China held its second edition of the World Humanoid Robot Games in late August 2026 at the Beijing National Speed Skating Oval, an event intended to showcase the country’s progress and commitment to robotics. The field has attracted a lot of research and development – not to mention interest – in recent years, especially after the arrival of large language models like ChatGPT and Gemini. Similar AI technology could provide the brains for humanoid robots. But the 2026 Robot Games in China didn’t make news because of new robotic innovations that would allow a humanoid to fold clothes faster than before. Instead, the event surprised the world with robots capable of running the 100 meters even faster than humans. A few robots repeatedly broke Usain Bolt’s 2009 world record (9.58 seconds), with the best time falling to 8.64 seconds in the final. Although these humanoids ran much faster races than last year, they were unable to stop like a human could. Instead, they crashed at full speed into a padded wall after crossing the finish line, requiring assistance.
A Tiangong Ultra robot model won the 100-meter race on Aug. 26 with a time of 8.64 seconds, a significant feat compared to its time of 21.50 seconds in 2025. A version of the robot finished in 9.39 seconds in a round earlier in the competition. A similar model improved to 8.86 seconds in a semi-final. The Honor Lightning model is another humanoid that beat Bolt’s time during the Games. The robot finished in 9.47 seconds in the opening round which was won by Tiangong Ultra. Before the Games, Lightning was even faster (9.32 seconds) in a test.
Robot games have also shown other types of humanoids involved in other games, including soccer, table tennis, and jumping, in addition to running.
Why is it important to beat Usain Bolt’s time?
Having a robot beat Bolt’s record, even if it means the machine would take significant damage from crashing into a wall at full speed at the end of the race, accomplishes two goals. First, it is a powerful marketing tool to show the rapid progress of the humanoid robotics industry, with a focus on Chinese companies. Second, a robot capable of completing a 100-meter run and reaching a speed high enough to beat Bolt’s record shows that the various technologies needed to allow a robot to reach that speed work outside of the laboratory.
A human running 100 meters would instinctively know what to do, from starting the race to accelerating, moving hands and feet, following the race, crossing the finish line, and braking. Researchers would need substantial development work for a robot to work in the same way. The batteries provide power to the motors that move the robot’s limbs. The humanoid’s joints and gears would perform the movement, while a cooling system would ensure the battery and mechanical parts don’t overheat. Also crucial are the software algorithms that allow the robot to plan the movement of its legs in real time. The software must take into account various factors, including landing on each foot without losing balance and planning the next step. The software would also determine the stride length and cadence required to win the race.
While this is just speculation, these technologies were all exposed from the race that went viral after the Games opened, regardless of the humanoid model. Honor’s Lightning robot initially led this race, with Tiangong eventually overtaking it towards the finish. Both robots broke Bolt’s record and both crashed into the wall.
What’s next for humanoid robots?
Humanoid robots broke other human records in longer races at the Games. Tiangong Ultra won the 400 meters race in 38.15 seconds, well below Wayde van Niekerk’s world record of 43.03 seconds. In the 1,500-meter race, the winning robot finished in 2:21.64, more than a minute faster than the human record of 3:26.00 (Hicham El Guerrouj).
It may not be necessary for a commercial humanoid robot to run faster than a human. Humanoids designed for warehouse work will need to perform specific tasks accurately and safely. Humanoids designed for the home will need to perform specific tasks, such as loading and unloading the dishwasher and taking care of the laundry. In this regard, China has already established humanoid robot schools that teach robots how to fold and unfold laundry, a key step in training AI algorithms related to precision movements. In other words, breaking Tiangong Ultra’s new record may not be a priority for researchers. But teaching a robot driving at high speed how to stop safely could be one of the things researchers study next.
At the very least, the researchers proved that their hardware was capable of faster speeds than last year. The batteries, motors, and joints worked reliably, and the algorithm allowed the robots to complete races. These are important developments for the industry, as they can offer additional data compared to laboratory testing. The robots may have been damaged by hitting the wall at high speed, but researchers will walk away with important data regarding a robot’s AI capabilities, mobility and durability while racing. These events could lead to key innovations for future business models, including a robot’s ability to understand its environment and adapt, or new battery technologies involving rapid energy discharge and thermal management.
