CPS-VIDA Lab Members

Jyotirmoy V. Deshmukh

Applies formal methods to cyber-physical systems, embedded control systems, temporal logics, and time-series analysis.

Merve Atasever

Merve completed a double major in Mathematics and Physics at Bogazici University and a master’s degree in Differential Geometry at Istanbul Technical University. Before beginning doctoral studies at USC, she worked as a data scientist on AI projects. Her research interests include reinforcement learning and robotics.

• Multi-agent reinforcement learning
• Representation learning
• Robotics

Sheryl Paul

My research is about how collective behavior emerges in multi-agent systems, and how we can specify, monitor, and steer that behavior with formal guarantees. I draw on control theory, formal methods, and game theory to connect what individual agents do to the population-level patterns they produce across opinion dynamics, transportation and mobility, formal methods for search and rescue settings, and population models with evolutionary game theory.

• Multi-agent systems
• Formal methods
• Control theory
• Game theory
• Reinforcement learning

Vidisha Kudalkar

Yiqi “Nick” Zhao

Yiqi (Nick) Zhao is a PhD Candidate in Computer Science at the University of Southern California (USC). He is advised by Professor Jyo Deshmukh and Professor Yue Wang. He also works closely with Professor Lars Lindemann. His research focuses on multirobot design and verification.

• Robotics
• Cyber-physical systems
• Systems and control theory
• Formal methods

Sam Williams

Sam Williams is a USC Computer Science Ph.D. student and CPS-VIDA Lab member. His research applies game theory to reason about strategic behavior in multi-agent cyber-physical systems. He received a B.S. in Computer Science from the University of Colorado Boulder in 2022.

• Formal methods
• Game theory
• Multi-agent systems
• Optimization

Iurii Biktairov

Edward Kim

Edward's research concerns predictive monitoring and reachability analysis. He is interested in using techniques from causal inference, statistics, and physics-informed methods to handle monitoring in high dimensions.

• Formal methods
• Causal inference
• Nonlinear dynamics
• High-dimensional Probability

Ryan Rubel