Sony’s Humanoid Robot Dominates High-Level Table Tennis Matches
111A new breakthrough in sports robotics has emerged as Sony AI unveils a humanoid robot that can compete with and beat top professional table tennis players. The robot, named Ace, has demonstrated its skills in official matches, showing that artificial intelligence can perform in fast-paced, real-world sports environments. This development marks a significant step forward in the field of physical AI, where machines are designed to operate effectively outside controlled settings.
How Ace Outperforms Human Players
Ace was built to handle the rapid decision-making and precise motor control needed in competitive table tennis. The robot uses a combination of high-speed perception systems and AI-driven controls to execute shots under match conditions. Equipped with nine synchronized cameras and three vision systems, Ace can track the ball’s movement and spin at speeds that are difficult for the human eye to follow. This allows it to react quickly and make accurate shots even during intense rallies.
The robot’s mechanical design includes eight joints that control the racket’s position, orientation, and force. This setup ensures it can meet the minimum requirements for professional play. Interestingly, Ace was trained entirely in simulated environments, allowing it to develop its own strategies and play styles. Unlike previous table tennis robots, which relied heavily on human demonstrations, Ace’s self-training approach enables it to adapt and improve through trial and error, much like a human player.
The Challenges of Playing Real-World Sports
Playing sports like table tennis is much more complex than digital games like chess or video games, where conditions are fully controlled. The speed of the ball, the spin, and changing trajectories make real-time responses extremely challenging. Sony AI’s project aims to explore how robots can respond with speed and accuracy in such dynamic environments. The team’s research has shown that capturing fast-moving objects and predicting their behavior requires advanced sensing and processing capabilities.
In tests conducted in April 2025, Ace won three out of five matches against elite players and lost two. By December 2025 and early 2026, the robot had also secured wins against professional players, demonstrating its growing competence. Professional player Mayuka Taira, who faced off against Ace, noted that the robot was unpredictable because it showed no visible signs of human-like hesitation or mistake. This makes it hard for opponents to anticipate its moves, adding a new layer of difficulty to competitive play.
Researchers believe that this kind of technology could eventually be used to train athletes or even develop new strategies for playing sports. The project’s findings were detailed in a study published in the journal Nature, highlighting the potential of AI and robotics to push the boundaries of human and machine performance in real-world activities.
What do you think?
It is nice to know your opinion. Leave a comment.