Max Robotics — Global Service Robot Directory + US Certification

About

Avadesh Meduri is a leading researcher in legged robotics, whose work centers on bridging the gap between theoretical control and real-world dynamic locomotion. His primary contributions lie in nonlinear model predictive control (MPC), whole-body motion planning, and state estimation for highly agile robots. Meduri’s most influential work, the BiConMP framework (88 citations), revolutionizes online whole-body trajectory generation by efficiently exploiting robot dynamics structure, enabling real-time nonlinear MPC for complex maneuvers. He has also advanced high-frequency MPC for manipulators (73 citations), demonstrating that reactive, optimal control can be deployed on physical hardware despite computational constraints. Beyond control, Meduri tackles the critical challenge of state estimation during dynamic locomotion, developing visual-inertial and leg odometry fusion methods that maintain reliable posture estimates even during flight phases. His recent explorations into diffusion-based learning for contact planning and risk-sensitive filtering show a forward-looking approach to robust, agile locomotion in constrained environments. With over 200 total citations and publications spanning top robotics venues, Meduri’s work is foundational for researchers seeking to deploy principled optimization and estimation algorithms on real, fast-moving robots.