Sony AI unveils table tennis robot capable of defeating elite human players

Sony’s artificial intelligence division has announced a major breakthrough in robotics, revealing a table tennis-playing machine capable of outperforming elite human opponents, Sport360NG reports. 

In research published on Thursday in the journal Nature, engineers from the Japanese technology giant introduced an AI-powered robotic arm designed to handle the speed, spin, and unpredictability of high-level table tennis. The system, known as “Ace,” is equipped to perceive incoming shots, react instantly, and return them with remarkable precision.

Despite its imposing size, comparable to a full table tennis setup, Ace demonstrates impressive agility. The industrial robot executes rapid, coordinated movements, making split-second decisions during rallies. Its operation is marked by a constant mechanical hum, with whirring and rattling sounds accompanying its fast-moving arm.

During testing, the robot faced a group of seven high-level opponents, including five elite players and two professionals. It secured victories in three of five matches against the elite group, while the remaining contests were described as closely fought.

Sony AI noted that the robot’s performance has continued to improve beyond the initial research phase. In more recent trials, Ace has recorded additional wins against professional players, displaying enhanced shot speed, sharper placement near the table edges, and greater intensity during rallies.

The company believes the development represents more than just a sporting milestone. According to Sony AI, solving the challenges of real-time perception and control in such a fast-paced environment could pave the way for advanced AI systems capable of operating safely in dynamic, real-world conditions.

Earlier generations of table tennis robots were limited to basic rallies and failed to reach competitive standards. However, Sony AI’s Chief Scientist, Peter Stone, said the achievement signals a turning point.

“Reaching expert human-level performance in such conditions opens the door to a new range of real-world applications that were previously unattainable,” he said.