This study explores using Artificial Intelligence (AI) for structural health monitoring of nuclear power plants. Mechanical systems like pipelines deteriorate due to erosion and corrosion. Vibrations from routine industrial operations can lead to material fatigue and pipeline cracks. Subsequent leakage may trigger catastrophes like nuclear reactor coolant loss. This research proposes an AI-powered condition monitoring system that can identify degraded locations and assess severity before cracks develop. A digital replica of the system, termed a digital twin, is created to collect vital sensor data. A methodological approach is formulated to extract degradation indicators and train the AI. An intelligent neural network then identifies and categorizes degradation as minor, moderate, or severe.

Speaker: Dr. Saran Bodda

Saran Bodda is a Research Faculty in the Center for Nuclear Energy Facilities and Structures (CNEFS) at NC State. He earned a Ph.D. in Civil Engineering with a major in Structural Engineering and Mechanics and a minor in Statistics, along with an M.S. in Civil Engineering from NC State University. His Ph.D. work was recognized with the Shibata Early Career Award at SMiRT 25 Conference, 2019. His undergraduate studies were completed at the Indian Institute of Technology (IIT) Guwahati, India. Dr. Bodda’s research interests encompass uncertainty quantification using statistical approaches, risk- informed validation, probabilistic risk assessment (PRA) for external hazards, AI-based approaches for structural health monitoring, development of advanced algorithms for PRA, Bayesian approaches to update probabilistic seismic hazard curves, and coupled analysis of primary-secondary systems.