Esports-Level Micromotor Skills: Sony Ace Robot Masters Table Tennis

Physical AI is breaking through another barrier. On April 23, 2026, a study by Sony AI was published in the journal Nature, demonstrating an autonomous robot named Ace playing table tennis at the level of elite athletes.

If the recent Chinese half-marathon proved chassis autonomy and endurance, Sony’s project demonstrates an previously unimaginable level of physics calculation and micromotor skills. The robot analyzes the ball’s spin, trajectory, and the opponent’s actions in real-time (fractions of a second), adapting its servos for the perfect strike. This is a fundamental breakthrough in Reinforcement Learning (RL) in the physical world. Algorithms capable of beating a human at ping-pong represent a direct path to the creation of next-generation surgical robots and ultra-precise assembly manipulators.

Source: Sony AI / Nature