and ambiguous attributes linked to lean implementation, most measures are described subjectively using linguistic terms that cannot be handled effectively using conventional approaches. However, fuzzy logic provides an effective means of dealing with problems involving imprecise and vague phenomena. It was exploited to translate linguistic judgments required for relative importance of LAs, relationships, and correlations matrices into numerical values. In this step, the degree of relationship between LAs and LEs was stated by the corresponding TFNs and placed in the HOQ matrix. Moreover, the degree of correlation between LEs was then expressed by TFNs in the fuzzy HOQ. Both of these correspondences are shown in Tables 2, 3 and 4. TABLE 2 Degree of relationships, and corresponding fuzzy numbers[1] Degree of relationship Fuzzy no. Strong (0.7; 1; 1) Medium (0.3; 0.5; 0.7) Weak (0; 0; 0.3) TABLE 3 Degree of correlations, and corresponding fuzzy numbers[1]

5 Degree of correlation Fuzzy no. Strong positive (SP) (0.3; 0.5; 0.7) Positive (P) (0; 0.3; 0.5) Negative (N) (-0.5; -0.3; 0) Strong Negative (SN) (-0.7;-0.5;-0.3) TABLE 4 The 4 point linguistic scale for importance judgment Importance judgment Fuzzy No. Very High (VH) (0.7; 1; 1) High (H) (.5; 0.7; 1) Low (L) (0; 0.3; 0.5) Very Low (VL) (0; 0; 0.3) 1) TFN (TRIANGULAR FUZZY NUMBER) The TFN can be denoted as a triplet (a, b, c), as shown in Fig. 3., where, a b c. When a = b = c, it is a nonfuzzy number by convention. The membership function can be defined as[2] : (x-a)/(b-a), x is a function of [a,b] µn(x):(c-x)/(c-b) x is a function of [b c] 0 otherwise 1 Fig 3 Triangular fuzzy number (TFN). If M = (a1, b1, c1) and N = (a2, b2, c2) represent two TFNs, then the required fuzzy calculations are performed as given below. Fuzzy addition : M + N = (a1 + a2; b1 + b2; c1 + c2 ) 2 Fuzzy multiplication : M N = (a1 a2; b1 b2; c1 c2 )..3 M 1/N = (a1/c2; b1/b2; c1/a2).4 Fuzzy and natural number multiplication: r M = (r.a; r.b; r.c)... 5 D. CALCULATE RELATIVE THE IMPORTANCE (RIJ) AND PRIORITY WEIGHTS OF LEs (RIJ * ) The aim of computing these two parameters was to determine which LE has the most effect on the lean supply chain. RIj was computed by fuzzy multiplication of Wi to Rij. RIj = Wi Rij j=1,2,3 m 6 RIj * = RIj+ Tkj RIk j=1,2,3 m 7 Tkj was shown in the roof part of HOQ. The mentioned parameters are shown in Fig. 3. Furthermore, normalization was performed by dividing each RIJ * by the highest one according to the fuzzy set algebra. Then, in order to rank the LEs, the normalized scores of RIJ * were de-fuzzified. Suppose M (a, b, c) is a TFN; then, the de-fuzzified value

6 is computed as (a+4b+c)/6. 8LEs with high crisp values indicate that they can be usefully exploited to enhance relevant LAs. Thus, such enablers must be selected for implementation. RESULT AND DISCUSSION This section presents an example of the proposed approach through a case study in the liquor industry to illustrate the usefulness and ease of application of the method as well as considering the practical implications of the approach. Focusing on the methodological point of view, the definition of a specific set of and LEs for applying the approach was not dealt with in this project; they should be identified according to the special characteristics of the company under consideration.. Then, 8 LAs and 11 LEs were identified these are shown below in Fig.4 Step 1: identifying the competitive bases a company is willing to achieve competitive advantage; Step 2: identifying lean attributes enhancing the selected competitive bases and filling the first HOQ; Step 3: identifying lean enablers to be exploited in order to achieve the required lean attributes, and filling the second HOQ. APPLICATION STEP 1 As the starting point of the approach proposed, a company should identify the relevant competitive bases. For illustration purpose, the set of competitive bases used in the numerical example has been ground on existing studies and information available in literature. Specifically, a viable list of five competitive bases (CBi, i = 1,...,5), namely speed, cost, responsiveness, competency and quality, was derived. They are listed as whats in fig. 4. In real case applications identifying the relevant competitive bases of a company would require direct contacts with company s members (in particular, marketing manager), either in the form of interviews or roundtable discussion. To support the application of the methodology in practice, and to quickly collect the required information, it is suggested to setup an appropriate workgroup, headed by academics and including firm s executives, reporting to the main business functions involved in the development of lean strategies. APPLICATION STEP 2 Step 2 requires filling the first HOQ, which, in turn, involve the following sub-steps: Defining the fuzzy linguistics scales; Assessing the relative importance of competitive bases; Listing the lean attributes Assessing the relationship between lean attributes and competitive bases; Identifying possible correlations between lean attributes. Sub-step i. Fuzzy linguistics scales to be used to assess weights of CBs, relationships between CBs and LAs and correlations between LAs could be either defined by the workgroup or derived from the literature. In this example, they were taken from Bottani and Rizzi (2006). Sub-steps ii. In this example, the relative importance wi of CBs was defined based on the work by Ren et al. Starting from findings by the authors, wi (i = 1,...5) were pondered based on a normalised 4-point fuzzy linguistic scale, ranging from very low (VL) to very high (VH), as shown in table4. Relative importance of CBs is listed in the second column of Fig.4. Whenever the procedure is applied to a real case, the same information can be derived asking company s members to express their judgment against the relative importance of competitive bases with regard to the over all strategy of the company. Judgment will be thus translated into fuzzy numbers according to the scale defined in the previous sub-step. Sub-step iii. A viable list of LAs should be defined depending on the specific case in example. For the purpose of this example 8 lean attributes suggested by Zarei et al. Where listed as LA1,..., LA8 in columns in the HOQ, and their acronyms are detailed more in fig 4. Sub-step iv. Selecting the set of LAs proposer by Yusuf et at. Greatly simply the assessment of the relationships.

