Possibilistic programming for designing reverse logistics networks to handle e-commerce returns under uncertainty

Returns in e-commerce are a significant cost factor, impacting the firms' profitability and customer satisfaction. This paper proposes a fuzzy possibilistic programming model for designing a multi-echelon reverse logistics network to optimise return management under uncertainty from returned demand, transportation and inventory costs. The model determines the optimal number and location of facilities, material flow and inventory level. Finally, a case study is conducted on an e-commerce business in Vietnam to evaluate the model's effectiveness. The network includes 36 collection sites, six intermediate warehouses and a regional hub with many convenience stores serving as collection centres. Sensitivity analyses on pertinent parameters have been performed to provide valuable managerial insights. Using the fuzzy model with a confidence level of 90%, the total cost is 326,281.6 USD, while the deterministic model generated a total cost of 361,004.5 USD. As a result, our model is efficient in providing solutions for enhancing business performance.