Reducing risks in spare parts service contracts with a long downtime constraint

This article investigates spare parts service contracts for capital goods. We consider a single-item, single-location inventory system that serves one customer with multiple machines. During the contract execution phase, the true demand rate is observed. It can differ from the estimated demand rate because of two factors: increased demand variation in finite horizon settings and a shift in the mean utilization of the machines by the user during the contract. When the true demand rate is higher than the estimated demand rate, the Original Equipment Manufacturer (OEM) is faced with higher-than-expected costs for the execution of the contract, and the asset user is generally faced with a higher number of extreme long downtime events. Therefore, we introduce the flexible-time contract, which ends after a predetermined number of demands. Using a Markov decision process, we prove that a state-dependent base stock policy is optimal under a flexible-time contract. Using simulation, we compare the flexible-time contract with the standard fixed-time contract. Our results show that the flexible-time contract reduces the costs for the OEM by up to 35% and prevents not meeting the agreed-on service level. We obtain similar results in a multi-item setting.