A revolution in robotics is unfolding on the track. Humanoid machines have already beaten humans over the half-marathon distance and are now within striking distance of the men's 100-meter sprint record, according to recent performance data from leading labs.

Engineers at the Dynamic Robotics Institute confirmed Wednesday that their latest bipedal prototype, HR-7, recorded a time of 10.8 seconds over 100 meters in controlled tests. This is just 1.5 seconds shy of Usain Bolt's world record of 9.58 seconds.

“We are seeing a convergence of hardware and AI that is accelerating at an extraordinary pace,” said Dr. Ellen Kwan, director of robotics at the institute. “The sprint record is now a matter of fine-tuning, not fundamental redesign.”

Background

The race to build faster humanoid robots has intensified over the past five years. Early models struggled to maintain balance at slow walking speeds. Today, machines like Boston Dynamics' Atlas and Honda's ASIMO have demonstrated parkour and stair climbing.

In April, the Chinese robot 'StrideMaster' completed a half-marathon in 1 hour 12 minutes—faster than the average human time. That milestone prompted a shift in focus toward explosive acceleration and top-end speed.

“The half-marathon showed us endurance is feasible,” noted Dr. Kwan. “Now we’re optimizing for the short burst. The biomechanics of a sprint are fundamentally different, but our simulations suggest a 9.9-second 100m is possible within 18 months.”

What This Means

The immediate question for industry watchers is: why build a robot that can sprint but has no clear use in homes or factories? Sales of humanoid robots for domestic chores remain negligible, and industrial arms are typically fixed or wheeled.

Analysts argue the sprint record is primarily a marketing and engineering benchmark. “Speed demonstrates control, power density, and real-time computation—all of which transfer to practical applications like emergency response or logistics,” said Marcus Voss, a senior analyst at TechNexus.

Yet critics warn that chasing athletic records distracts from solving pressing challenges in manipulation and autonomy. “A robot that can run a 100-meter dash but can’t pick up a cup of coffee is a party trick, not a product,” said Dr. Akiko Tanaka of the Human-Robot Interaction Lab in Tokyo.

Proponents counter that every breakthrough in locomotion pushes the envelope. The US Defense Advanced Research Projects Agency (DARPA) has funded several of these projects, citing potential for search-and-rescue in rubble or disaster zones where wheeled vehicles cannot go.

“The 100-meter record is a milestone, not a destination,” added Dr. Kwan. “Once we crack the sprint, we will redirect those control algorithms to more utilitarian tasks. This is about building a platform that can eventually do everything a human can do physically.”

The immediate next step is an official attempt under IAAF rules. The Institute plans to submit HR-7 for certification and hopes to stage a record attempt at a major robotics conference in 2025. If successful, it will mark the first time a humanoid has legally broken the 10-second barrier.