Sony's table tennis bot is already crushing ranked pros

Most ping-pong robots you see at tech expos are glorified ball launchers. Feed them a gentle topspin, and they will nudge it back; throw in heavy sidespin or a sharp cut, and their mechanical joints instantly choke.

Sony just shattered that benchmark. On June 10, the company showcased Ace, an autonomous table tennis robot that recently took on active human professionals under official competition rules officiated by Japan Table Tennis Association umpires. Between February and April, Ace defeated seven world-ranked pros.

Its victims include Miyuu Kihara, who was ranked world No. 26 in women's singles, and two-time Olympic silver medalist Miu Hirano. After her match, Hirano sounded genuinely stunned, asking if anyone could actually beat it.

Reading spin in under a millisecond

Anyone who has played table tennis knows the game is all about spin, not just trajectory. Topspin, backspin, and sidespin alter how the ball bends in midair and jumps off the rubber. If a bot cannot decode spin in real time, it simply cannot survive a rally against a pro.

Ace solves this through pure vision. The system can estimate the ball's rotation and rotational axis in less than a millisecond just by tracking it visually. When it strikes, it does not merely block; it actively slices, loops, and varies spin and placement to force human errors.

The mechanical design is equally brutal. Ace does not bother with bipedal legs. Instead, it rides on a heavy-duty electric rail capable of sliding the entire torso side to side almost instantaneously.

Hirano highlighted a telling detail about Ace's footwork. Great human players rarely hit the ball while on the run; they sprint to the spot, plant their feet, and then execute cleanly, which makes it nearly impossible for the opponent to read the shot. Ace plays exactly like that. It arrives early, comes to a dead stop, and strikes, hiding its intentions completely.

A top-10 contender in women's singles

Sony first published its foundational work on Ace in Nature last year, where the bot handled advanced amateurs but fell short against professionals. In December 2025, Ace notched its first pro win. By this spring, it was consistently beating ranked players across full-length matches.

Hirano gave a remarkably candid assessment of where the machine stands. In steady rallies, she said, Ace makes virtually zero mistakes. In her view, the robot already has the potential to rank in the global top ten among female players.

Its main limitation comes down to sheer explosive power. Against elite male players with heavier serves and harder drive shots, Hirano believes Ace's defensive blocks could still be overpowered, placing it around roughly 50th in the men's rankings.

Why it matters

This is widely considered the first time an autonomous robotic system has defeated ranked professionals in a head-to-head adversarial sport under official regulatory standards.

Sony made an exceptionally smart engineering trade-off here. By ditching human-like legs in favor of an industrial electric gantry, the team prioritized raw lateral acceleration, low latency, and rock-solid repeatability over mobility hype. For a game decided in milliseconds, that pragmatism made all the difference.

For professional sports, this changes how elite talent trains. National teams typically spend enormous resources training human sparring partners to emulate specific rival styles. A machine like Ace never fatigues, maintains near-perfect rally consistency, and can theoretically be dialed to simulate any opponent's spin signature on demand.

That said, the machine is still bound to a custom high-power rail, meaning you will not see it outside specialized research centers anytime soon. But as a proof of concept, Ace shows that when real-time computer vision meets raw industrial actuators, human dominance in lightning-fast sports is no longer guaranteed.