This research area focuses on developing intelligent decision-making frameworks and adaptive control strategies for autonomous agents operating in complex environments. We integrate classical control theory with modern machine learning and decision theory to create systems that can reason about uncertainty, learn from experience, and adapt their behavior in real-time.
Our research encompasses formal methods for safety assurance, bounded-rational decision models, learning-enabled components with formal guarantees, and hierarchical control architectures for autonomous systems. We develop techniques for handling incomplete information, adversarial scenarios, and human-in-the-loop decision-making.
Applications include autonomous vehicles, assistive robotics, cyber-physical systems, and human-agent collaboration. Our work emphasizes both theoretical foundations and practical implementation on real robotic platforms and autonomous vehicles.