Sony AI Robot Ace Defeats Elite Human Table Tennis Players
A research team led by Sony Group Corp.'s artificial intelligence unit has developed Ace, an AI-powered robot capable of competing with and defeating skilled human table tennis players. According to findings published in the scientific journal Nature, Ace successfully beat three out of five elite players with over a decade of competitive experience, although it lost matches against two active professional athletes. The robot utilizes an eight-jointed arm and multiple cameras to track high-speed ball movements, enabling it to handle complex spins and trajectory changes, including those off the net. This achievement marks a significant milestone, as Ace is the first known AI system to outperform humans in interactive, physical skill-based games using unaltered professional equipment and International Table Tennis Federation rules. While not yet superior to top-tier professionals, Ace's performance demonstrates advanced real-time control and perception capabilities. Researchers suggest that the techniques underlying Ace's success have broader applications beyond sports, potentially revolutionizing areas requiring fast, precise human-robot interaction, such as manufacturing automation and service robotics.
Wire timeline
Sony AI Robot Ace Defeats Elite Human Table Tennis Players
A research team led by Sony Group Corp.'s artificial intelligence unit has developed Ace, an AI-powered robot capable of competing with and defeating skilled human table tennis players. According to findings published in the scientific journal Nature, Ace successfully beat three out of five elite players with over a decade of competitive experience, although it lost matches against two active professional athletes. The robot utilizes an eight-jointed arm and multiple cameras to track high-speed ball movements, enabling it to handle complex spins and trajectory changes, including those off the net. This achievement marks a significant milestone, as Ace is the first known AI system to outperform humans in interactive, physical skill-based games using unaltered professional equipment and International Table Tennis Federation rules. While not yet superior to top-tier professionals, Ace's performance demonstrates advanced real-time control and perception capabilities. Researchers suggest that the techniques underlying Ace's success have broader applications beyond sports, potentially revolutionizing areas requiring fast, precise human-robot interaction, such as manufacturing automation and service robotics.
The Mainichi