AI is already setting the standards for the next nuclear reactors

Last December, during the first International Symposium on Artificial Intelligence and Nuclear Energy, the IAEA announced the signing of an agreement with Atomic Canyonan American technology company that develops AI-based solutions for the nuclear energy sector. Your manager, Trey Lauderdale had taken a big step by creating a program capable of processing the mountains of paper documents at the Diablo Canyon nuclear power plant in California. Building on this experience, also through an agreement with the Idaho National Laboratory to begin designing industry standards that test the capability of AI software for nuclear projects, similar to updates from ChatGPT or Perplexity, made his effort applicable on a global scale. Last July, Lauderdale partnered with Oak Ridge National Laboratory (ORNL) to streamline the nuclear licensing process using artificial intelligence and review license applications. The goal was to use high-performance computing to create high-fidelity simulations that ensure design security, while speeding up licensing using artificial intelligence and automating aspects of the review process.

Global standards

Now the new agreement with the United Nations nuclear agency gives that company the power to start cataloging IAEA data and laying the foundation for global standards for how AI software can be used in industry. “We’re going to start building proofs of concept and models together, and we’re going to build a framework on AI opportunities and use cases,” explained Lauderdale, CEO of Vienna-based Atomic Canyon, while waiting for details to be decided on what international standards or guidelines might look like. In the United States, Atomic Canyon began moving forward earlier this year with a project supported by the Institute of Nuclear Power Operators, the Nuclear Energy Institute and the Electric Power Research Institute to create a virtual assistant for nuclear workers.

Atomic Canyon isn’t the only company applying AI to nuclear power. Last month, the nuclear giant Westinghouse He unveiled new software he is designing with Google to calculate ways to reduce the cost of key reactor components by millions of dollars. The Nuclear Company, a startup that aims to build reactor farms, also announced a deal with software giant Palantir to create the software equivalent of what the companies described as the “Iron Man Suit”, able to quickly obtain regulatory details and plans for engineers tasked with building new nuclear power plants.

Pending tasks

“There is a lot of work to be done in the nuclear power sector, from a management perspective, at every link in the nuclear supply chain, from how we design reactors to how we license them, how we regulate, how we do environmental assessments, how we build them and how we maintain them,” Lauderdale says.

The IAEA statute grants specific programming powers to the country with the largest fleet of nuclear reactors, the United States. The other 30 countries with nuclear energy have largely aligned their regulations and approaches with those standardized in Washington. But the rapid deployment of new reactors in China appears to be poised to put the Asian country ahead of the United States and Beijing would gain influence over other countries’ atomic energy programs. But if AI turns out to have as big an impact on nuclear operations as Lauderdale predicts, the rules of this balancing game will change.