Reverse Supply Chain Design: A Neural Network Approach

The success of a reverse supply chain heavily relies on the efficiency of the collection facilities and recovery facilities chosen while designing that reverse supply chain. In this chapter, we propose a neural network approach to evaluate the efficiency of a facility (collection or recovery) of interest, which is being considered for inclusion in a reverse supply chain, using the available linguistic data of facilities that already exist in the reverse supply chain. The approach is carried out in four phases, as follows: In phase I, we identify criteria for evaluation of the facility of interest, for each group participating in the reverse supply chain. Then, in phase II, we use fuzzy ratings of already existing facilities to construct a neural network that gives impacts (importance values) of criteria identified for each group in phase I. Then, in phase III, using the impacts obtained in phase II, we employ a fuzzy TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) approach to obtain the overall rating of the facility of interest, as calculated by each group. Finally, in phase IV, we employ Borda’s choice rule to calculate the maximized consensus (among the groups considered) rating of the facility of interest. Kishore K. Pochampally Southern New Hampshire University, USA