A predictive analytics approach for demand forecasting in the process industry

Abstract Anticipating demand changes is critical in the process industry with high capacity utilization. The developed solution for demand forecasting is based on a new predictive analytics approach and sophisticated information technology, which allows combining company data and economic information specifically matched to the market environment of product segments. The approach systematically selects the best information for a business outlook. Using the data from the global chemical company BASF, the first empirical results show that this approach significantly outperforms statistical methods based on historical demand data only.