Date/Time
Date(s) - 16/02/2022
3:30 pm - 5:00 pm

Speaker: Dr. Mehmet Erdem Coskun

Topic: Retail Reverse Supply Chain Management and Optimization, Managing Retailer’s Inventory via Reverse Logistics

Date of Presentation: Wednesday, February 16, 2022

Location:  Online

Abstract:

In our research, we model a comprehensive and realistic retail reverse supply chain network where there are multiple stores under different store types, namely, Company Owned and Franchise stores, multiple warehouses, multiple distribution centers, multiple vendors and liquidators. Each store holds certain products that are not selling and stores want to get rid of the items of these products from their inventory and hence the retail company wishes to salvage them optimally. The objective of the retailer is to minimize related costs for relocating these products within and outside of the network. However, individual franchise stores have their own goals of how their ineffective inventory should be handled. Hence, there is a conflict between franchise stores’ objectives and the retail company’s objective in terms of how much inventory should be chosen from each store to be removed. The strategic decision-making process of which Distribution/Return Centers should be activated is implemented and designed into the model structure since it might be very costly to use all of the return centers for the reverse logistics activity. Some products have demand at other parts of the network and reverse network has a strict way of routing this excess inventory via return centers and warehouses. Return Centers are a pathway to inspect-sort-separate of the returned products and warehouses are the suppliers to the stores with limited capacity for this activity since they also serve the stores with a regular forward supply chain. Vendors use a multi-layer penalty structure for receiving products from the retailer. We model the realistic structure of the supply chain network where stores can return products to certain distribution/return centers, distribution centers route products to certain warehouses under their network or certain distribution centers. The realistic nature of the network, logistics\&distribution, cost and profitability structure of the products are modelled by considering facility activation costs, capacity limitations, demand for the products at other locations, product dimensions, weight \& volume of the products, rebate rates, detailed cost structure of the product, the unit cost of transportation from location to location, and receiving\&handling costs. We model the problem as a Mixed Integer Linear Program (MILP) and solve test problems using CPLEX. In order to overcome the complexities of the realistic nature of the problem, we develop a multi-stage heuristic that has sub-routines to solve the problem close to optimality within reasonable solution times and compare the heuristic algorithm performance results with CPLEX results. We investigate the results and summarize our findings.