Humanoid robots can now play tennis in the style of Nadal, Federer, and Djokovic, learned from broadcast footage.
AdaPT (Noitom Robotics, Shanghai AI Lab, Dobot, SJTU) extracts professional player motions from TV broadcasts, retargets them to a humanoid, and learns both rallying and serving while preserving each player's distinctive style.
The core problem: decoupled planning-and-tracking works in sim but degrades badly on real hardware. Their fix is speed adaptation, training the tracker on randomized execution speeds while the planner learns to adjust motion speed to the incoming ball.
Deployed on Unitree G1 and the full-size Dobot Atom.
Worth noting on the sensing: the robot has no onboard vision. Ball and robot position both come from external infrastructure, a 35-camera motion-capture arena in the lab.
Paper: https://t.co/xrHLo2u9ts
Forehand. Backhand. Seamless transition.
Humanoid table tennis is getting seriously smooth.
Maybe sports are indeed a commercial use case for humanoid?
On August 19, Unitree Robotics (https://t.co/AfvedBvKl4) debuted on the STAR Market of the Shanghai Stock Exchange, becoming China’s first publicly traded “humanoid robotics stock.” The shares opened 629.44% above the IPO price at approximately $163.04 per share, giving the company a market capitalization of roughly $65.94 billion.
According to the offering announcement, Unitree’s IPO price was RMB 150.80, or approximately $22.35 per share, implying an IPO valuation of about $9.04 billion. A standard winning IPO allocation consists of 500 shares, requiring an initial payment of approximately $11,175. Unitree was the second-highest-priced A-share IPO so far this year, behind Pinzhun Laser, which also listed on the STAR Market on August 18.
At the opening price of approximately $163.04 per share, a 500-share IPO allocation would have generated an unrealized gain of more than $69,600.
Unitree is currently trading at approximately $133.10 per share.
A true world first: a humanoid robot autonomously serving a ping-pong ball with a dexterous hand.
The HMI Lab research team from Peking University has achieved autonomous table-tennis serving through coordinated whole-body control and high-DoF dexterous manipulation.
The robot autonomously tosses the ball with its dexterous hand, coordinates its body, arm and wrist, and completes the serve — bringing perception, decision-making, locomotion and fine manipulation into one unified loop.
A small serve for a robot, a major step toward general-purpose embodied intelligence.