Xingzhe Taishan, a humanoid robot developed by Jinan-based Yobotics, took second place in the humanoid robot racing event at the inaugural CMG World Robot Mount Tai (Taishan Mountain) Climbing Competition on Nov 2.

Held in Tai'an, Shandong province, the competition attracted nearly 90 teams and turned Taishan Mountain's rugged trails into a testing ground for robotics in real-world conditions. Billed as the world's first robotics competition staged on natural mountain terrain, it featured four official events – humanoid robot racing, quadruped robot racing, quadruped robot load-bearing, and exoskeleton-assisted climbing – plus an all-terrain mobility challenge.

The humanoid robot race covered about 1.6 kilometers and more than 600 steps, with steep slopes, winding paths, and uneven stone surfaces. Some steps were so narrow that they could not accommodate a robot's entire foot, demanding precise movements and reliable balance.

In preparation, the team tested different sole materials before adopting a biomimetic composite design that offers better grip and durability. A new generation of high-power-density servo motors increased the robot's maximum leg-lifting height from 20 to 30 centimeters, enabling it to tackle steps of varying heights.

To improve terrain recognition, the team integrated visual, LiDAR, and foot-pressure sensors, trained a specialized model on tens of thousands of images from Taishan Mountain, and refined foot placement through trajectory planning. It also optimized the robot's hardware, algorithms, and mechanical structure to improve energy efficiency, achieving more than two hours of operation in relatively flat outdoor conditions rather than relying on a larger battery.

Xingzhe Taishan climbs stone steps during the CMG World Robot Mount Tai Climbing Competition in Tai'an, Shandong province. [Photo/The New Times]