A stochastic model for better planning of product flow in retail supply chains

Abstract

Retail supply chains operate in a constantly changing environment and need to adapt to different situations in order to increase their reliability, flexibility and convenience. Holding and transportation costs can amount to up to 40 per cent of the product value, so that the proper coordination of interrelated activities plays an essential role when managing retail flows. In order to provide a relevant model we first focus on future demand satisfaction, whereas pricing policies, perishability factors, etc., are subjected to a complementary model for operative planning. The idea is to obtain a preferable distribution plan with minimal expected distribution costs, as well as minimal supply risks. The used methodology produces a set of solutions and quality estimates which can be used in order to find a desired distribution plan which is near-optimal. While considering stochasticity on the demand side, a multi-objective optimisation approach is introduced to cope with the minimisation of transport and warehouse costs, the minimisation of overstocking effects and the maximisation of customer’s service level. The optimisation problem that arises is a computationally hard problem. A computational experiment has shown that the version of the problem where the weighted sum of costs is minimised can be handled sufficiently well by some well-known simple heuristics.