The effect of smoothing filters on supply chain performance

This paper examines how the form of the sales smoothing function affects the performance of a supply chain. Modelling uses Laplace and z transforms with Simulink™ implementations of the equations for an APVIOBPCS inventory system. Inventory stock-out depends on the delay function representation and strongly on the sales smoothing function. The models presented here use a control-theoretic formulation to determine sales smoothing functions, which can be designed to substantially reduce the inventory response to changing sales. The analysis shows the effect of replacing the traditional moving average technique (EWMA) with one based on control theory can result in a much lower stock-out value for a shorter time improving both bullwhip effect and customer service levels. Performance of this filter in an EPOS enabled supply chain is sufficient to reduce the stock-out for the retailer by 50% with a substantial reduction for other members of the chain and almost eliminating bullwhip.