Application of a modern statistical forecasting technique to a materials management decision support system

The ability to manage inventory levels in a cost effective and efficient manner is of vital importance to companies, particularly those in the fast moving consumer goods industries. Crucial to the control of inventory for many of these companies is materials management decision support software. Mercia Software are in the process of upgrading their APL DOS based system, MerciaLincs PC, to a Windows 95 based client/server system MerciaLincs Client/Server, which is being developed in Visual C++. Central to any inventory control system is the statistical forecast. As part of the overall upgrade of the system, Mercia are incorporating new forecasting techniques based on Bayesian Learning and the dynamic linear model (DLM). This modem technique has a number of advantages over existing techniques. The DLM has important advantages over the existing methods. It is, however, more complicated and is, therefore, more difficult to implement in this environment.