Michael Posa

Michael Posa is an Assistant Professor in the Department of Mechanical Engineering and Applied Mechanics at the University of Pennsylvania, where he leads the DAIR (Dynamics, Algorithms, and Intelligent Robotics) Lab. His research focuses on planning, control, and formal analysis for robots that physically interact with the world, with an emphasis on enabling autonomous systems to move dynamically and safely while managing the complexities of contact. Posa develops computationally tractable algorithms that combine numerical optimization, non-smooth dynamics, and control theory, allowing robots—both legged systems and manipulators—to handle impacts, friction, constrained footholds, and multi-contact interactions. His work explores contact-implicit trajectory optimization, MPC under uncertainty, hybrid dynamics modeling, impact-invariant control, geometry-aware contact learning, and scalable approaches that merge reinforcement learning with physics-based models. By rigorously studying the interplay between contact mechanics and algorithmic decision-making, Posa aims to build systems that can reliably maneuver in challenging real-world environments. His group validates these methods on hardware platforms ranging from bipedal robots to dexterous manipulation systems, advancing the foundations of high-performance, safety-critical robotic control.