Berkeley’s HiTTER Robot May be Better at Table Tennis Than You
SMRTR summary
UC Berkeley's HiTTER robot showcases impressive table tennis abilities, playing over 100 consecutive shots against humans. Its two-part system includes a "brain" for ball tracking and response calculation, and a "body" controller trained on human motion data. By combining planning with reinforcement learning, HiTTER achieves natural movements while reacting swiftly to fast-moving balls. This advancement could impact various robotic applications beyond sports, including warehouses, homes, and surgery.
SMRTR provides this summary for quick context. The original article belongs to TechEBlog.
Read the original article