Nusrat Hossain Zerin ? & Dewan Maisha Zaman ? I. Introduction lectronics industry is introduced in Bangladesh during the post liberalization era. Due to extension of information technology, many Bangladeshi companies start to produce electronics goods and Walton is the first local electronics industry in Bangladesh. Electronics waste or e-waste is the result of this. . E-waste can be made profitable and useful by establishing RL system. Reverse logistics means the reuse of products and materials. In reverse logistics system, the source goes at least one step back in the supply chain. Normally in supply chain products move manufacturers to distributors or customers. Any process or management which maintain or continue after the sale of the product called reverse logistics. Defective product can be returned by customer. The manufacturing firm would then have to organize shipping of the defective product, testing the product, dismantling, repairing, recycling or disposing the product. The product would move in reverse through the supply chain network in order to retain any use from the defective product. The logistics for such matters is Author ? ?: Dept. of Industrial Engineering and Management, Khulna University of Engineering & Technology, Bangladesh. e-mails: [email protected], [email protected]

Reverse logistics [4] found that effective RL focuses on the back-ward flow of materials to maximize value from returned items and guarantee their proper disposal [1,5,6]. However, many companies are not yet ready to implement RL including Walton. Walton Hi-Tech Industries Ltd. has been selected for the study because it has a high consumption volume, and major source of waste generation. Also, this is one of the few sectors which come under e-waste regulations [1, 5, and 6].A thorough study of CSFs and their ordered implementation is essential for successful RL implementation. The major intention of this study is to understand various CSFs for RL implementation in Walton Hi-Tech Industries Ltd. The identification and prioritization of these factors will help the researchers and the managers in strategic decision making for RL implementation. After review of literature on RL and the opinion of experts, 11 CSFs factors of RL implementation were identified. The experts were asked to rate each of these 11 factors in terms of their importance. A decision matrix was developed from these responses which are used in the application of fuzzy-TOPSIS methodology for prioritizing CSFs [1, 5and 6].

Fuzzy TOPSIS algorithm is used for Prioritizing Reverse Logistics Implementation Factors. This algorithm consists of 8 steps. These steps are presented in detain as follows: (0.5,0.7,0.9) High (0.7,0.9,1.0)

Step 2: The TOPSIS method evaluates the following fuzzy decision matrix.

Abstract-The reverse logistics (RL) try to discover to assist the return of consumer supplies or its ingredient materials in the production phase. Nowadays reversed logistics has been achieved more attention in supply chain management. Political, economic, green image and social responsibility etc. influence firms to develop strategies to their current systems because, substitute uses of resources must be promoted that may be cost-effective and ecology friendly by extending products' routine life cycles in the turbulent business environment. Reverse logistics activities i.e. storing, transporting and handling of used products poses a great challenge to reverse logistics managers as there is always chances of uncertainty in terms of quantity, quality and timing of return of end of life (EOL) products. This approach intends to investigate the going-over dynamics of reversed logistics implementation based on survey opinion polls for Walton electronics industry by a real life case study. Fuzzy Analytic Hierarchy Process (FAHP) was selected for identify the vital gauges from different companies' professionals of confirmation and electronics firms to do the related ecofriendly management work. The prirotization of each factor has been obtained. The overall weights and performance ranking of the evaluation criteria have been shown by the result of this study with respect to the reverse logistics implementation.

Step 1: Collecting the required data containing linguistics terms. A proper scale must be chosen to represent the data. Respondents must be asked to choose the best alternative among the linguistics terms for a given question. Fuzzy numbers for the selected linguistics terms are presented in Table 1. (1)

Where is a fuzzy number corresponding to the linguistic term assigned by the I the Decision Maker (DM) to the j th factor. i=1, 2,?., m are the number of DMs and j = 1, 2, ?, n are the number of factors (CSFs).

Step 3: This step includes neutralizing the weight of decision matrix and generating fuzzy un-weighted matrix (R).

Where ?? ?? * =max i ?? ????

Step 4: Calculate the weighted normalized decision matrix

The weighted normalized value ?? ???? is calculated as

Where ?? ?? is the weight given to each decision maker? ?? ?? = (1, 1, 1, 1, 1)??? ? ??, because all the DMs are considered to have same weight for this study.

