Systematic Review Maps Physics-Informed Machine Learning in Prognostics and Health Management

A systematic literature review published on arXiv (ID: 2608.10047) examines the integration of Physics-Informed Machine Learning (PIML) into Prognostics and Health Management (PHM). The review, which analyzes 212 studies, proposes a four-class classification scheme for PIML approaches: observational bias, inductive bias, learning bias, and hybrid approaches. It further categorizes the studies by PHM task. The motivation behind the review is the recognized limitations of purely data-driven machine learning models, which often suffer from poor generalization, lack of causal inference, and limited interpretability. By incorporating prior physical knowledge into the machine learning pipeline, PIML aims to mitigate these issues, leading to growing interest in its application to PHM. The review is intended to provide a structured overview of how PIML is being leveraged in the field, potentially guiding future research and applications. The paper is authored by researchers and was announced as a cross-type submission on arXiv. The review does not specify the authors' names or affiliations in the provided content. The findings are expected to be of interest to researchers and practitioners in industrial reliability, maintenance, and machine learning.