The U.S. Department of Energy recently announced funding for the first projects selected under its new Genesis Mission, a national initiative dedicated to harnessing AI to advance energy, engineering, and national security. Almost 300 projects were selected, including a submission from Dr. Yang Liu, Director of the Texas A&M Institute of Data Science’s (TAMIDS) Generative AI for Science and Engineering (GAISE) Lab. His team’s project, “SHIELD: Secure Human-in-the-loop Intelligence for Engineering and Licensing Deployment of Advanced Nuclear Systems,” optimizes workflows throughout the design analysis, safety evaluation, and licensing processes. Dr. Shuiwang Ji, professor in the Department of Computer Science and Engineering and co-director of the GAISE Lab, serves as co-principal investigator, leading the adaptation of open-source AI models so partner companies can run SHIELD securely on their own systems.
Today, engineers move information by hand between design tools, safety analysis codes, and licensing documents, checking formats and reconciling results at each step. SHIELD is an AI-powered workflow platform that connects these steps into a single, governed process. The system retrieves relevant technical references, prepares and runs simulations, diagnoses errors, and assembles results into review-ready reports. Every action is logged so reviewers can trace exactly where each result came from, and engineers approve every consequential step. The goal is to cut repetitive labor in design and licensing work while keeping humans in charge of all major decisions.
Testing SHIELD on safety analyses with Oklo and Westinghouse
This project will first be tested on the safety analyses of sodium fast reactors at Oklo as well as thermal-structural analyses at Westinghouse, two major nuclear technology companies. Dr. Liu’s team anticipates SHIELD cutting the time spent on repetitive tasks by at least 50% and the overall time needed to complete safety analyses and licensing work by roughly 30%.
“Advanced reactors will deploy faster only if the engineering and licensing work behind them speeds up. SHIELD automates the repetitive steps in that process while engineers keep authority over every decision that matters. GAISE made this project possible by uniting AI and nuclear engineering expertise, and the Genesis Mission gives us the chance to prove the approach with partners like Westinghouse and Oklo,” said Dr. Liu.
At TAMIDS, the GAISE Lab aims to advance scientific discovery and engineering through state-of-the-art generative AI methods, with an emphasis on nuclear energy and materials science. GAISE was instrumental in securing the Genesis Mission award by providing the resources and platform needed to develop the initial ideas. GAISE also supported PhD candidate Zaid Abulawi from Texas A&M’s Department of Nuclear Engineering as a TAMIDS Fellow, who worked as an intern at Idaho National Lab supporting another Genesis Mission Project, Prometheus, this summer. We would like to congratulate Dr. Liu and his team on being selected for the Genesis Mission grant. Their work with GAISE and the larger Texas A&M community is invaluable and demonstrates TAMIDS’ commitment to collaborative research.