Innovation in Scientific Knowledge Based on Forecasting Assessment

Abstract

This chapter presents a study of forecasting methods applicable to the spare parts demand faced by an automotive company that maintains a share of nearly 25% of the automotive market and sells approximately 13,000 parts per year. These parts are characterized by having intermittent demand and, in some cases, low demand, which makes it difficult for such companies to perform well and to obtain accurate forecasts. Therefore, this chapter includes a study of methods such as the Croston, Syntetos and Boylan, and Teunter methods, which are known to resolve these issues. Furthermore, the rolling Grey method is included, which is usually used in environments with short historical series and great uncertainty. In this study, traditional methods of prognosis, such as moving averages, exponential smoothing, and exponential smoothing with tendency and seasonality, are not neglected.