Adjusting condition monitoring quality for condition-based maintenance optimization of deteriorating systems

International audienceCondition monitoring (CM) is an essential part in a condition-based maintenance (CBM) programas it can provides useful information about the system health for maintenance decision-making. The quality ofCM information depends closely on implemented inspection and instrument technologies, and eventually oninvested costs. While numerous works in the literature have considered problems related to CM quality (e.g.,imperfect inspection models, detection and localization techniques, etc.), few of them focus on adjusting CMquality for CBM optimization. In this paper, assuming that the quality of CM information can be controlledby adjusting inspection costs, we investigate how such adjustment can contribute to save on the total costof a CBM program. The study is done for a single-unit deteriorating system subject to random failures andto periodic inspection. We model the quality of the degradation information on the inspected system usingthe idea that its variance is a decreasing function of inspection cost (i.e., a more accurate CM informationrequires a more sophisticated inspection, hence a higher cost). A new inspection and replacement strategywhose decisions are adapted to this variance is proposed and compared to a non-adapting classical one. Thecomparison results under various configurations of system characteristics and maintenance costs allow us toconclude on the interest of CM adjustment in CBM optimization