AI-powered robot beats elite table tennis players

2026-04-22 · Source: AI (artificial intelligence) | The Guardian · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Robotics & Autonomous Systems · Depth: Intermediate, short

Summary

Sony AI's robotic system, Ace, has achieved a significant milestone by beating elite human table tennis players in matches played under official rules. Ace won three out of five matches against elite players, though it lost both matches against professionals. The robot demonstrated mastery of spin, handled difficult shots like net catches, and executed a rapid backspin shot previously thought impossible by a professional. The system utilizes an eight-jointed arm on a movable base and multiple cameras positioned around the court to track the ball's position and spin. Its ability to estimate spin by zooming in on the ball's logo and its decision-making for shots were refined through 3,000 hours of computer simulations, with serves learned from expert players. This feat is considered a major advancement in robotics, addressing the complex demands of real-world competitive sports.

Key takeaway

For robotics engineers developing systems for dynamic, real-time environments, Ace's success highlights the importance of advanced perception and simulation-based training. Your projects could benefit from implementing multi-camera vision for precise object tracking and leveraging extensive simulated gameplay to refine decision-making and motor control, especially for tasks requiring rapid, adaptive responses against human opponents. Consider how the absence of human cues in your robot's design might impact interaction.

Key insights

Sony AI's Ace robot demonstrates advanced real-time perception and control in complex human-robot interaction.

Principles

• Multi-camera vision enhances real-time object tracking.
• Simulation training refines complex motor skills.
• Robots lack human tells, altering competitive dynamics.

Method

Ace estimates ball spin and axis of rotation using camera systems that zoom on the ball's logo. Shot decisions are honed via 3,000 hours of simulated gameplay, while serves are derived from expert player data.

In practice

• Integrate multi-camera systems for enhanced perception.
• Utilize simulation for rapid skill acquisition.
• Design robotic systems for specific task constraints.

Topics

• AI Robotics
• Table Tennis
• Sony AI
• Robot Perception
• Machine Learning Simulation

Best for: Computer Vision Engineer, AI Scientist, Robotics Engineer, Research Scientist

Related on AIssential

• This robot can beat you at table tennis — The "Ace" robot demonstrates AI's ability to surpass elite human performance in complex physical sports.
• Watch Sony’s elite ping-pong robot beat top-ranked players — Sony's Ace robot demonstrates AI's capability to excel in complex physical sports like table tennis against top human players.
• Sony AI table tennis robot outplays elite human players — Ace is the first robot to defeat elite human players in a competitive physical sport.
• Meet Ace, the table-tennis robot that can beat elite players — High-speed perception systems enable robots to achieve elite performance in complex, dynamic tasks like table tennis.
• Game, Set, Bot: Sony AI’s ‘Ace’ Serves Up a Defeat to Table Tennis Pros — Gaming environments serve as safe, scalable testbeds for developing and benchmarking advanced physical AI agents.
• Table tennis robot defeats some of world’s best players – why this has major implications for robotics — Ace robot demonstrates AI's ability to achieve superhuman performance in complex, real-world physical environments.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by AI (artificial intelligence) | The Guardian.