Research on Performance Evaluation Metrics and Test Methodology of Predictive Maintenance

Predictive maintenance is an important scenario for the application of artificial intelligence technology in intelligent manufacturing, guiding enterprises to carry out equipment operation and maintenance through fault diagnosis and prediction, which can effectively reduce equipment downtime and improve production efficiency. However, the application of predictive maintenance technology has been in a bottleneck for a long time, and the fundamental reason is the lack of predictive maintenance performance evaluation index system, which makes it difficult for users to judge the effect of the application of this technology and thus lack confidence in it. Some scholars have proposed a preliminary evaluation index system, but the lack of testing technology to support the quantitative indexes makes it difficult to promote and implement the system. In this paper, by reviewing the related research on predictive maintenance performance assessment index system, we innovatively propose an on-off line assessment index system that can be standardized, and test some quantitative indexes through two different testing methods, namely, online testing and offline testing, in order to verify the applicability of the assessment index system.