Abstract:

Industrial equipment is usually repairable. Reliability of repairable systems depends on maintenance actions, and is characterized by mean cumulative function or failure intensity function of failure processes. Evolution trend of the failure intensity function is called the system-level failure pattern in the literature. The failure pattern plays an important role in preventive maintenance decision analysis. The purpose of this paper is to develop a robust non-parametric method to estimate the failure intensity function based on the failure processes observed from a single system or several independent and identical systems. After examining a large number of real-world datasets published in the literature, it is found that the intensity function of repairable systems is typically of roller-coaster curve shape. This type of failure pattern is illustrated and the possible causes are discussed. The usefulness for planning the preventive maintenance and determining the retirement time of repairable products is also illustrated.