LAUSR

Abstract Inventory optimization of high-value spare parts may generate a significant reduction of cost to allow a better allocation of resources in maintenance management. Sherbrooke's METRIC (Multi Echelon Technique for Recoverable Item Control) is the most common method to define an overall optimization process adopting a system-approach. Its main assumption consists of adopting a Poisson distribution to describe the demand pattern of the items. However, many studies proved that in high-availability systems, high-cost spare parts often follow irregular demand patterns, with very frequent zero-demand values. For this purpose, we propose an innovative model for a single site, the ZIP-METRIC, to take advantage of a distribution yet widely adopted in the healthcare and biological sciences, i.e. the Zero-Inflated Poisson. A case study of 1745 items of a European airline fleet demonstrates the model effectiveness, confirming that the ZIP-METRIC outperforms the traditional Poisson-based approach.