The goal of jet engine health management is to predict failures accurately and reliably at an early stage, allowing for maintenance to be performed at the right time and place. Effective Engine Health Monitoring (EHM) requires early fault detection with no missed detection and an acceptable level of false alarms. FD&I capabilities have been built with sensors placed all over the engine, and machine learning and artificial intelligence can be used to extract features from sensor signals and identify fault signatures. However, false alarms can be an issue when detection is pushed too early, and some failure modes have multiple effects, making reliable detection difficult. To address these issues, the proposed project objective is to study and apply machine learning and artificial intelligence in support of early, accurate, and reliable fault detection and isolation with 100% recall rate and acceptable precision. The approach will use unsupervised learning to reveal causality, and physics-informed machine learning will be employed to incorporate the underlying physics and engineering principles into the ML tool. The proposal is to use Bayesian Machine Learning to develop a generic framework for incorporating physics into machine learning. The goal of the project is to develop a tool that is reliable, explainable, and reduces the problem associated with the rarity of faults.