Unlike most of the existing literature on reverse supply chains, that focuses on product recovery or waste management, in this thesis we consider reverse supply chain operations for an independent retailer. The latter have forward and reverse supply chains that are independent of the manufacturers. We study three major problems related to Retail Reverse Supply Chains (RRSC) for independent retailers. In RRSCs, each retail store holds some products that are not selling (and/or under-selling) and wishes to salvage them optimally. We refer to these products as Ineffective Inventory. Salvage can be in many forms and take place by relocating a product within the reverse supply chain (RSC), such as sending the product from a franchise store back to a Distribution/Return Center (RC) and then forward to another franchise store, or returning it to a vendor, liquidation, etc. The RRSC network may includes system members such as stores (retailer owned and/or franchise), RCs, warehouses, vendors and liquidators. Each of the stores carries some inventory that is underselling, and it is important to reduce the inventory of such products in order to refill the space with inventory that is more likely to sell.
In the first problem, we consider a basic RRSC with retail stores, vendors and a warehouse. The retail company allocates a budget for its RRSC activities. We refer to this budget as a Profit-Loss budget, due to lost income from the items that will be removed from the stores that was a part of the gains resulting from the previous year tax calculations. The objective is to use this Profit-Loss budgetary limitation as effectively as possible with the most suitable products to relocate products within the supply chain and/or return them back to their vendor. A heuristic algorithm is developed to solve this problem, by making use of the problem structure, and results are compared with the solutions of an exact state-of-the-art commercial solver.
In the second problem, we consider a network optimization model with inventory decisions. The goal is to optimize ineffective inventory levels in stores and the disposition of their returns. We model a comprehensive RRSC network with multiple stores that could be Company-Owned or Franchise Stores, multiple warehouses, multiple RCs, multiple vendors, and liquidators. The objective of the retailer is to minimize costs for relocating some of this ineffective inventory within the network or scrapping. However, individual franchise stores have their own goals of how their excessive inventory should be handled. The franchisee goals may be conflicting with those of the franchisor in terms of how much inventory should be chosen from each store to be relocated. In return, this conflict may lead to a conflict among franchise stores. This issue is addressed and resolved through inventory transparency among all the supply chain members. The tactical decision making process of which RC should be used for handling returns is incorporated into the model. In order to overcome the complexities of the large size problem, a multi-stage heuristic is developed to solve this problem within reasonable times. The results are then compared with the solutions of state-of-the-art commercial solver.
In the third problem, we focus on the strategic decision of developing optimal vendor contract parameters for the retailer, using optimization models. Specifically, we identify optimal return penalties and associated return thresholds, between an independent retailer and its vendors. This model will support the retailer in their contract re-negotiation for its RSC activities. Vendors use a multi-layered penalty structure that assigns higher penalties to higher returns. The objective is to find the optimal penalties and/or optimal return thresholds that should be negotiated with the vendors in order to pay a lower penalty in the upcoming return cycles compared to existing penalty structures. We first design a Mixed Integer Non-Linear Program (MINLP) where the model makes the decision of vendor penalty fees and return thresholds simultaneously for each vendor. We generate small size to large size problems and solve them via MINLP solvers such as DICOPT and ANTIGONE. In order to gain insights to the inner workings of the MINLP, the decision variables, vendor penalty fees and return thresholds, are considered as parameters and hence, two models are designed to find the optimal penalty structure and optimal return thresholds, respectively. Useful insights from both of the models’ solutions are derived in order to generate rule-of-thumb methodologies to find approximate solutions close to optimal penalty percentages and return thresholds via identifying all possible scenarios that can exist in the problem structure. / Thesis / Doctor of Philosophy (PhD) / This thesis deals with Retail Reverse Supply Chain (RRSC) management. We consider an independent retail company's and its franchise stores' ineffective inventory which may be constituted of unsold, under-selling, slow-moving, customer-returned, end-of-life, end-of-use, damaged, and faulty products within their inventory. We take into account the retailer's reverse supply chain structure and investigate the following problems: 1) How to manage a store's product returns under a given budgetary limitation for financial planning and taxation reasons, due to lost income from returned items, 2) Inventory optimization by taking into account the reverse supply chain structure of the retailer, and 3) Providing insight to the retailer on how it can best re-negotiate its vendor (buy-back) contracts for its product returns. The thesis covers decision making in all three levels: day-to-day operational decisions such as which products to be returned and where to allocate them within its reverse supply chain options, mid-term tactical decisions such as which Return Centers (RC) to be activated for the Reverse Logistics (RL) activities, and long-term strategic decisions such as what should be the optimal contract terms to re-negotiate with the vendors in order to cut future return costs.
| Identifer | oai:union.ndltd.org:mcmaster.ca/oai:macsphere.mcmaster.ca:11375/27582 |
| Date | 16 June 2022 |
| Creators | Coskun, Mehmet Erdem |
| Contributors | Hassini, Elkafi, Computational Engineering and Science |
| Source Sets | McMaster University |
| Language | English |
| Detected Language | English |
| Type | Thesis |
Page generated in 0.0027 seconds