Cost-effective condition-based maintenance using Markov decision processes

Investigations conducted in several industries indicate that there is no direct relationship between equipment failure and equipment age in the majority of cases. Most failures are caused by events or conditions that occur during component operation and manufacturing processes. Therefore, optimal maintenance decisions should be based on the actual deterioration conditions of the components. Condition-Based Maintenance (CBM) is a methodology that strives to identify incipient faults before they become critical to enable more accurate planning of preventive actions. For the ultimate success of CBM methodology, we must have sound methods for modeling deterioration (the propagation of faulty conditions), the conditions and their effects, and the optimal selection and scheduling of inspections and preventive maintenance actions (the right action at the right time). In this paper, we present a generalized CBM model that can be applied to a wide range of applications. The CBM model includes a stochastic deterioration process, a set of maintenance actions and their effects, and a scheduled inspection policy that identifies the condition of deterioration. Using Markov Decision Processes (MDP), we provide an optimal cost-effective maintenance decision based on the condition revealed at the time of inspection. In addition, we present a procedure for finding optimal inspection schedules.