Condition-based maintenance optimization considering improving prediction accuracy

Condition-based maintenance (CBM) aims to reduce maintenance cost and improve equipment reliability by effectively utilizing condition monitoring and prediction information. It is observed that the prediction accuracy often improves with the increase of the age of the component. In this research, we develop a method to quantify the remaining life prediction uncertainty considering the prediction accuracy improvement, and an effective CBM optimization approach to optimize the maintenance schedule. Any type of prognostics methods can be used, including data-driven methods, model-based methods and integrated methods, as long as the prediction method can produce the predicted failure time distribution at any given inspection point. Furthermore, we develop a numerical method to accurately and efficiently evaluate the cost of the CBM policy. The proposed approach is demonstrated using vibration monitoring data collected from pump bearings in the field as well as simulated degradation data. The proposed policy is compared with two benchmark maintenance policies and is found to be more effective.