UC San Diego Researcher Awarded LANL-UC Fellowship

Article Content

Yuanyuan Shi, Assistant Professor in the Department of Electrical and Computer Engineering at the University of California San Diego Jacobs School of Engineering, has been awarded the Michael R. Anastasio LANL-UC Early Career Faculty Fellowship from Los Alamos National Laboratory (LANL). The fellowship will support Shi’s research to develop a new artificial intelligence (AI)-driven framework that can rapidly model, simulate and control complex physical systems, while also estimating what is happening inside them using limited sensor data. The project aims to overcome longstanding barriers in studying systems governed by multiple interacting physical processes, where traditional simulations are often too slow and computationally expensive.

The fellowship offers a unique opportunity for rising tenure-track UC faculty to advance their research in computational science and strengthen their connection to Los Alamos National Laboratory through a joint appointment. Shi’s work aligns with the program’s focus on basic machine learning and AI methods for physical system modeling and control.

Many real-world, complex physical systems — including those involving fluid flow, electromagnetic fields, plasma dynamics and fusion — are shaped by multiple physical processes that interact simultaneously. Capturing these interactions in detail is challenging because it typically requires massive computing resources. This limits researchers’ ability to test ideas quickly or design effective control strategies. Another challenge is that scientists often rely on sensors located at only a few points in a system, which leaves them without a reliable way to observe what’s happening inside.

To address these challenges, Shi is developing a foundational neural operator learning framework that combines physics-based knowledge with advanced machine learning. Her approach could enable faster and more accurate simulations; smarter automated control; and the ability to estimate what’s happening inside a complex physical system with limited data.

“The selection process for the Los Alamos Michael Anatasio Fellowship was very competitive with applications from across the UC system,” said Tri Tran, a program officer at Los Alamos National Laboratory and a member of the fellowship committee. ”Professor Shi’s research work was selected out of almost 20 proposals over two rounds of reviews.”

The project has potential applications in autonomous lab development for particle accelerators at Los Alamos National Laboratory, where understanding and controlling beam dynamics behavior is essential, as well as in applications in building heating and ventilation control, fusion energy and other complex physical systems.