Integrating AIs into model-based intelligent maintenance

About this project

Introducing AI technologies into critical equipment to assist next generation intelligent maintenance is a hot research topic. Moreover, model-based dependability analysis is an advanced technology integrating system dependability-based development, analysis, and verification. we have combined the traditional maintenance technology i.e., RCM (reliability centred maintenance) with the state-of-the-art model-based dependability analysis methods and have proposed a model-based RCM analysis framework. The benefit of this combination is that it automates the RCM analysis process by automatically obtaining important failure-based analysis artifacts such as FT and FMECA. According to the decision of FMECA, the technical maintenance scheme for a certain system can be decided. For example, for some critical equipment, a predictive maintenance scheme (e.g., condition-based maintenance) involving fault monitoring and diagnosis must be applied to ensure safety or dependability concerns. To further enhance the analytical performance of the framework, in this project, we effort to integrate several intelligent maintenance models e.g., intelligent feature learning and fault recognition, intelligent fault monitoring and diagnosis, intelligent remote maintenance, intelligent health status assessment and residual life prediction into the framework, and to explore a new intelligent maintenance model. The research topic is open so that students can either choose one of described intelligent maintenance models or propose an interested and related topic.

In any case, the following key research questions need to be explored:

1. What makes a field a good choice for an intelligent approach?
2. What are the potential limitations of selected intelligent model in the field of system maintenance?
3. Which intelligent model is best suited for the specific application field (e.g., fault diagnosis) of the maintenance?