Maintenance Decision and Spare Part Selection for Multi-component System

Abstract

Due to the advancement of technology over time higher technology machines are being used in the production and service sectors. Companies suffer great financial losses if these machines stop working due to a breakdown. To avoid these losses maintenance has become increasingly important for companies over time. Condition based maintenance aims to intervene in a system as close to the point of failure as possible using information received from the system. Sensors are used to obtain information about the wear and tear of the machine. However since sensors are costly they are not installed on every machine component but rather on the system. While this reduces costs it also means that we now obtain partial information from the system rather than from each component. In these systems we need to make two types of decisions. The first decision is when to intervene in the system. The second decision is how many spare parts to carry with us once we decide to intervene. We simulated several different experiments for a periodic system composed of identical components and found optimal policies based on the two decisions we made. Our managerial insights indicate that as the number of components in the machine increases the importance of selecting spare parts for the system also increases leading to a tendency to maintain the system as late as possible before the system fails. Moreover in situations where the penalty for maintenance is lower after a failure occurs in optimal policy we maintain later and carry more spare parts during our interventions. © 2024 Elsevier B.V. All rights reserved.