Abstract:
Accurate material demand forecasting can lead to significant cost savings, greater competitiveness and improved customer satisfaction. However, more often than not, demand forecasting as a business function is carried out poorly, with forecast accuracy often not significantly better than the naïve forecast. To appropriately address these concerns and satisfy overall business objectives, it's increasingly important to have a holistic strategy to improve demand forecast accuracy through a well thought-out enterprise data strategy, applications of advanced forecasting methods as well as synchronized cross functional business processes. This paper describes data types that are essential to demand forecasting, investigates advanced analytics methods such as ARIMA and survival analysis and discusses the application of these methods for the purpose of fleet sustainment demand forecasting. Lastly, this paper addresses the business process needed to continuously monitor and improve forecast performance.