Sony AI's Project Ace Robot Defeats Professional Table Tennis Players in Historic First

Sony AI has achieved a landmark milestone in robotics and artificial intelligence: its autonomous system, Project Ace, has defeated elite and professional table tennis players under official International Table Tennis Federation regulations — the first time a robot has reached expert-level competitive performance in a physically demanding sport.

The robot secured three victories in five matches against elite players, with competitive performances in all remaining matches. The achievement marks a significant leap from previous robotic sports demonstrations, which had largely operated in controlled, simplified conditions rather than facing world-class human opponents under real match pressure.

Project Ace relies on three core innovations. Its perception system employs nine high-speed cameras with Sony's IMX273 sensors to precisely determine ball position, supplemented by three gaze control systems using event-based vision sensors capable of measuring ball spin and angular velocity in real time. The system successfully returned over 75% of serves with spins reaching 450 radians per second — far beyond what previous robots could handle. A model-free reinforcement learning control system enables the robot to adapt rapidly without relying on pre-programmed physical models, while state-of-the-art high-speed robotic hardware executes the resulting actions with millisecond-level precision.

Perhaps most striking is Project Ace's scoring record: the robot landed 16 aces against elite players, compared to just eight aces scored against it collectively. Performance continued to improve across matches held in December 2025 and March 2026, suggesting the system is still far from its ceiling.

The broader significance extends well beyond table tennis. Competitive physical sports represent one of the most demanding real-world testbeds for AI — requiring simultaneous mastery of ultra-fast perception, real-time decision-making, and precise physical execution with no tolerance for lag. Sony AI researchers suggest this capability profile could translate to robots operating in safety-critical environments, manufacturing floors, and interactive service roles where speed and adaptability matter as much as raw intelligence. Project Ace's success signals that the gap between AI performance in digital domains and the physical world is narrowing faster than many expected.