Optimization of Time Delay based Preventive Maintenance using Markov Decision Process

In operating Indonesian Navy Vessels (KRI), users often find some indicators leading to the prediction that the system experiences a decline in performance or a breakdown, and requires repair at great expense. To support the reliability-based maintenance system, an analysis is required to determine the exact breakdown rate and the state of the system based on time delay. In time-delay-based maintenance, before experiencing a breakdown, the system will show a decline in performance. However, time-delay-based maintenance is difficult to apply in the field since it requires the appropriate data to form the model. In this study, time-delay-based maintenance is applied in combination with policy patterns in the operation and observation, using Markov Decision Process. By applying time-delay-based preventive policy, it can be concluded that the policy pattern 1, 2, and 3 in this study can minimize operational and maintenance expenses when KRI experiences a breakdown in general.