TL;DR — Key Takeaways
Texas A&M landed four projects in DOE’s Genesis Mission, spanning critical minerals, particle accelerator operations and nuclear reactor safety.** Two teams will use multimodal AI to improve mineral exploration**, combining geological, geochemical, microbial, hydrological and satellite data.** Agentic digital twins could help scientists maximize costly beam time**by recommending equipment settings, monitoring sensors and detecting problems.
Four Texas A&M research teams have been selected for the U.S. Department of Energy’s Genesis Mission, where they will develop and test AI systems for critical mineral research, particle accelerator operations and nuclear reactor safety analysis.
The projects are part of the Genesis Mission’s recently announced initial research portfolio, drawn from what DOE called its largest response ever to a funding opportunity. This includes 278 projects involving 342 institutions, with universities leading 168 of them. The work is organized around 26 national science and technology challenges DOE identified in February, including fusion energy, quantum computing, chip design and advanced manufacturing. DOE said the teams will gain access to the initiative’s shared platform of AI agent frameworks, advanced models, industry software and high performance computing resources at national laboratories and partner facilities.
That infrastructure is central to the Genesis Mission’s premise of connecting AI with the scientific data, supercomputers and research facilities scientists already use to help them move more quickly from experiments to results. DOE has set a goal of doubling the productivity and impact of U.S. research within a decade through faster data analysis, better use of scarce computing and laboratory resources, and greater coordination across research institutions.
At Texas A&M, two of the four projects will apply multimodal AI to critical minerals. A team led by civil and environmental engineering professor Kung-Hui Chu will work with Lawrence Berkeley National Laboratory to combine microbial, hydrological and geochemical data to build AI tools for locating mineral deposits and studying biological methods of recovery. A second team, led by geology and geophysics professor Nicholas Perez, will analyze geological, geochemical, geophysical and satellite data to identify patterns associated with rare earth deposits. The researchers will study Texas, the Colorado Mineral Belt and other parts of the Southwest in collaboration with Pacific Northwest National Laboratory.
Another project, led by Texas A&M assistant professor Jonas Karthein in collaboration with MIT’s Laboratory for Nuclear Science, will test agentic digital twins at precision nuclear physics facilities, where data is limited and beam time can cost hundreds to more than $10,000 per hour. The system will learn how equipment behaves, recommend settings, monitor sensors and flag problems, helping researchers make better use of scarce experimental time. The team plans a nine-month proof of concept at Texas A&M and MIT, with possible expansion to larger facilities in a second phase.
A fourth project will explore AI support for reactor safety and licensing. Headed by nuclear engineering professor Yang Liu, the SHIELD system will help run reactor models and simulations and prepare documents for regulatory review. The team will test it on a sodium-cooled reactor design and conventional large light-water reactors.
DOE cautioned that the selections are subject to award negotiations and do not yet guarantee funding. But projects like these could give the Genesis Mission concrete measures of progress, including better mineral exploration decisions, more efficient use of beam time and faster preparation of reactor safety analyses. Their results could also help show where AI can make the greatest impact across scientific research.