Sony's Robot Beats Pro Table Tennis Players In Nature First
Sony AI's Project Ace, an autonomous robot that uses event-based vision sensors and reinforcement learning, has become the first machine to beat elite human players at table tennis under International Table Tennis Federation rules, according to a paper published on the cover of Nature last week. The system reacts in 20 milliseconds, more than ten times faster than a human pro, and returned spins of up to 450 radians per second with a 75 per cent success rate. Across formal matches against five elite players and two professionals, Ace took three wins from five and continued to beat both elite and one pro player in additional bouts in December and March. Sony Semiconductor Solutions supplied the high-speed image sensors at the heart of the perception stack.
Why it matters for Asia
Sony has pulled off what looks like the next "AlphaGo moment" for physical AI, and it has happened in Tokyo rather than Silicon Valley. For Asia's industrial heartland, where Japanese, Korean and Chinese factories are racing to deploy humanoid and bimanual robots into electronics assembly, automotive and warehouse floors, Ace shows that real-time reinforcement learning at sub-human latency is now production-ready. Manufacturing, logistics and consumer robotics buyers across the region should expect Sony's perception and control stack to start showing up in commercial products well inside this decade.