A mathematical model for choosing between centralization and decentralization in distribution networks for perishable products

An optimal supply chain (SC) configuration improves the performance of distribution companies for perishable products since it reduces logistic costs while ensuring high service levels and preventing product spoilage. The SC configuration should not be optimized only once, but rather continuously reviewed since customer demand for perishable products varies over time. A key decision to be reviewed is whether to centralize or decentralize inventories. This decision allows for matching the allocation of stocks in distribution centers (DCs) with customer needs, ensuring product availability, reduced waste due to product spoilage, and lower logistic costs. However, distribution companies for perishable products struggle in reviewing their SC configuration since structured methodologies for choosing between centralization and decentralization are missing. Due to this gap, the performance of distribution companies for perishable products is often undermined. To address this gap, this work provides distribution companies for perishable products with a novel mathematical model that allows comparing five SC configurations, where inventories are either centralized, decentralized, or managed with three hybrid configurations. Through the mathematical model, the optimal SC configuration is identified as the one with minimum total logistic cost, which also respects pre-established service levels and the product shelf life. The applicability of the mathematical model is tested in the case study of an agri-food distribution company. The results show that, by adopting the proposed model, distribution companies can associate individual perishable products with the optimal SC configuration, achieving cost savings, high service levels, and reduced waste. Moreover, the results confirm that decentralization is advantageous for perishable products with short shelf lives, while centralization is cost-effective in other cases.