Advertisement

Energy Department speeds up nuclear regulatory process with AI

The agency called the experiment, which was conducted in collaboration with two national labs, two vendors and the Nuclear Regulatory Commission, a “milestone test case.”
Listen to this article
0:00
Learn more. This feature uses an automated voice, which may result in occasional errors in pronunciation, tone, or sentiment.
An entrance to the James Forrestal Building, the headquarters of the Department of Energy, in Washington, DC on June 22, 2022.

The Department of Energy is touting the results of an AI experiment that it says demonstrated the technology’s ability to streamline the nuclear regulatory process. 

Using AI, a DOE team converted a safety analysis document into a 208-page license application. The timeline for the process typically requires a team of people over the span of four and six weeks — but it took just one day with AI, according to the Energy Department. 

The agency worked in collaboration with the Idaho National Laboratory, Argonne National Laboratory, Microsoft, Everstar and the Nuclear Regulatory Commission. DOE called the effort a “milestone test case.” 

“It has the potential to transform how industry prepares its regulatory submissions and deploys nuclear energy while upholding the highest standards of safety and compliance,” Rian Bahran, deputy assistant secretary for nuclear reactors at DOE, said in the Thursday announcement of the partnership. 

Advertisement

The intersection of AI and the nuclear energy industry is one of the core focuses of the Energy Department-led Genesis Mission. The agency identified 26 challenges last month for multidisciplinary teams to rally around and tackle, with more than a third centered on the nuclear industry. 

While Genesis Mission teams are still being set up, stakeholders have already been working to shorten the historically long development timelines, including through the DOE’s latest experiment. The 26 challenges also outline other goals, like reducing operational costs for reactors and lessening nuclear cleanup liability by using the latest AI technologies throughout the lifecycle of workflows. 

As AI becomes more embedded into processes, workers will be tasked with ensuring the technology produces accurate outputs and remains an effective tool. 

IT leaders at DOE labs have emphasized the need to manage expectations amid immense AI hype. Tech officials at NRC have echoed the sentiment, comparing the implementation of AI to needing a designated driver after a night out. 

“The subject matter experts who are reviewing what is going on; they’re there with you the whole time,” Nicholas Buggs, chief enterprise architect within the NRC’s Office of the Chief Information Officer, said during a symposium earlier this month. “We’re always validating, always looking at the outputs to make sure that that’s exactly what we expect it to do.”

Advertisement

Latest Podcasts