Sony Group Corporation announced a significant milestone in robotics and artificial intelligence on April 22, 2026, with the publication of a study in the journal Nature documenting a robotic system capable of defeating professional table tennis players. The system, developed by Sony AI and named Ace, represents a breakthrough in the field of high-speed physical automation and reinforcement learning. According to the study, Ace is the first robotic entity to consistently challenge and occasionally overcome elite human athletes in a sport requiring millisecond-level reaction times and complex physical coordination.
The technical architecture of Ace relies on a sophisticated array of nine high-speed cameras positioned around a standard Olympic-sized court at Sony’s headquarters in Tokyo. These sensors allow the robot to track the ball’s trajectory and, crucially, its spin by monitoring the rotation of the manufacturer’s logo on the ball. This level of visual perception is paired with a robotic arm capable of rapid, precise movements. Peter Dürr, a lead researcher at Sony AI and co-author of the Nature paper, stated that the complexity of table tennis made traditional programming methods obsolete. Instead, the team utilized reinforcement learning, a subset of AI where the machine learns optimal behaviors through millions of simulated and real-world trials.
During the controlled experiments, Sony pitted Ace against a range of players, including highly skilled amateurs and professional athletes. While human players still maintain advantages in strategic unpredictability and long-term stamina, the robot demonstrated superior consistency in returning high-velocity shots and adjusting to varied spin patterns. The study notes that Ace achieved a competitive win rate against professional-tier opponents, a feat previously considered a grand challenge in the robotics community due to the inherent latency in mechanical systems.
This development underscores Japan’s continued strategic focus on robotics as a pillar of its national industrial policy. As global competition intensifies between the United States, China, and Japan for dominance in the AI sector, Sony’s achievement highlights a shift from generative AI models to embodied AI—systems that can interact physically with the real world. The Japanese government has historically supported such initiatives through the Society 5.0 framework, which aims to integrate advanced technologies into all aspects of social and industrial life.
Beyond the sporting arena, the researchers indicated that the technologies refined through the Ace project have broader implications for industrial automation, disaster response, and precision manufacturing. The ability of a machine to process visual data and execute precise motor responses in real-time is a critical requirement for the next generation of collaborative robots, or cobots, intended to work alongside humans in complex environments. Sony’s report concludes that while Ace is a specialized athlete, the underlying neural networks provide a template for more agile and adaptable machines across various sectors.