Welcome to the website of SMART Lab!
At the SMART Lab, our mission is to pioneer advancements in nuclear engineering by integrating state-of-the-art machine learning with modeling and simulation techniques. Our focus is on three areas:
- Physics-Informed Machine Learning : We develop and apply physics-informed machine learning within the MOOSE framework to support advanced reactor system modeling and simulation. This approach balances accuracy and efficiency, incorporating uncertainty quantification and data assimilation to ensure reliable predictive capabilities.
- AI-enhanced, Digital Twin-Integrated Cyber-Physical Testbed: We have built a thermal-fluid testbed that integrates hardware, software, and digital twin technology for real-time monitoring and autonomous control. This platform enables flexible operation, remote experimentation, and scalable AI-driven control strategies, providing a versatile environment for diverse reactor simulation scenarios.
- Generative AI for Science and Engineering: We explore the use of large language models (LLMs) for nuclear engineering applications—ranging from real-time status monitoring to AI-assisted operations of complex nuclear tools. By incorporating domain-specific knowledge into LLMs, we aim to develop robust solutions for a variety of nuclear engineering challenges.
Openings
We have openings for 1-2 graduate students. The students should be highly motivated and are expected to work on AI/ML and its applications to nuclear technologies.
Highly motivated undergraduate students are welcome to participate in our research. Undergraduate students are encouraged to publish their research results.