Sony AI has introduced Project Ace — the first autonomous system capable of defeating professional table tennis players in real matches. The AI-powered robot won three out of five matches against elite players.
Table tennis is considered one of the most challenging tasks in robotics. This is due to the extreme speed and unpredictable spin of the ball, which is very difficult to calculate within fractions of a second.
Thanks to a system of nine cameras and sensors, Project Ace tracks the ball’s trajectory in 3D. This allows the machine to instantly adapt to spins of any complexity.
During the matches, the robot demonstrated superior technique. It scored twice as many points on direct shots as professional athletes and returned more than 75% of the most difficult hits.
Project lead Peter Dürr called this achievement a milestone moment. Unlike computer games, this requires AI to have perfect coordination of movements in real physical space.
ORIENT
