Humanoid Robotics

Humanoid robotics studies how to build and control human-shaped machines for tasks that combine walking, balance, manipulation, and interaction in spaces designed for people. Common approaches include model-based whole-body control and model-predictive control, reinforcement learning and motion imitation, and perception systems that use cameras, depth, touch, or proprioception. Human demonstrations, video, and simulation are increasingly used to train skills, while benchmarks test locomotion, object handling, collaboration, and social behavior. A central challenge is making skills learned in simulation or from demonstrations reliable under real-world contact, uncertainty, and disturbances, while evaluating safety and task success consistently.