tech

Chinese Humanoid Robots Beat Usain Bolt’s 100-Meter Record—But It Wasn't Pretty

A robot named Tiangong Ultra ran the 100-meter dash in 9.39 seconds, but reports suggest the display was more a showcase of raw power than human-like grace, revealing the current state of robotic agility.

SignalEdge·August 24, 2026·3 min read
Engineers testing the gait and speed of a bipedal humanoid robot in a high-tech laboratory.

Key Takeaways

  • A Chinese humanoid robot, Tiangong Ultra, ran the 100-meter dash in 9.39 seconds, surpassing Usain Bolt's 9.58-second world record.
  • The feat occurred at the World Humanoid Robot Games in Beijing, a new showcase for robotics.
  • A second robot from the company Honor, named Lightning, also beat the record with a time of 9.47 seconds, according to The Verge.
  • Despite the speed, the robots' movements were described as awkward, indicating a gap between achieving a single performance metric and replicating fluid, human-like motion.

A 15-year-old human world record has been broken, but not by a human. At the World Humanoid Robot Games in Beijing, a bipedal robot named Tiangong Ultra ran the 100-meter dash in 9.39 seconds, eclipsing Usain Bolt’s 9.58-second record set in 2009. The Verge reports that a second robot, developed by electronics firm Honor and named Lightning, followed closely with a time of 9.47 seconds, also besting Bolt's historic time. This marks a significant milestone for robotic locomotion, achieved on a controlled track.

A Record Falls, An Asterisk Remains

The record-breaking sprints occurred during a preliminary heat on Saturday, positioning the Beijing Humanoid Robot Innovation Center—the creator of Tiangong Ultra—as a leader in a very specific metric: straight-line speed. While the raw numbers are impressive, they don't tell the whole story. Engadget noted that the robots looked “absolutely ridiculous” while performing the feat. This observation points to the core challenge in bipedal robotics. Achieving speed on a flat, predictable surface is an optimization problem that can be solved with powerful actuators and a control system narrowly focused on forward momentum. The result is a gait that, while effective for the singular task of sprinting, lacks the nuance, balance, and efficiency of a human athlete. The awkwardness is the data. It reveals that the current state of the art prioritizes raw performance over the complex, dynamic stability required for general-purpose movement. While Boston Dynamics focuses on parkour and Tesla's Optimus is shown folding laundry, these Chinese teams have targeted a classic, headline-friendly athletic benchmark.

More Benchmark Than Breakthrough

Together, these reports point to a classic engineering trade-off. The robots were fast, but not graceful. This suggests their design was heavily optimized for a single variable—speed—at the likely expense of others like stability on uneven terrain, energy efficiency, or the ability to stop and change direction quickly. This event is less a sign that robots are ready to walk among us and more a demonstration of what's possible within the tightly controlled confines of a competition. Breaking the 100-meter record is a powerful marketing tool and a clear benchmark of progress. It is not, however, a demonstration of the general-purpose agility needed for a robot to navigate a factory floor or a disaster zone. The structural force at play is the global race to establish dominance in robotics, and easily digestible metrics like a 100-meter sprint are effective for capturing public and investor attention. The real test is what happens when the track ends and the unpredictable real world begins.

SignalEdge Insight

  • What this means: Humanoid robot locomotion is advancing rapidly in raw performance metrics like speed, even if general-purpose agility and grace remain distant goals.
  • Who benefits: The Beijing Humanoid Robot Innovation Center and Honor, which gain significant PR and validation for their engineering programs.
  • Who loses: No one directly, but it applies pressure on Western competitors like Tesla and Figure AI to demonstrate comparable performance benchmarks.
  • What to watch: Whether these sprinting robots can be adapted for tasks requiring dynamic stability and navigation in unstructured, real-world environments.

Sources & References

Daily Newsletter

Stay ahead of the curve

Get the most important stories in tech, business, and finance delivered to your inbox every morning.

You might also like