- This event has passed.
NEEP Seminar: Jamie Coble, University of Tennessee, Knoxville
April 1 @ 12:00 PM – 1:00 PM
Tuesday, April 1
12:00 – 1:00pm
ERB 106
Remote Participation: Please contact office@ep.wisc.edu for the Zoom link.
Title: Enhancing Economic Operations and Maintenance of Nuclear Power Plants through Applied AI/ML
Abstract: Nuclear power plants continue to be faced with a challenging economic reality that has led to the premature closure of many plants in the US fleet. Recent interest in clean, reliable nuclear power (primarily driven by the tech industry’s increasing energy demand and carbon-neutral commitments) brings opportunities to re-commission shuttered plants and build new plants, expanding the US nuclear footprint. Even with this new push for nuclear power, economic considerations remain critical to success. Controlling the day-to-day operations and maintenance (O&M) costs associated with nuclear power is one of the primary avenues for improving the economic outlook for the current and future nuclear industry. The current approach to O&M relies primarily on license-based periodic inspection and maintenance activities scheduled to preclude in-service degradation and failure. While effective at maintaining plant safety, reliability, and availability, this approach often leads to unnecessary and costly activities.
Artificial Intelligence (AI) and Machine Learning (ML) offer transformative potential to modernize nuclear O&M. Predictive maintenance programs can leverage AI/ML algorithms to continuously monitor equipment health, detect anomalies, and predict remaining useful life (RUL) of critical components. Additionally, ML-powered anomaly detection systems provide early warning of abnormal conditions, enhancing operational reliability and supporting real-time risk management. These predictive insights can also be leveraged to optimize maintenance scheduling and resource allocation. This seminar will overview research efforts in online equipment condition assessment, fault detection and diagnostics, and failure prognostics for active equipment in nuclear power plants and methods to integrate this knowledge into robust decision making, including maintenance planning and integration with Digital Twins for O&M support.
Speaker: Jamie Coble, University of Tennessee, Knoxville
Bio: Dr. Jamie Baalis Coble is a Professor, Southern Company Faculty Fellow, and Associate Department Head in the Nuclear Engineering department at the University of Tennessee-Knoxville. Dr. Coble’s expertise is primarily in data analytics, machine learning, and artificial intelligence approaches for equipment condition assessment, process and system monitoring, anomaly detection and diagnosis, failure prognosis, and integrated decision making. Her research interests expand on past work in nuclear system monitoring and prognostics to incorporate system monitoring and remaining useful life estimates into risk assessment, operations and maintenance planning, optimal control algorithms, and digital twin simulations, as well as other applications of machine learning and AI in support of nuclear power operations. Prior to joining the UT faculty, she worked in the Applied Physics group at Pacific Northwest National Laboratory. Dr. Coble is currently pursuing research in prognostics and health management for active components and systems; advanced control strategies for integration of small modular reactors and multi-modular reactors with deep renewable penetration; methods to improve decision making in O&M of nuclear power facilities; use of GenAI in nuclear operations and regulatory contexts; and applications of digital twins and AI for nuclear security.
This seminar is presented by the Institute for Nuclear Energy Systems and the Nuclear Engineering & Engineering Physics Department.
