Improving the inventory management of parts using an order-based performance approach at Fluiten

The inventory management at Fluiten, a company manufacturing mechanical seals in Pero, Italy, has been analysed and improved. Using a stepwise classification method, 19,000 parts have been grouped based on their importance, predictability of demand and expected demand during lead time. Their lead time demand has been approximated using a Normal, Gamma or Generalised Negative Binomial distribution. Two methods to incorporate order-based performance in the individual target fill rates of the parts have been tested on a small scale. The best-performing method used the number of orders a certain stock-keeping unit was part of to represent its criticality and has been applied to the full scale. The proposed management model has been tested using a simulation on a daily basis in Python, which showed an improvement in order fill rate of 18% while reducing on-hand inventory value. The research has proven to provide a funded inventory management proposal, a prototype for an inventory management tool and new insights regarding backorder performance.