Sony AI Announces Breakthrough Research in Real-World Artificial Intelligence and Robotics

Published today in Nature - introduces Ace the first robot to beat elite human players in competitive physical sport

Demonstrates physical AI that can perceive, decide, and act faster than humans

Tokyo, Japan - April 23, 2026- Sony AI today announced its major breakthrough in robotics and artificial intelligence (AI) with the publication of its project Ace, the first known real-world autonomous system competitive with elite and professional-level human table tennis players. The research outlining the achievement, "Outplaying Elite Table Tennis Players with an Autonomous Robot," was published today on the cover of Nature .

This marks the first time a robot has achieved human, expert-level play in a commonly played competitive sport in the physical world - a longstanding milestone for AI and robotics research.

A Leap from Virtual to Physical AI

For decades, AI systems have demonstrated 'superhuman' performance in digital domains - from chess to Go to complex video games. However, applying AI to the physical world, especially in the high-speed domain where perception, planning, and control unfold in milliseconds, has remained one of the field's most significant challenges.

With Ace, Sony AI combined Sony's novel advanced sensor technology, reinforcement learning, and precision hardware to achieve expert-level play in a sport that demands fast, precise, and adversarial interactions near obstacles and at the edge of human reaction time. Building on Sony AI's breakthrough research on its superhuman AI agent Gran Turismo Sophy™ in the high-speed virtual domain, Ace extends this progress into real-world environments,exploring how robots can perceive, plan, and act with high-performance human speed and accuracy in dynamic environments.

The research implications extend beyond sport. By solving a problem that requires exceptional real-time sensing and control, this research lays the groundwork for AI systems that can safely and reliably operate in dynamic physical environments, ranging from safety-critical settings to real-time interactive domains, where outcomes can benefit from interactions at the edge of human performance.

"This research has shown that an autonomous robot can, in fact, win at a competitive sport, matching or exceeding the reaction time and decision making of humans in a physical space," said Peter Dürr, Director of Sony AI in Zürich, and project lead for Ace. "Table tennis is a game of enormous complexity that requires split-second decisions as well as speed and power. This research breakthrough highlights the potential of physical AI agents to perform real-time interactive tasks, and represents a significant step toward creating robots with broader applications in fast, precise, and real-time human interactions."

Pushing the Limits of Human-Robot Interaction

Table tennis is one of the most demanding and complex real-world tests for robotics, requiring rapid decision-making, precise physical execution, and continuous adaptation to an unpredictable opponent. The ball's high speed, spin, and complex trajectories - especially spin, often not sufficiently handled under official match environments in prior work - are central to competitive play.

To meet these demands, Ace, was designed with three novel components:

• A high speed perception system composed of nine active pixel sensor (APS) cameras equipped with "IMX273" image sensors from Sony Semiconductor Solutions (SSS) to determine the ball's precise 3D position, combined with three gaze control systems (GCS) that use event-based vision sensor (EVS) cameras with "IMX636" event-based vision sensors from SSS, pan/tilt mirrors, and telephoto tunable lens to measure the ball's angular velocity and spin in real time.
• A novel control system based on model-free reinforcement learning to enable rapid adaptation and decision-making without reliance on pre-programmed models.
• State-of-the-art high-speed robotic hardware capable of executing precise, high-speed control for agile physical interaction.
  

Proving Performance at the Edge of Human Reaction Time

For the results reported in the Nature publication, Ace was evaluated in matches against five elite players and two professional table tennis players, under International Table Tennis Federation (ITTF) regulations. Ace achieved three victories in five matches against the elite players, along with competitive performances in the remaining matches. Other interesting results from the evaluations included:

• Ace was able to return a wide range of spins, consistently achieving over 75% return rate up to 450 rad/s, demonstrating its excellent capability of handling spin, and far exceeding previously reported values in competitive table tennis robots.
  

• Ace scored 16 direct points after serving, sometimes called "aces," against the elite players, while the elite players collectively scored only eight.

• Ace's low-latency perception and control systems also allowed for quick reaction to unusual shots, such as balls bouncing off the net. This behavior illustrates the ability of our approach to generalize to situations that are both rare and hard to model in simulation.

These results demonstrate, for the first time, the potential of physical AI agents to outperform human experts in interactive, real-time tasks. While past researchers have built robots to play table tennis, most have been demonstrated only on cooperative rallying, and none has surpassed an amateur level in competitive play.

Continued Progress Following Publication

Following submission of the Nature manuscript, the team conducted additional competitive matches in December 2025 and March 2026. In the December matches against four new players: two professional and two elite, Ace defeated both elite players and one professional player, while losing to the second professional opponent. In the March 2026 matches against three new professional players Ace defeated all three players at least once. Compared with earlier evaluations, Ace demonstrated higher shot speeds, more aggressive placement closer to the table edge, and faster-paced rallies, reflecting continued performance gains under competitive conditions.

"This breakthrough is much bigger than table tennis," said Peter Stone, Chief Scientist at Sony AI. "It represents a landmark moment in AI research, showing, for the first time, that an AI system can perceive, reason, and act effectively in complex, rapidly changing real-world environments that demand precision and speed. Once AI can operate at an expert human level under these conditions, it opens the door to an entirely new class of real-world applications that were previously out of reach."

Further details about the Ace project and December and March matches can be found here: https://ace.ai.sony

About Sony AI

Sony AI, a division of Sony Research, was founded as a subsidiary of Sony Group Corporation on April 1, 2020, with the mission to "unleash human imagination and creativity with AI." Sony AI aims to combine cutting-edge research and development of artificial intelligence with Sony Group's imaging and sensing technology, robotics technology, and entertainment assets such as movies, music, and games to accelerate Sony's transformation into an AI-powered company and to create new business opportunities. To achieve this, Sony AI is working across six Flagship Projects that are aimed at the evolution and application of AI technology in the areas of AI for Creators, Gaming and Interactive Agents, Ethics, Scientific Discovery, Imaging and Sensing, and Robotics. For more information, visit https://ai.sony.com .

Media Contact:
Sony AI
Rebecca Garcia
[email protected]

Or

[email protected]