Bipedal Robot Cassie

Cassie, a highly agile bipedal robot, serves as a prominent platform for advancing research in legged locomotion. Current research focuses on developing robust control algorithms, often employing model predictive control (MPC) and reinforcement learning (RL), to enable Cassie to navigate challenging terrains, including uneven surfaces and stairs, and perform dynamic maneuvers like jumping and turning. These efforts leverage reduced-order models to improve computational efficiency and incorporate techniques like Bayesian optimization for safe and efficient parameter learning, bridging the gap between simulation and real-world performance. This work contributes significantly to the broader field of robotics by pushing the boundaries of dynamic stability and adaptability in legged robots.