Smarter inventory control of spare parts can result in savings of 50%

This way both the number of parts in stock and the waiting time for spare parts can be reduced, with theoretical savings of up to 50%. This is possible thanks to fundamental mathematical models developed by PhD candidate Bram Kranenburg MSc. With his research Kranenburg hopes to obtain a doctorate from the Technical University Eindhoven (TU/e) on Thursday 23 November.

Big Business

The storage of spare parts is big business in the Netherlands, involving billions of euros every year. Every branch of industry or service that works with complex machinery needs spare parts. Just think of electronics, hospitals, industrial machinery, and the car industry. One small, defective part can put a complete machine out of operation for quite some time. That is why there have always been strict requirements for stocking and distributing spare parts.

Pooling storage facilities

A great deal of research has already been done to optimize the entire logistic process. Still, inventory control is usually done separately for each warehouse. ASML approached the TU/e to find out if there was not a smarter way to do this and this question became a central theme in Kranenburg’s PhD research. Kranenburg: “The crux of my model is the pooling of different warehouses. If a local warehouse does not have a certain part in stock, it can contact another local warehouse instead of the central warehouse. If you want to do this on a structural basis, there is much to be won by planning your inventory control around this. But if you want to do this right, it becomes very complex mathematically to work this all out. That is the problem I worked on in my PhD research and ASML has been able to implement my model and algorithms right away.”

Universally applicable

The models are very relevant to everyday practice and can lead to great savings. Kranenburg: “Some theoretical data sets yielded cost reductions of up to 50%!” In his research Kranenburg further worked on models for integrated inventory control for different machines and for different groups of customers. All models are universally applicable. In his next job as consultant with CQM Kranenburg will strive to further increase the applicability of his research.