7 Since the impact of those lean attributes on the above mentioned set of competitive bases was investigated by Ren et al. Accordingly, as well grounding on the fuzzy scale proposed in table 2 to4, the relationships matrix of this example was built as shown in the centre of fig.4. The same assessment, in real case, requires inner viewing the workgroup, to drive information concerning the impact of each LAs on the CBs identified in sub-step ii. Specifically, company s member should be asked to assess how, and to what extent, an LA has potential to enhance a given CBs; judgments could be expressed on a linguistics scale and translated into fuzzy numbers according to previous sub-step i. Sub-step v. Outcomes proposed in table 5 were used as a guideline to derive correlations required in the roof of the HOQs, as all the studies cited in the table suggest possible links between LAs. According to such findings, the roof of correlations of the first HOQ was built as shown in fig.4. Then, the relative importance RIj (j = 1,...,8) and the final scores of lean attributes were computed applying Eqs. (6) and (7). Outcome of the computation are presented in the last row of fig.4. The application of the methodology to a real case would require interviewing company s members, to get, based on their in-field experience, information concerning the possible impact of lean attribute on another. Such information should be expressed following the linguistics scale defined in sub-step i, and translated in fuzzy numbers for computational purpose. None the less, findings from the literature could be useful to suggest possible interactions between lean attributes to interviewees, thus helping in the evaluation Sub-step vi. To find the normalized value of LAs so this can be use in second fuzzy QFD as the relative weight (wi) as shown in fig 4 the normalized value shown in table 5 APPLICATION STEP 3 Requires building the second HOQ, which involves the following sub-steps: defining the fuzzy linguistics scale; assessing the relative importance of lean attributes; listing the relevant lean enables; assessing the relationship between lean attributes and lean enablers; and identifying possible correlations between lean enablers. Sub-step i. To be consistent with results of the previous application step, in this example we adopt the same fuzzy linguistics scales to assess weights of LAs, relationships between LAs lean LEs in the second HOQ. Sub-step ii. Depending on the specific case study, the second HOQ can be built starting from all LAs examined in the previous step, or the analysis can be limited to those attributes which got the highest score in the first HOQ. In order to thoroughly illustrate the application of the methodology, in this example all LAs previously examined will be considered as whats in the second HOQ. Importance weights of LAs can be derived from the final score obtained in the first HOQ. However, since a normalized fuzzy scale has been adopted to express importance judgment of competitive bases in the previous step, a preliminary normalization of fuzzy scores of lean attributes is suggested before they are used as importance weights in this step. Normalization is performed by dividing each score by the highest one, i.e. score of LA4, according to the fuzzy sets algebra. Lean attributes and related normalized

8 Fig 4 First Fuzzy QFD approach For finding Wi of LAs TABLE 5 Normalized value of lean attributes Lean Attributes Relative weight (Wi) or normalized score LA LA LA LA LA LA LA LA importance weights are thus listed in the first columns of the second HOQ, shown in fig.5. Sub-step iii. As per the previous steps, LEs to be implemented by companies in order to acieve leanness should be defined according to the case study company considered. Interviews with company s members could be useful to

