Robots Take On Mount Tai in China’s First Major Mountain Challenge
Robots climbing Mount Tai may sound like science fiction, but the feat has now become reality. On Monday, the inaugural China Media Group World Robot Climbing Mount Tai Competition concluded in Tai’an, Shandong province. More than 50 embodied AI companies, universities and research teams from across China tackled a 3.9-kilometer course from Hongmen to Zhongtianmen.
The event was billed as the world’s first professional robotics competition held in a mountainous UNESCO World Heritage Site. It was no ordinary walking test on a flat surface: the route included more than 3,100 stone steps and a cumulative elevation gain of nearly 600 meters. It passed landmarks including Wanxian Tower, Doumu Palace and Stone Sutra Valley, while varying step heights and widths, uneven surfaces and changing gradients added to the difficulty.
There were no smooth laboratory floors or tightly controlled testing conditions. Organizers described the actual mountain trail as a natural testing ground for pushing embodied AI systems to their limits. The competition featured four events—humanoid robot racing, legged robot racing, legged robot load-bearing and powered exoskeleton assistance—plus a separate all-terrain mobility and pathfinding challenge. China Media Group broadcast the event across its multimedia platforms.
Legged robots delivered especially strong performances. In the legged robot race, a robot from Beijing Lingyi Technology, a subsidiary of Unitree Robotics, crossed the finish line first in 43 minutes 41 seconds to take the championship. Shanghai Gaoyi Technology, whose team also used Unitree robots, finished second, while Linxai Intelligent’s Lingju team was third.
The load-bearing event focused more on endurance and sustained performance. Linxai Intelligent’s Lingshi team won in 1 hour 51 minutes 29 seconds, followed by teams from Shanghai Gaoyi Technology and Shenzhen Technology University. The humanoid robot race was adjusted because of the demanding terrain, covering Hongmen Square to Doumu Palace; the Zhishen Lingyue team claimed first place in 27 minutes 26 seconds.
Beyond the race results, the competition exposed the technical limitations of robots operating in real-world environments. In laboratories, robots typically work on standardized surfaces under controlled conditions. Mount Tai was a completely different story. Uneven steps required more precise terrain perception and gait planning, while continuous uphill climbs tested dynamic balance and joint control. Prolonged high-intensity operation also put considerable pressure on motor cooling, battery life and overall system stability.
Robots also had to deal with unexpected obstacles, avoid them autonomously and recover independently after losing balance or falling—capabilities that become essential when machines move beyond controlled environments. During the competition, some robots experienced motor overheating, automatic power reductions and sideways posture deviations. Several load-bearing robots also developed unstable gaits after changing payloads at supply points.
While these may look like minor setbacks during a race, they highlight the engineering challenges legged robots still face when operating over long distances, carrying heavy loads and navigating complex terrain.
After all, walking steadily on a flat surface does not mean a robot is ready for outdoor work. Maintaining stability on uneven terrain, balancing payload capacity with battery life and movement efficiency, and adapting autonomously to unexpected situations are key challenges embodied AI must overcome before moving from impressive demonstrations to practical applications.
The event also featured the Mount Tai Embodied Intelligence Industry Summit, a robotics industry-chain exhibition and an investment matchmaking event. Together, these activities gave companies, research teams and investors a platform to exchange ideas, showcase products and explore potential partnerships.
The event was about much more than organizing a robot climbing competition. Mount Tai’s rugged trails offered a rare real-world testing environment where robots could collect operational data under continuous uphill climbs, complex terrain conditions and extended periods of operation. The results could help guide future product development and provide insights into the potential use of humanoid and legged robots in specialized inspection, mountain rescue and outdoor transportation.
As robots take on Mount Tai, competition in embodied AI is increasingly shifting from laboratory demonstrations to real-world performance. The next challenge for these machines is not simply proving they can walk, run or climb, but showing they can perform tasks reliably and consistently in complex environments.