Chinese Humanoid Robots Outrun Usain Bolt's 100-Metre Record at Beijing Games
A humanoid robot named Tianzhuo, developed by China's Beijing Innovation Centre of Humanoid Robotics, ran 100 metres in 9.39 seconds at the World Humanoid Robot Games in Beijing, surpassing Usain Bolt's human world record of 9.58 seconds set in 2009. BBC News
The World Humanoid Robot Games opened on Saturday, August 22, and feature a programme spanning boxing, football, and household tasks such as folding laundry. The sprint event drew three competing robots. Tianzhuo took first place. Among the runners-up was Lightning, a humanoid robot built by Chinese smartphone manufacturer Honor. BBC News
Lightning, a thin robot with a black and red exterior, had already attracted attention before the main competition. China's state broadcaster reported that Honor's Lightning ran the 100 metres in 9.32 seconds, also beating Bolt's record. However, that time was recorded during a preparatory test event for the second World Humanoid Robot Games, not during the main games themselves. The Guardian Reuters ABC News Al Jazeera
During that same preparatory run, Lightning reached a peak speed of 14.5 metres per second. NBC News Reuters ABC News Al Jazeera
The distinction between the test-event performance and the official competition result matters. Lightning's 9.32-second run and Tianzhuo's 9.39-second run are not directly comparable, since they occurred in different settings. Honor's faster time was achieved in a preparatory context, while Tianzhuo's 9.39-second performance came during the Games themselves.
The broader context here is the accelerating convergence of humanoid robotics with competitive benchmarking. The World Humanoid Robot Games functions simultaneously as a sporting spectacle and a public stress test for bipedal locomotion, balance, and dynamic motor control. Sprinting at these speeds demands not just raw motor power but real-time adjustments to stride pattern, management of the forces the robot's feet generate against the ground, and rapid feedback from sensors that tell the robot where its own limbs are, a sense similar to human balance. The fact that multiple Chinese-built robots independently broke a 17-year-old human record, in the same week and at the same venue, points to a level of progress in humanoid movement that is difficult to dismiss as a one-off engineering display.
The competitive field also reflects an emerging industrial pattern: companies whose core businesses lie elsewhere are entering humanoid robotics. Honor is primarily a smartphone manufacturer. Its participation, alongside a dedicated national robotics innovation centre, points to a breadth of institutional investment in the humanoid form factor that extends beyond specialist robotics firms.
There are caveats worth weighing. The sprint times, particularly Lightning's, were reported by Chinese state media rather than independently verified by international athletics or robotics adjudicators. The conditions of the track surface, wind measurements, and timing methodology have not been detailed in the reporting. Bolt's 9.58-second record was set under World Athletics regulations with electronic timing, wind gauges, and doping controls. Whether the robotic performances were subject to comparable standardisation is unclear from the available sources.
None of this diminishes the engineering achievement. A bipedal machine covering 100 metres at peak speeds of 14.5 metres per second without collapsing is a significant control-systems challenge, and the times posted at the Games place humanoid sprint performance firmly within the range of elite human capability.
What remains less certain is how these benchmarks translate into practical deployment. The Games' inclusion of household tasks alongside athletic events suggests the organisers are framing humanoid robots as general-purpose platforms, not single-discipline specialists. The gap between sprinting on a controlled surface and navigating an unstructured home environment is enormous. But the sprint results provide a measurable, comparable data point for tracking progress in humanoid locomotion, and the fact that the conversation has shifted from whether robots can match human athletic performance to which robot does it best is itself a meaningful signal.
