US Edition
Sony AI Robot Ace Beats Elite Ping-Pong Players
Ace used reinforcement learning and high-speed sensors to beat elite players and challenge professionals, showing how adaptive robots can handle fast real-world competition.
- On Wednesday, April 22, 2026, researchers published a study in Nature detailing how Sony AI's robot, Ace, achieved "expert-level play in a commonly played competitive sport in the physical world."
- Between April 2025 and December 2025, Sony scientists refined Ace's performance, enabling the robot to beat both elite players and professional Japanese league players Minami Ando and Kakeru Sone.
- Using high-speed cameras and proprietary reinforcement learning, Ace identifies the ball's 3D position, scoring 16 unchallenged "aces" and maintaining a 75% serve return rate.
- Peter Dürr, director of Sony AI in Zurich, noted that table tennis requires "split-second decisions," highlighting potential for physical AI agents in complex, real-time interactive tasks.
- "This breakthrough is much bigger than table tennis," said Stone, chief scientist at Sony AI, as the field has advanced rapidly since Google DeepMind's robot was defeated two years ago.
58 Articles
58 Articles
Robot beats elite human players at table tennis
A robot beat elite human players at table tennis, a landmark victory in a competitive sport. Sony Ace defeated top amateurs in three matches out of five, although it lost twice against pros. It follows a running robot beating the human half-marathon record, albeit in a highly controlled environment. Ace had it easy to some extent — it has no legs so needn’t worry about balance, and has nine video cameras around the table, while humans tend to ha…
72 revolutions per second, speed up to 70 km/h – the robot "Ace" plays table tennis at the top level in lightning speed and precision. Impressive videos prove the progress. One point, however, makes experts skeptical.
A new artificial intelligence robot takes on and beats professional ping pong players, paving the way for adaptive robotics in the physical world
Med blixtsnabba riktningsförändringar för att ta en svår skruv och analyser av bollbanan för att Komma med en bra retur kan AI-roboten
Coverage Details
Bias Distribution
- 43% of the sources are Center
Factuality
To view factuality data please Upgrade to Premium