Since the positive and negative ideas introduced by Chen (1997) are used for the research. The following terms are used for ideal and negative ideal

?? ?? ? = (0, 0, and 0) (8)

Step 6: Calculate the sum of distances from positive and negative ideal solution for each factor.

) is the distance between two fuzzy numbers which can be calculated using the vector algebra. For example distance between two numbers A1

Similarly, the separation from the negative ideal solution is given as

Step 8: Prioritize the preference order based on the order of the values of ?? ?? .

Many CSFs are common to all of these studies and these factors can be utilized as base for discussion with expert from Walton Hi-tech Industries ltd. Eleven CSFs were identified after pertinent literature review including studies discussed in ''Introduction'' and discussion with the number of experts from the Walton Hi-tech Industries ltd. These factors are Total manufacturing cost (C1), Recycling (C2), Environment concern (C3), Recycled volume (C4), Economic Need (C5), Pressure with stakeholders (C6), Reverse logistics management information syst (C7), Top management awareness (C8), capabilities & skilled workers (C9), Increase of sales volume for new product (C10), Consumer awareness &Social acceptability (C11).

The fuzzy TOPSIS methodology, presented in this research paper has been evaluated in context of Walton electronics industry. Four experts from electronic companies participated in this study. Profile of the decision makers and their respective organization is given as follows and their respective organization is given as follows:

First decision maker (DM1) is a supply chain manager in Walton electronic industry. Second decision maker (DM2) is a logistics manager in Walton electronic industry. Third decision maker (DM3) is a logistics manager in Walton electronic industry. Fourth decision maker (DM4) is vice president of operations management of same industry.

Step 7: Calculate the relative closeness to the ideal solution. The relative closeness with respect to ?? * is defined as

Step 5: Determine the ideal and negative-ideal solution for the CSFs (?? 1 ,?? 1 , ?? 1 )and A2(?? 1 , ?? 1 , ?? 1 )can be calculated as To generate R, following relationship can be applied solution. Here all decision maker weight is 1 and ?? * is (1, 1 ,1) and ?? ? is (0,0,0). So the table for Normalized fuzzy decision matrix for criteria and weighted normalized alternatives, FPIS and FNIS are same as table 2.

To prioritize the CSFs for RL implementation in Walton Electronic industry, 11 factor legislation, Total manufacturing cost (C1), Recycling cost (C2), Environment concern(C3), Recycled volume (C4), Economic Need(C5), Pressure with stakeholders (C6), Reverse logistics management information system(C7), Top management awareness(C8), Process capabilities & skilled workers(C9), Increase of sales volume for new product (C10), Consumer awareness &Social acceptability (C11),identified in section ''Identification of CSFs for RL implementation'' are considered for the prioritization. Four decision makers DM1, DM2, DM3, and DM4 were asked to rate the importance of the above mentioned each CSF on a 5-point scale having the linguistic terms low (L), fairly low (FL), medium (M), fairly high (FH), and high (H). The decision makers used the linguistic variables shown in table 2to assess the importance of the CSFs. A decision matrix was prepared based on the responses received from the DMs shown in table 3. As mentioned in the fuzzy-TOPSIS methodology step 1, triangular fuzzy numbers were used to convert linguistics variable into the fuzzy numbers. By converting the fuzzy linguistic variables into triangular fuzzy numbers using table 1, the fuzzy decision matrix D was obtained. In the next step unweighted fuzzy decision matrix R was enumerated. Further steps were followed to obtain the weighted fuzzy normalized decision matrix, to find the ideal and negative-ideal solutions for the CSFs. The distance ?? ? and ?? * of each CSF is derived, respectively, by using Eqs. (7), (8), (9), and (10). The?? ? and ?? * closeness coefficient C for each CSF is obtained by using Eq. (11 The findings of the research will help the managers and academicians in the development of RL strategies and practices in Bangladesh electronics industry. These CSFs can also be used for RL implementation in other sectors of Bangladesh industry. Like other studies, this study also has some limitations. This study is conducted using for experts from the Walton electronic industry. Future studies may consider larger sample size to assess the methodology and the effectiveness of the proposed solution to enable generalization. Furthermore, the wider rating of the 7 or 11-point linguistic scale could be used instead of using a 5-point linguistics scale. Researchers may utilize other methodologies including other MCDM methodologies and may compare the results. Future studies may be carried out to identify company-specific or productspecific identification of CSFs for RL implementation.