Tiangong Ultra's Victory and Yushu G1’s Hesitation: An Analysis
20 April 2025 · Uncategorized ·
On the morning of April 19th in Beijing’s Economic-Technological Development Area (Yizhuang), a groundbreaking global first—a humanoid robot half-marathon—took place, quickly becoming one of this weekend's most talked-about events. The competition highlighted companies with robots demonstrating speed and endurance, unique designs, and engaging performances, captivating audiences. Beijing Humanoid Robot Innovation Center’s ‘Tiangong Ultra’ emerged as the dominant victor, capturing widespread attention; however, Yushu Technology G1, a highly anticipated contender, faced challenges throughout the race, occasionally pausing or deviating from its course. Songyan Dynamics N2 secured second and third place with an innovative split-strategy approach. Other noteworthy observations included robots requiring hydration breaks and protection from sunburns.
Following the event, industry experts began to analyze what these results signify for humanoid robotics development both in China and globally: Why were so many strong competitors absent? How advanced is embodied intelligence technology currently at a national and international level? And what commercial or industrial applications are now viable for these robots?
In fact, Tiangong Ultra’s victory was widely anticipated by industry insiders.
Firstly, Beijing Humanoid Robot Innovation Center possesses an inherent advantage due to its location in Yizhuang—one of China's most important robot agglomeration areas. This region offers complete supply chains and robust manufacturing capabilities for humanoid robots, fostering a cluster of innovative enterprises around the center.
Secondly, the innovation center was jointly founded by prominent entities including Xiaomi Robotics, Shenzhen-based Embodied Intelligence Walker (which subsequently relocated to Yizhuang), and Beijing Machinery Industry Group. Supported by substantial capital investment and strong research & development capabilities, they are well-positioned within this field; further demonstrating their commitment, Xiaomi’s Mi Auto SU7 served as the official timing vehicle for the event.
Consequently, given that Cyberone from Xiaomi Robotics and Walker from Embodied Intelligence focused on optimizing Tiangong Ultra with combined resources—and did not participate in the race themselves—it would have been surprising if any other robot had prevailed.
Regarding Yushu Technology G1’s less-than-expected performance, several factors may be at play. Some industry insiders suggest that its showing doesn't necessarily reflect a deficiency in intelligence or balance; they highlighted impressive actions such as independently rising from hands and knees during the competition.
Another potential reason is insufficient adaptation to this specific racing scenario for Yushu Technology G1, which was one of only a few teams participating outside of Yizhuang. Notably, Xiaomi has also invested in Yushu Tech.
A Shenzhen-based researcher told Observer that currently, innovative enterprises across China possess relatively comparable research and manufacturing capabilities but primarily differ in strategic positioning and prioritization of application scenarios.
From the perspective of long-term investors or companies focused on industrial applications, commercializing robots to replace repetitive or hazardous tasks can significantly enhance production efficiency; therefore, giants like Huawei and Midea have taken a leading role by integrating embodied intelligence into various industries.
A researcher from Shanghai emphasized that China should prioritize applying humanoid robotics in practical scenarios rather than blindly following the footsteps of Western technology leaders. He also noted the need to incorporate critical components for improved performance under specific conditions—most participating models were relatively small, resembling children’s sizes due to technical and cost considerations.
Following the event, industry experts began to analyze what these results signify for humanoid robotics development both in China and globally: Why were so many strong competitors absent? How advanced is embodied intelligence technology currently at a national and international level? And what commercial or industrial applications are now viable for these robots?
In fact, Tiangong Ultra’s victory was widely anticipated by industry insiders.
Firstly, Beijing Humanoid Robot Innovation Center possesses an inherent advantage due to its location in Yizhuang—one of China's most important robot agglomeration areas. This region offers complete supply chains and robust manufacturing capabilities for humanoid robots, fostering a cluster of innovative enterprises around the center.
Secondly, the innovation center was jointly founded by prominent entities including Xiaomi Robotics, Shenzhen-based Embodied Intelligence Walker (which subsequently relocated to Yizhuang), and Beijing Machinery Industry Group. Supported by substantial capital investment and strong research & development capabilities, they are well-positioned within this field; further demonstrating their commitment, Xiaomi’s Mi Auto SU7 served as the official timing vehicle for the event.
Consequently, given that Cyberone from Xiaomi Robotics and Walker from Embodied Intelligence focused on optimizing Tiangong Ultra with combined resources—and did not participate in the race themselves—it would have been surprising if any other robot had prevailed.
Regarding Yushu Technology G1’s less-than-expected performance, several factors may be at play. Some industry insiders suggest that its showing doesn't necessarily reflect a deficiency in intelligence or balance; they highlighted impressive actions such as independently rising from hands and knees during the competition.
Another potential reason is insufficient adaptation to this specific racing scenario for Yushu Technology G1, which was one of only a few teams participating outside of Yizhuang. Notably, Xiaomi has also invested in Yushu Tech.
A Shenzhen-based researcher told Observer that currently, innovative enterprises across China possess relatively comparable research and manufacturing capabilities but primarily differ in strategic positioning and prioritization of application scenarios.
From the perspective of long-term investors or companies focused on industrial applications, commercializing robots to replace repetitive or hazardous tasks can significantly enhance production efficiency; therefore, giants like Huawei and Midea have taken a leading role by integrating embodied intelligence into various industries.
A researcher from Shanghai emphasized that China should prioritize applying humanoid robotics in practical scenarios rather than blindly following the footsteps of Western technology leaders. He also noted the need to incorporate critical components for improved performance under specific conditions—most participating models were relatively small, resembling children’s sizes due to technical and cost considerations.