Strategic and tactical optimization of distribution networks for perishable products

This work investigates strategic and tactical optimization of distribution networks for perishable products. This kind of products is characterized by a given remaining lifetime that progressively reduces until a minimum value is reached (typically, the zero level). In other words, perishable products are characterized by a certain utility that remains more or less constant until the expiration date is reached. After expiration, the utility goes to zero (Nahmias, 1982). Optimization of the delivery of such products poses important problems such as reduction of wastes and efficient exploitation of avail- able resources within the distribution network. First, a discrete-time dynamic model of the network is constructed: the model is made up of a directed graph, where nodes and arcs represent warehouses and transportation links, respectively. Second, strategic goals consisting in the selection of optimal values of stock replenishment cycles, safety stocks, and amounts of products to transfer are pursued via optimization techniques taking into account uncertainties in the demands of customers. Third, strategic decisions are tuned according to a rolling-horizon optimization approach implementing a tactical decision viewpoint, where no uncertainties on the demands of customers exist. This framework, originally proposed by Gaggero and Tonelli (2020), is extended by considering new opti- mization models in the strategic step to better account for uncertainties in the requests of customers. Simulation results in different demand scenarios confirm the effectiveness of the proposed framework as compared to a traditional lot-for-lot approach.