AI in Nuclear Research at LSU

Two research teams from Louisiana State University (LSU) are set to collaborate with the U.S. Department of Energy’s Genesis Mission, a significant national endeavor aimed at integrating artificial intelligence into nuclear research. These efforts, in partnership with Oak Ridge National Laboratory and Brookhaven National Laboratory, are designed to tackle contemporary challenges facing physicists.

Led by LSU Department of Physics and Astronomy professors Jeffery Blackmon and Phillip Sprunger, these teams were among 278 projects nationwide selected to receive funding. Collaborating with faculty from the University of Kentucky and scientists from national labs, they submitted two substantial proposals to the U.S. Department of Energy.

Kristina Launey, an associate professor of physics, noted, “This initiative has been compared to the Manhattan Project. So being part of it, it’s amazing. It’s an amazing achievement from LSU, which means what we do, the research here, brings LSU to the national level.”

The Genesis Mission aims to unite leaders from government, industry, academia, and science to further the U.S. Department of Energy’s ambitions in developing AI systems that can address national defense and energy challenges.

Launey elaborated on the mission’s collaborative spirit, stating, “The initiative aims to bring scientists from different areas together, and by doing this, it practically combines the expertise of various fields.” For instance, one project merges nuclear experimental physics with data analysis to enhance the understanding of nuclei.

Blackmon’s project aims to shorten the duration needed for data analysis following nuclear decay or collision events. When such an event occurs, valuable data is emitted in the form of neutrons and gamma rays, which can provide insights into crucial nuclear processes applicable to real-world challenges like cancer treatment and defense strategies.

“As humans, it can be hard for us to think in more than a few dimensions at a time,” Blackmon remarked. “We often see our graduate students struggle for months, even years, to analyze data sets that might have been captured in a few hours.” The integration of AI is expected to streamline the organization and analysis of experimental data, significantly reducing the time required to produce actionable results.

By leveraging over a decade’s worth of measured nuclear event data from Oak Ridge, Blackmon’s team intends to create AI systems that can interpret this complex information. “We hope AI can help us use the correlations between the many measured parameters to greatly speed the pipeline of data analysis and provide new insights into the properties of nuclei,” he added.

Similarly, Sprunger’s project focuses on enhancing energy processes nationally through AI applications. As the interim director of the Louisiana Light Source, he oversees a synchrotron responsible for generating exceptional visible light, X-ray, and UV beams for experimental purposes.

“We use that light to then probe materials. It is that light source, or synchrotron, that produces the light that enables us to understand materials that you can’t otherwise do with any other kinds of probes,” Sprunger explained.

Managing a synchrotron presents numerous challenges, but Sprunger believes AI can facilitate smoother operations. His team, in partnership with Brookhaven National Laboratory, is investigating AI solutions to predict potential machine failures and maintain equipment functionality.

Timur Shaftan, the director of the National Synchrotron Light Source II, highlighted that this collaboration will unlock future opportunities in accelerator science and engineering. “Rapidly-developing AI tools offer revolutionary opportunities to advance accelerator facility operations, such as those at NSLS-II or Louisiana Light Source, and for new insights into accelerator physics,” he stated.

While the implementation of AI in this substantial research avenue may appear daunting, Sprunger views it as a necessary evolution. “It doesn’t matter if we’re trying to understand where the electrons going at 100 times the speed of light are flying and how we can control those better, or if we’re counting fish out in the gulf,” he remarked. He believes the principles derived from AI applications are universally applicable across various fields.

Both teams’ initiatives are in alignment with the U.S. Department of Energy’s objective to incorporate AI into experimentation and reflect LSU’s aspirations to rank among the top 50 research institutions in the country. Chancellor Jim Dalton stated, “These projects capture exactly what LSU aspires to be as we drive toward the ranks of the nation’s Top 50 research universities — a place where our faculty work shoulder to shoulder with the national laboratories on the defining scientific challenges of our time.”

“From our own synchrotron in Baton Rouge to the instruments developed by Oak Ridge and Brookhaven, LSU is bringing artificial intelligence to the front lines of discovery,” Dalton concluded.