Transfer Learning for Prognostics and Health Management: Advances, Challenges, and Opportunities

Abstract

As failure data is usually scarce in practice upon preventive maintenance strategy in prognostics and health management (PHM) domain, transfer learning provides a fundamental solution to enhance generalization of data-driven methods. In this paper, we briefly discuss general idea and advances of various transfer learning techniques for PHM domain, including domain adaptation, domain generalization, federated learning, and knowledge driven transfer learning. Based on the observations from state of the art, we provide extensive discussions on possible challenges and opportunities of transfer learning for PHM domain to direct future development. Conflict of Interest Statement The authors declare no conflicts of interest.