AI for Nuclear Safety: Predicting Remaining Useful Life

Join Muthukumar Ganesan, a scientist at the Indira Gandhi Centre for Atomic Research (IGCAR), as he explores the vital role of AI in securing the next generation of nuclear energy. The session focuses on the shift from reactive and preventive maintenance to AI-driven predictive strategies, specifically within the high-stakes environment of Fast Breeder Reactors.

The discussion highlights the development of a robust system to estimate the Remaining Useful Life (RUL) of critical components, such as the primary sodium coolant pump. By utilizing multivariate time series sensor data and virtual modeling, Ganesan demonstrates how machine learning can identify degradation patterns, like bearing wear and shaft misalignment, before they lead to operational failure. The talk emphasizes a "sovereign-by-design" approach to safety, showcasing the use of redundant hardware and diversified deployment strategies to ensure 100% system reliability in safety-critical industrial environments.