Shimiao Li

Shimiao Li is an Assistant Professor in the Department of Electrical Engineering at the School of Engineering and Applied Sciences, University at Buffalo, specializing in power systems analysis, smart grid technologies, and AI-driven grid optimization. Her work develops physics-ML synergy tools for situational awareness and blackout prevention in critical infrastructure.

Education:

• PhD in Electrical and Computer Engineering, Carnegie Mellon University (2024)
• B.E. in Electrical Engineering, Tianjin University, China (2018)

Her research integrates circuit-theoretic modeling with machine learning to create sparsity-exploiting tools for state estimation, contingency analysis, and anomaly detection. Key focus areas include grid cybersecurity, real-time threat response, and physics-informed ML frameworks that bridge system dynamics and data-driven approaches for enhanced grid resilience.

Analysis of her 14 publications (2017-2024) reveals a dominant trend toward circuit-theoretic state estimation and probabilistic grid modeling, with increasing emphasis on ML integration post-2021. Her work consistently targets practical deployment challenges in power grid operations, evidenced by industry commercialization (SUGAR toolbox) and recognition from IEEE and Microsoft.

Awards:

• Microsoft Accelerate Foundation Models Research Initiative Award (2023)
• Best Paper Award, IEEE PES General Meeting (2021)
• CIT Dean’s Fellowship, Carnegie Mellon (2019)
• Chinese National Scholarship (2015, 2016)
• National Mathematics Contest First Prize, Tianjin (2015)

Dr. Li actively contributes to academic service as an NSF EPCN review panelist (2025) and DAC 2025 Technical Program Committee member. Her research has been supported through industry partnerships (Pearl Street Technologies) and Microsoft's foundation models initiative, with her SUGAR toolbox components commercialized for grid analysis.

She collaborates with Pacific Northwest National Laboratory (2022 internship) and leads projects advancing physics-ML synergy for grid situational awareness, including GridWarm for contingency analysis acceleration and node-breaker modeling for state estimation.