Combined routing and product take-back strategies in reverse logistics

Reverse logistics concerns ways of recovering some or all parts of previously shipped products to customers for possible reuse, recycling, remanufacturing, or disposal. This recovery process results in a reverse flow of materials and products from users upstream to producers. We study integrating forward and reverse movement of products at the routing level for a system consisting of a depot and a number of geographically dispersed customers who order certain quantities of a single product that can be recovered in a short time. The problem is considered under a multi-period planning horizon where future customer demand is uncertain. Customer demand is delivered either by the depot owned single vehicle or by outside carriers. The depot owned vehicle is utilized to pick up all the returned products during its delivery trips with possibility of postponing pick ups of some returned products with penalty.; For a given vehicle route, finding a pick up policy is formulated as a Markov Decision Process (MDP). Optimal pickup policies are derived for three special cases of this MDP and a heuristic pick up policy is proposed for the general case. The heuristic pickup policy is tested against derived lower bounds. We discuss a rudimentary algorithm for finding a joint solution of a vehicle route and a pick up policy and ways of expediting running this algorithm.