MIT Researcher Develops Autonomous Control Systems for Nuclear Plants

A former U.S. Navy nuclear operator is working to solve one of the industry's most pressing operational challenges: how to run small nuclear reactors without large on-site crews.

Lauren Fortier, a second-year doctoral student in MIT's Department of Nuclear Science and Engineering, is developing autonomous supervisory control systems that could enable remote operation of next-generation nuclear facilities. Her research addresses a fundamental economics problem facing the nuclear industry as it scales down reactor size.

From carrier operations to academic research

Fortier's path to MIT began aboard a U.S. aircraft carrier in the South China Sea, where she supervised nuclear plant operations after earning her undergraduate degree in materials science and engineering from Northwestern University on an ROTC scholarship. Operating a reactor that powered the entire vessel through the ocean, she observed that many plant procedures remained intensely manual—a model that works for large facilities running at full capacity but becomes economically unviable for distributed microreactors.

The Navy's offer to pursue graduate education brought Fortier to MIT, where her master's work developed a supervisory control system using thermal hydraulic simulations. But the research revealed a deeper challenge: existing operational frameworks were designed exclusively for human operators, limiting flexibility in how procedures could be executed.

Building human-machine collaboration

Fortier's doctoral research tackles the question of transitioning to autonomous operations through an integrated approach rather than multiple interlinked systems. Her framework allows humans and machines to share responsibilities strategically, with human intervention delivered only when necessary.

Crucially, Fortier is not using AI-driven machine learning for this automation. Instead, she employs finite state automata—a transparent, event-driven system where every action follows clear if-then logic based on current plant conditions. This approach addresses validation concerns that currently prevent AI deployment in nuclear operations.

"We're not using a data-driven statistical approach like machine learning because we do not yet have the tools to validate the operation of such systems," Fortier explained in an interview first reported by MIT News.

Her research benefits from collaborations across institutions. Sacit Cetiner, her advisor with joint appointments at MIT and Idaho National Laboratory (INL), connected her with Katya Le Blanc, a senior human factors scientist at INL, to develop effective human-machine interfaces. Additional work with Westinghouse during a 2025 summer internship allowed real-world testing of autonomous operations concepts.

Why it matters

Small modular reactors and microreactors represent a potential path to distributed clean energy, but their economic viability depends on dramatically reducing operational staffing requirements. Legacy nuclear plants justify large crews because they operate at 100 percent capacity; small reactors in remote locations cannot support the same cost structure. Fortier's transparent automation framework could enable commercial deployment of these systems while maintaining the rigorous safety standards required for nuclear operations.

Fortier's work earned recognition in the 2025 Innovations in Nuclear Energy Research and Development Student Competition from the Department of Energy's Nuclear Energy University Program. Her next phase involves scaling the supervisory control system beyond individual components to full plant operations.

Details of the research were first reported by MIT News.