Michael Posa

Michael Posa is an Assistant Professor in Mechanical Engineering and Applied Mechanics at the University of Pennsylvania's School of Engineering and Applied Science. He also holds affiliations with the Departments of Computer and Information Science, and Electrical and Systems Engineering. As the head of the Dynamic Autonomy and Intelligent Robotics (DAIR) Lab, part of the GRASP Lab, his research focuses on control, learning, and planning for robots interacting dynamically and safely with complex environments. Key interests include non-smooth dynamics of contact, machine learning, and numerical optimization, with applications in legged robots and robotic manipulation. His work emphasizes computationally efficient algorithms for real-time control and decision-making.

Recent achievements include the 2024 Best Paper Award for contributions to multi-contact model predictive control. He actively mentors students like Brian Acosta and William Yang, whose theses address bipedal walking and dynamic manipulation. The DAIR Lab collaborates on interdisciplinary projects and participates in robotics conferences like ICRA. Michael’s lab emphasizes diversity and innovation, recruiting students across MEAM, ESE, and CIS departments.

His research bridges theory and practice, with publications spanning model reduction for legged systems, vision-based contact localization, and impact-aware control strategies. Ongoing efforts explore contact-implicit MPC frameworks and integrating physics-driven perception (e.g., Vysics) for robust autonomy.