Forecasting Techniques in Logistics

Abstract

This article deals with the selection of distribution control techniques and discusses one element of distribution control, the type of forecasting techniques. It first goes into the detail of a classification of distribution control decisions, the bigger framework underlying this article which can be used to select appropriate distribution control techniques. The type of forecasting technique is one of these distribution control decisions, and this is the topic of discussion for the rest of the article. The results of case study research and literature research on the application of forecasting techniques are described. A simulation model is presented, which is used to research the usefulness of forecasting techniques. As opposed to most research on forecasting, the relation between characteristics of products, processes and markets on different types of forecasting techniques is investigated from a logistics perspective in this simulation. The effect of forecasting techniques is assessed based on the impact the techniques have on the inventory level, not on a forecast accuracy measure. Two techniques are discriminated: techniques that can incorporate demand patterns and techniques that can not incorporate these patterns. It appears that the combination of a seasonal pattern or a trend in demand with demand uncertainty has a significant impact on the choice between these two techniques. Only if demand uncertainty is low and if demand contains clear patterns, forecasting techniques that can incorporate demand patterns outperform those that can not.