A Predictive Model for Forecasting Spare Parts Demand in Military Logistics

Abstract

The proportion of the stock range that is devoted to spare parts is often considerable in industrial context. Accordingly, even small improvements in forecasting spare parts demand might lead to substantial cost savings. Time series analysis has been the most popularly applied method in the prior spare part demand forecasting models. However, these approaches need to be improved in terms of prediction accuracy. In this study, we gathered component consumption data including structured and unstructured data from a spare part management information system in military logistics. We proposed demand forecasting models based on data mining and text mining techniques. The results show that our approach can improve the prediction performance compared to that of existing approaches.