9 derive a possible list of enablers. In this example, starting from the works of Gunasekaran, and gunasetaran and yusuf, 11 viable lean enablers were identified and listed as LE1,...,LE11 in columns in fig.5. Acronyms are explained in note to the figure. Sub-step iv. To our knowledge, no specific studies are currently which thoroughly describe the impact of lean enablers on lean attributes, thus directly providing the relationships matrix of the second HOQ. Such relationship, however, are partially dealt with by scientific literature. The resulting values are proposed in the centre of fig.5.findings from the literature could be useful even in a real case application, to suggest possible interviewees to help in the evaluation. Sub-step v. As per the previous sub-step, in this example possible correlations between LEs were derived from the literature. On the basis of the literature examined, as well as on the degree of correlations proposed by company, the roof of correlations of the second HOQ is built as shown in fig. 5 Findings from the literature could also be useful in the case the second HOQ is built based on experts opinion, to suggest possible interactions between enablers, as well as correlations between enablers, thus helping in the assessment. The relative importance RIk (k = 1,...,7) and the find scorek of LEs were computed according to Eqs. (6) and (7). Eq. (8) was finally adopted to derive crisp scores. Outcomes of the computation are presented in the last rows of fig. 5. AS a result of the computation, supply chain management practices Quality improvement (LE8) got the highest crisp score, due to both the wide number of positive correlations with others LEs and strong relationships with several LAs. Thus, such an enabler has the highest implementation priority in order to achieve leanness, followed by Service level improvement (LE1) and Human resource training (LE7) shown in table 6. FIG 5 Second Fuzzy QFD For ranking of Lean enablers TABLE 6 RANKING OF LEAN ENABLERS

10 Rank LEs Crisp Value 1 LE LE LE LE LE LE LE LE LE LE LE CONCLUSIONS AND SUGGESTIONS In this study, an integrated Fuzzy-QFD approach was proposed to enhance the leanness of the supply chain. The approach showed the applicability of the QFD methodology, especially of the HOQ, to identify viable lean enablers for achieving a defined set of LAs. The FQFD was used to prioritize lean attributes. In order to cope well with the vagueness of linguistic judgments required in building the HOQs, relationships, and correlations, Wi, relative importance (RIJ), and priority weights (RIJ * ) of LEs were all defined with TFNs. A case study was presented to illustrate the ease of application of the proposed approach. The focus of attention of future researches can be on the integration of useful methods with QFD to prioritize LEs in order to enhance supply chain leanness. Future researches can also consider utilizing other ranking methods instead of the FQFD, such as TOPSIS, AHP, to prioritize the LAs and compute their priority weights. Moreover, Wi, i.e. LA priority weights, obtained from different ranking methods, can be compared. In this project, a case study of a company in the liquor industry was presented. More case studies for other supply chains should be presented. Acknowledgment The author wish to express their sincere appreciation to the top management of belsend sugar company (distillery division ). DR. Ghanshyam Singh for his useful advice and comment. REFERENCE [1] Bottani E. Rizzi, AStrategic management of Logistics service: A fuzzy-qfd approach. International Journal of Production Economics 103 (2), 2006, pp [2] Chamodrakas I., Alexopoulou N., Martakos D.,Customer, evaluation for order acceptance using a novel class of fuzzy methods based on TOPSIS. Expert Systems with Applications 36, [3] Chien C.J, Tsai H.H.,Using fuzzy numbers to evaluate perceived service and quality. Fuzzy Sets and Systems 116, 2000, pp [4] Dohnal.M, Vystrcil.J, Dohnalova.J, Marecek.K, Kvapilik.M, Bures.P, Fuzzy food engineering. Journal of Food Engineering 19 (2), 1993, pp [5] Georgiadis, P, Vlachos. D, Iakovou. E. A system dynamics modeling framework for the strategic supply chain management of food chains. Journal of Food Engineering 70 (3), 2005, pp [6] Hopp, W.J., Spearman, M.L., To pull or not to pull: what is the question? Manufacturing and Service Operations Management 6 (2), 2004, pp

11 [7] Hunt, I., Wall, B., Jadgev, H., Applying the concepts of extended products and extended enterprises to support the activities of dynamic supply networks in the agri-food industry. Journal of Food Engineering 70 (3), 2005, pp [8] Nambiar, A.N., Mahalik, N.P., Trends in food packaging and manufacturing systems and technology. Trends in Food Science and Technology 21 (3), 2010, pp [9] Schonbergerm, R.J., Japanese production management: an evolution with mixed success. Journal of Operations Management 25, 2007, pp [10] Scherrer-Rathje, M., Boyle, T.A., Deflorin, P., Lean, take two! Reflections from the second attempt at lean implementation. Business Horizons 52, 2005, pp