Industrial systems, including wind turbines, manufacturing equipment, aerospace structures, and energy infrastructures, form the backbone of modern society and sustainable development. These complex systems often operate under harsh and dynamic conditions, making them vulnerable to degradation and unexpected failures. Traditional maintenance approaches, relying on fixed schedules or reactive measures, are increasingly inadequate in terms of cost, adaptability, and operational efficiency. The overarching goals of advanced Prognostics and Health Management (PHM) are to ensure productibility, quality, availability, and safety across the lifecycle of these critical assets.

Recent advances in artificial intelligence (AI), multi-sensor fusion, signal processing, and digital twin (DT) technology have driven a paradigm shift in PHM. Data-driven and hybrid modeling now enable real-time fault detection, remaining useful life (RUL) prediction, and intelligent maintenance decisions. However, purely data-driven methods often lack physical consistency, robustness, interpretability, and reliable uncertainty quantification, limiting their adaptability and applicability in safety-critical industrial settings. In parallel, physics-based modeling offers strong interpretability and extrapolation but faces challenges in scalability, complexity, and real-time deployment.

My research aims to bridge physics-based models and modern AI to develop physics-integrated, uncertainty-aware, and explainable PHM frameworks that emphasize robustness, interpretability, and practical applicability. I focus on DT–enabled PHM, physics-informed neural networks (PINN), hybrid modeling, and trustworthy AI, integrating expertise in signal processing, probabilistic modeling, and deep learning.

If you are a KIT student interested in writing a Bachelor’s or Master’s thesis in PHM, please feel free to email me.
If you are from industry and interested in collaborating in any of the following areas, I also welcome your contact:

• Condition Monitoring
• Health Assessment and Anomaly Detection
• Remaining Useful Life (RUL) prediction
• Data Analytics and Feature Extraction
• Uncertainty-Aware Decision-Making
• Industrial AI and Transfer Learning
• Physics-based Modeling and Digital Twin
• Large Language Model (LLM) and Foundation Model
• Intelligent Manufacturing and Energy System Management