Measuring the impact of (s, S) ordering policy on the bullwhip effect by means of simulation optimization

Abstract

An important supply chain research problem is the bullwhip effect (BE), which its presence makes supply chain planning difficult. In fact, demand fluctuations increase as one move up the supply chain from retailer to manufacturer. In this paper, a single-stage supply chain for a single-item inventory system is considered. The aim is about how identify the critical parameters that affect BE phenomenon when the retailer employ a (s, S) policy. To investigate and measure this impact, a simulation model is developed using Arena 10 software. A simulation-optimization based-approach is adopted, which combines metamodel through the factorial design experiments and the desirability function approach. Our findings suggest that when we want to reduce the BE, this doesn't mean directly reduce costs. The BE must be introduced in the supply chain management such a constraint not such an objective function (BE must be lower to 1 to have not a bullwhip effect), and the effect of S is always more significant than the effect of s, and it is valid for all studied situations. Otherwise, in a (s, S) policy, the choice of S is more determinant that s, to reduce costs in the supply chains as well BE.