A Chinese humanoid robot has shattered Usain Bolt’s 100-metre world record, but the machine’s extraordinary speed is also highlighting the areas where robots remain far behind humans: perception, decision-making and the ability to react to an unpredictable world.
Tiangong Ultra, developed by the Beijing Humanoid Robot Innovation Centre, ran 100 metres in 8.64 seconds in the final of the World Humanoid Robot Games in Beijing on Wednesday, almost a second faster than Bolt’s 9.58-second human record set in Berlin in 2009.
The robot averaged almost 42 kph (26 mph), setting the latest in a series of increasingly rapid times during the five-day competition. Tiangong had previously run 9.39 seconds and 8.86 seconds before lowering the mark again in the final.
The performance was a striking demonstration of how quickly humanoid robotics is advancing in China, where government support and intense competition among technology companies have accelerated investment in machines designed to move and work like people.
But sports scientists and robotics experts say comparing the robot directly with a human sprinter is misleading.
Tiangong did not run like Bolt. Its upright starting position, short and extremely rapid steps and powerful electric motors give it a very different set of physical characteristics from a human athlete.
“It is a remarkable advancement in technology where humanoids can keep balance and sprint at a speed that humans can’t catch up with,” said Jung Moon-hyun, a sports science professor at South Korea’s Chungnam National University.
At the start of the race, however, the robot was at a significant disadvantage.
Bolt exploded from the starting blocks in Berlin with a reaction time of just 0.146 seconds. Starting blocks allow human sprinters to generate horizontal force immediately and build speed rapidly.
Tiangong, by contrast, starts from a standing position. In its first record-breaking run, the robot also appeared to wait almost a second before moving after apparently failing to register the starting signal.
“If they could use starting blocks, I could see them getting a lot faster,” said Iain Hunter, a professor of exercise science at Brigham Young University and a sports scientist with USA Track and Field.
Hunter estimated that Bolt would have reached the five-metre mark almost two seconds ahead of Tiangong.
The robot’s advantages become clearer later in the race.
Its developer has redesigned parts of its torso and waist to reduce weight and lower the energy required to move the body while running. Engineers have also improved its motors, range of motion and control software, according to Guo Yijie, a manager in the humanoid innovation department.
The result is a running style that looks very different from that of an elite human sprinter.
See also: Google Changes EU Spam Policy to Avert Potential Antitrust Fine
Bolt completed his 100-metre world-record run in 41 steps. Tiangong appears to take more than 50 steps, according to a Reuters review of race footage, but its steps are much faster.
Park Suhan, a robotics professor at South Korea’s Kwangwoon University, said the robot’s gait appeared to be driven by the capabilities of its motors rather than an attempt to reproduce a human sprint.
Parunchaya Jamkrajang, a sports scientist at Thailand’s Mahidol University, said the robot’s short, rapid steps could also help it maintain balance.
“They try to increase speed, but if the stride becomes too wide, they will fall,” she said.
No fatigue, but another problem
One of the biggest differences between the robot and a human athlete is fatigue.
Human sprinters generate power through muscles, tendons and flexible joints. Their ankles, knees and hips absorb forces as they run and return some of that energy with each stride.
The system also comes with a biological limitation: muscles and the nervous system eventually fatigue.
Robots powered by electric motors do not experience fatigue in the same way.
“An electric motor simply creates rotation. It doesn’t need to pull and extend like muscles do. You can simply attach it at the joint,” said Park Hae-won, a professor at South Korea’s KAIST who leads the university’s humanoid robotics laboratory.
Matt Bundle, a biomechanics expert at the University of Montana, said elite human sprinters typically begin to slow in the final stages of a race as neuromuscular fatigue sets in.
Tiangong, by contrast, appeared to maintain its top speed through the closing stretch of its record run.
But that created another problem.
After crossing the finish line, the robot crashed into a padded barrier.
“Because the robot isn’t constrained like that,” Bundle said of the robot’s ability to maintain its speed. “It’s going to be able to keep going.”
The incident underscored a central challenge for humanoid robots: moving quickly is one problem, but knowing when and how to stop is another.
“It’s no good if we get up to a good top speed, but all we do is end up in a heap of broken bones because we can’t stop,” Bundle said.
Jamkrajang said future progress in humanoid robotics should be measured by more than raw speed.
Robots will need to be able to change direction, assess their surroundings and make decisions while moving, she said.
“Changing direction, or running and making a decision like, ‘I won’t run into this,’ and decelerating and reducing the robot’s speed. We haven’t seen this yet.”
From racing to real-world work
The contrast between speed and intelligence was on display throughout the World Humanoid Robot Games, where robots competed in events designed to test both athletic ability and practical skills.
The second edition of the competition brought together 2,056 humanoid robots from 666 teams across 16 countries, with events spanning sports and tasks intended to simulate real-world applications.
The Games included running, football, boxing and other athletic contests, as well as challenges involving tasks such as household work, hotel services and emergency response.
The competition offered a showcase for China’s rapidly expanding robotics industry, but it also exposed the limitations of machines that can perform spectacular feats under controlled conditions yet struggle when faced with unexpected situations.
Some robots fell during competitions, while others had difficulty completing relatively simple tasks. The contrast has raised questions about how quickly advances on the sports field can translate into useful robots for factories, warehouses, homes and other unpredictable environments.
The ability to run faster than the fastest human, for example, does not necessarily make a robot more useful.
A machine working alongside people may need to recognize an obstacle, understand that a person has suddenly entered its path, change direction, slow down and choose another route.
Those abilities depend not only on motors and mechanical strength but also on perception, software and artificial intelligence.
China has made humanoid robotics a strategic priority and is investing heavily in the development and manufacturing of machines capable of performing physical tasks.
The country already has a large industrial robotics sector, giving domestic manufacturers access to supply chains and production expertise that can help accelerate the development of humanoid machines.
The rapid improvement in Tiangong’s sprint times illustrates the pace of that development.
Last year’s winning time in the 100-metre humanoid race was 21.50 seconds. This year, Tiangong broke the record repeatedly before finishing at 8.64 seconds — faster than a human world record that has stood since 2009.
For robotics researchers, however, the next milestone may not be another fraction of a second shaved from a 100-metre time.
A robot that can run quickly, stop safely, turn around an obstacle and make an independent decision may represent a more important technological advance than one that can simply run faster than a human.
The spectacle in Beijing showed that robots are already capable of surpassing people in some narrowly defined physical tasks.
The harder challenge is making them capable of understanding the world well enough to know what to do next.
About the Author
Cecilia Attah is a tech analyst with a degree from Benue State University. She covers tech news and startups at TechRegard with a focus on how technology is transforming Africa and shaping the global landscape.
