Video: UC Berkeley humanoid robot plays table tennis with human-like agility
SMRTR summary
UC Berkeley's humanoid robot HITTER showcases impressive table tennis abilities, maintaining long rallies with humans. Its dual-system design combines a high-level planner using external cameras and a low-level calculator for coordinated movements. Trained on human motion data through model-based planning and reinforcement learning, HITTER reacts swiftly to fast-moving balls with natural movements. This advancement signifies progress toward more adaptable, human-like robotic interactions.
SMRTR provides this summary for quick context. The original article belongs to Interesting Engineering.
Read the original article