Skip to content

NMSU collaborates on NRC project to develop AI for nuclear plant safety

Release Date: 21 Nov 2022
NMSU collaborates on NRC project to develop AI for nuclear plant safety

"The China Syndrome" is a film and also a fictional scenario in which a nuclear reactor core could melt down through the basement of a power plant, contaminating soil and water.

The nightmare scenario became a reality in 2011 at the Fukushima Nuclear Reactor in Japan. A major earthquake, followed by a tsunami, disabled the power plant causing a meltdown of three nuclear cores, which forced a 200-square-mile evacuation. Nuclear regulators elevated the accident to level seven – the highest level on the scale created by the International Atomic Energy Agency – placing it in the same category as the deadly Chernobyl nuclear accident in 1986.

Most nuclear accidents or near misses are not so dramatic. A project funded by the U.S. Nuclear Regulatory Commission will help researchers create a framework to predict maintenance and operations issues and provide potential solutions before they turn into nuclear accidents.

"If a nuclear accident happens, what should the operator do?" asked Son Tran, New Mexico State University computer science professor and co-principal investigator on a grant from the U.S. Nuclear Regulatory Commission. "I expect to develop a prototype of the system that can take in different scenarios from the Nuclear Regulatory Commission and, through artificial intelligence, provide information to help the operators make decisions."

NMSU will be collaborating with North Carolina State University along with various companies and agencies on a project titled “Trustworthiness of Digital-Twin-based Automation Technology in Nuclear Power Plant Operation.” The $500,000 project seeks to establish a technical basis for the development and demonstration of a trustworthy automated technology framework that can be used to assess the condition of nuclear power plants and help operators make maintenance and operations decisions to prevent accidents or mitigate impact.

Tran, who has more than 20 years of experience and has developed several state-of-the art automated planning systems, and Nam Dinh, professor in North Carolina State University's Department of Nuclear Engineering, who has more than 20 years of research, development and engineering experience in areas of nuclear reactor thermal hydraulics and nuclear power safety, are co-principal investigators on the three-year project.

"Making sure that the tools developed for the control of a nuclear reactors are trustworthy is important," Tran said. "In this case, it's the digital twin (DT) technology that helps operators in the decision process. This allows for the operators to work together with the tools in making the best decisions to control the reactors, especially in emergency situations." 

Tran's research in the areas of automatic reasoning dovetails with Dinh's research in the engineering of nuclear reactor systems. Methods and lessons learned during the project will be instrumental in creating a framework for the development and deployment of DT technology, advanced modeling approaches, intelligent automation systems and unattended reactor systems to improve the reliability and economic viability of existing and next-generation commercial reactor designs.

"AI would only be playing the role of a supporter, but supporter in the sense that we provide operators with what we call informed decisions," Tran said. "We believe if we can develop a system that can predict the state of the nuclear reactor, given specific circumstances, then the automated technology can present data-driven options to the human operator. The human operator will decide on what to do next."

Once Tran and Dinh establish the technical basis of the framework and demonstrate its feasibility, the next step would be to deploy the technology into nuclear reactors.


adding all to cart
File added to media cart.