The upcoming Doctoral Research Seminar will present "Optimal control and learning for legged locomotion“ by Majid Khadiv

January 22nd, 2024, 11:00 – 12:00 in room 2026, Karlstr. 45.

Planning motions for loco-manipulation systems is extremely complicated due to their highly nonlinear dynamics, under-actuation and inherent instability, as well as the hybrid nature of contact interaction. The two dominant approaches to control legged robots in multi-contact scenarios are optimal control (OC) and reinforcement learning (RL). In this talk, I will present my recent research efforts on devising a general framework based on OC and RL to efficiently control legged systems. I will also discuss my perspectives on how to generalize this framework toward realizing a fully autonomous humanoid robot.

#robotics #humanoids