Sony AI's "Ace" robot has achieved a significant milestone in artificial intelligence and robotics, demonstrating its ability to defeat professional table tennis players in recent matches. As highlighted in a social media post by Min Choi, "Sony's AI robot just beat professional table tennis players." This advanced AI-powered robotic arm, developed by the Japanese electronics giant, has showcased superhuman perception and real-time control, challenging and overcoming human experts in a dynamic physical sport.
Central to Ace's prowess is its sophisticated architecture, integrating nine synchronized frame-based cameras and three event-based vision systems to track the ball with millimeter accuracy and measure spin at up to 700 Hz. Min Choi's post further emphasized the robot's sensory capabilities, noting, "9 cameras. Reads the ball's spin by tracking the logo." This allows the robot to precisely read the ball's trajectory and rotation, crucial for anticipating its flight and executing precise returns. The system's striking skills are trained entirely in simulation using reinforcement learning, then seamlessly transferred to the physical robot.
Initially, a study published in Nature in April 2026 detailed Ace winning three out of five matches against elite players, though it faced tougher competition from professionals. However, Sony AI confirmed that subsequent improvements enabled Ace to defeat professional players in December 2025 and again in March 2026. Professional player Mayuka Taira, who lost a match to Ace, noted the robot's strengths: "it is very hard to predict, and it shows no emotion," according to Sony AI.
This achievement marks the first time an AI system has reached expert-level performance in a competitive physical sport under official rules, with matches officiated by licensed umpires. Peter Stone, Chief Scientist at Sony AI, emphasized the broader implications, stating, "This breakthrough is much bigger than table tennis." He added that solving a problem requiring exceptional real-time sensing and control lays the groundwork for AI systems to operate safely and reliably in dynamic physical environments, extending beyond sports.
While Ace has demonstrated remarkable agility and precision, researchers acknowledge that professional human athletes are adept at adapting to opponents and finding weaknesses. The robot's unique playing style, not learned from human observation, creates surprising situations. This ongoing research continues to push the boundaries of physical AI, paving the way for future applications in safety-critical settings and real-time interactive domains.
