Spare Parts Management: an Optimized Service Level-based Model for Inventory Control

Spare parts optimization can significantly reduce inventory costs while avoiding compromising equipment availability. However, the distinctive trend of spare parts demand makes it difficult to establish the optimal stock level. Many literature studies address and classify the different criticalities of spare parts, but there is a lack of workable and structured tools related to the effective management of these items. For this reason, this paper proposes a practical approach to compute the stock quantities by setting a target service level, which is pre-defined by considering some critical factors, ad-hoc tuned to the company’s needs. The presented approach is composed of three sequential steps: the first aims at classifying the demand time series behavior as intermittent, lumpy, erratic, or smooth; then, forecasting methods are applied to predict consumption events, and the forecasting accuracy metrics are compared to identify the optimal one, per each item; lastly, reorder strategies are selected according to the results of the previous steps and reorder events are triggered by the probability of achieving the target service level for the next time bucket. This model has been validated in a pharmaceutical manufacturing facility, leading to excellent results in reducing stockholding costs.