AI for Smarter, More Powerful, More Efficient Particle Accelerators

Particle accelerators are powerful tools that have revolutionized research in physics, chemistry, materials science, and biology, and have been adopted for applications that improve medicine, national security, and manufacturing. The Multi-Office particle Accelerator Team (MOAT) will deploy artificial intelligence tools to optimize and transform how accelerators are designed and run, making them more efficient and impactful.

The effort is part of the Genesis Mission, a new national initiative led by the Department of Energy to advance AI and accelerate discovery, providing solutions for challenges in science, energy, and national security. A cornerstone of the Genesis Mission is the Transformational AI Models Consortium (ModCon), which will build and deploy self-improving AI models by harnessing DOE’s unique data, facilities, and expertise. MOAT is one of three AI model teams that Lawrence Berkeley National Laboratory (Berkeley Lab) leads or plays a key role in, building on AI expertise in high-performance computing, managing large datasets, and pioneering AI models in partnership with industry.

The project will take advantage of troves of experimental data, simulations, and expertise from across the DOE Office of Science’s accelerators and light sources, which account for half of the agency’s user facilities. MOAT will build and expand tools such as digital twins, intelligent assistants, and advanced AI models to simulate complex accelerator physics and operations. It will also make them platform agnostic — so shared knowledge and improvements can be adopted by national labs, universities, and industrial partners.

“By making intelligent AI tools that continuously learn and work across facilities and fields, we’re accelerating discoveries that particle accelerators can make in key applications such as fundamental physics, fission and fusion energy, advanced materials, and advanced medical technologies,” said Jean-Luc Vay, MOAT’s lead and the head of the Advanced Modeling Program in the Accelerator Technology & Applied Physics (ATAP) Division at Berkeley Lab.

During its initial phase, MOAT will expand on recent successes led by ATAP and collaborators, who deployed an AI “Accelerator Assistant” at Berkeley Lab’s Advanced Light Source (ALS). The powerful particle accelerator speeds electrons around a circular ring, where they emit ultraviolet and X-ray light that hundreds of experiments every year use to study physics, materials science, biology, chemistry, and more. The complex system has hundreds of thousands of variables.

For the first time, researchers demonstrated that an AI system driven by a large language model could autonomously prepare and run a multi-stage physics experiment on a synchrotron light source — and set up the experiment 100 times faster than humans alone. Engineers use natural language to prompt the Accelerator Assistant with their goal. The AI can then find different variables, generate and run code to analyze data and visualize results, and safely interact with the accelerator itself.

“We’re witnessing the emergence of a new layer of scientific infrastructure: AI systems that can interpret intent, plan actions, and safely operate complex instruments,” said Thorsten Hellert, an ATAP staff scientist and the lead author of the paper on the recent milestone work. “This approach can accelerate experiments at the ALS today and form the foundation for connected, AI-enabled facilities across the DOE complex.”

In MOAT’s first phase, researchers are generalizing this framework and testing it across multiple accelerator facilities, demonstrating cross-site interoperability and shared AI infrastructure.

MOAT currently includes collaborators from Argonne, Brookhaven, Fermi, Jefferson, Oak Ridge, and SLAC national laboratories, as well as industry partners.

###

Lawrence Berkeley National Laboratory (Berkeley Lab) is committed to groundbreaking research focused on discovery science and solutions for abundant and reliable energy supplies. The lab’s expertise spans materials, chemistry, physics, biology, earth and environmental science, mathematics, and computing. Researchers from around the world rely on the lab’s world-class scientific facilities for their own pioneering research. Founded in 1931 on the belief that the biggest problems are best addressed by teams, Berkeley Lab and its scientists have been recognized with 17 Nobel Prizes. Berkeley Lab is a multiprogram national laboratory managed by the University of California for the U.S. Department of Energy’s Office of Science.

DOE’s Office of Science is the single largest supporter of basic research in the physical sciences in the United States, and is working to address some of the most pressing challenges of our time. For more information, please visit energy.gov/science.