Sony AI develops groundbreaking table tennis robot
Sony AI has developed a robot capable of taking on some of the world’s finest table tennis players, marking what researchers describe as a “landmark moment” for artificial intelligence, reports BritPanorama.
The machine, named Ace, has defeated elite human opponents in official matches, a feat its creators believe stands alongside historic AI triumphs such as IBM’s Deep Blue defeating chess grandmaster Garry Kasparov in 1997 and DeepMind’s mastery of Go in 2016.
While AI systems have long excelled in board games, the physical demands of fast-paced sports had remained elusive. Table tennis, characterised by balls travelling at speeds exceeding 45 miles per hour and intricate spin patterns, presented a formidable challenge.
To match the speed of professional play, Ace employs a sophisticated vision system comprising nine high-speed cameras that monitor the ball’s position in three dimensions. Three additional event-based sensors focus specifically on detecting spin by tracking surface markings such as the manufacturer’s logo. Rather than capturing conventional images, these cameras register minute fluctuations in light as they occur, enabling them to follow the ball at extreme velocities. This data feeds into a computer trained through thousands of hours of virtual gameplay.
An eight-jointed robotic arm executes each stroke, replicating human speed and range of motion within a split second. Under official competition rules, Ace emerged victorious against three out of five elite amateur opponents. The robot successfully returned over three-quarters of heavily spun shots and demonstrated the ability to adjust when balls took unexpected deflections off the net. However, when facing professionals Minami Ando and Kakeru Sone, the machine lost both encounters but managed to claim one game from the seven played.
Rather than relying on superhuman power, Ace’s success stems from its ability to produce varied spin types and remarkable consistency in returning the ball across the table. During one rally, it executed a spin shot that human experts had deemed physically impossible. Kinjiro Nakamura, a table tennis specialist who competed at the 1992 Olympics, witnessed Ace intercept a ball early and apply backspin in a manner he had never seen before, stating, “No one else would have been able to do that.”
Nakamura added, “I didn’t think it was possible. But the fact that it was possible means that there is a possibility that a human could do it too.” Peter Stone, chief scientist at Sony AI, indicated that the significance of this technology extends beyond the realm of sport, saying, “This breakthrough is much bigger than table tennis. 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.”
As technology continues to blur the lines between human and machine capabilities, the implications for both sport and broader society are profound. The spectacle of a robot mastering the nuances of table tennis speaks volumes about the future of AI in competitive environments — a reminder that human ingenuity and innovation may still hold the edge in creative expression and unpredictability.
