Beijing’s robot half-marathon wasn’t just spectacle – it reveals China’s intent to dominate the emerging humanoid robotics sector and lead on a global level.
On April 19, just months after Unitree’s dancing humanoid robots captured international headlines for their appearance in the CCTV Spring Festival Gala, Beijing hosted the world’s first-ever humanoid robot half marathon. The event drew thousands of live spectators along the course and millions more watching via the CCTV livestream and online platforms, quickly becoming the most widely viewed robotics competitions in history.
Co-organized by Beijing municipal government bureaus, the Beijing Yizhuang Half Marathon and Humanoid Robot Half Marathon marked a milestone in China’s push to accelerate humanoid robot development. The event featured 1,200 human runners alongside 20 robot teams from private companies and state-backed projects, including industry leaders Unitree and X-Humanoid. While humans and robots raced on separate tracks, the event podium featured top finishers in both categories — symbolizing Beijing’s ambition to integrate humanoid robots into everyday life.
The results revealed that the performance gap between humans and robots remains substantial. While the men’s winner finished in 1:02:36 and the women’s champion in 1:11:07, Tiangong Ultra, the leading robot, completed the course in 2:40:27. Despite this gap, the event demonstrated the significant technological progress Chinese companies have made in relation to their U.S. counterparts. Tiangong Ultra maintained an average pace of 8.2 km/hr and reached speeds of up to 12 km/hr — surpassing the maximum 8 km/hr demonstrated by Tesla’s much-hyped Optimus. Meanwhile, although Boston Dynamics’ Atlas is faster than Tiangong Ultra, its battery life is much shorter when performing at maximum capacity.
“This represents a technological and engineering milestone,” noted ZongZe Wu, a research associate at Tsinghua University’s Institute for Artificial Intelligence International Governance. “Although humanoid robots still have a long way to go, simply completing a course of this distance is a remarkable feat.”
This sentiment resonated broadly across Chinese media and social platforms. Sam Zhu, a 16-year-old tech enthusiast from Wuhan, captured the public mood saying, “This feels like witnessing the start of something transformative — potentially as significant as the early stages of the Industrial Revolution.”
Although the event lasted only hours, its symbolism extends far beyond sport. For policymakers and technologists across China, the half-marathon represents a tangible demonstration of the country’s accelerating capabilities in “humanoid robotics” (人形机器人) and “embodied intelligence” (具身智能). It also marked a turning point in a field Beijing has explicitly prioritized as central to its vision of an AI-powered economic transformation – and the beginning of a new era in physical AI applications.
China’s Humanoid Robot National Strategy: From Policy to Execution
In November 2023, the Chinese government laid out its first official development plan for humanoid robots with ambitious two-stage targets. The initial phase, with a target date of this year, focuses on technical breakthroughs, industrial deployment, and cultivating globally competitive enterprises. So far, these goals that already appear to have been largely met.
China has developed advanced movement control technology, and launched “Gewu,” an open-source platform that trains over a hundred robot variants with a single codebase. Additionally, according to Morgan Stanley’s 2025 Humanoid 100 Report, Chinese companies now represent 35 of the top 100 firms in the humanoid robotics value chain and nine of the 22 companies capable of producing fully integrated humanoid robots — compared to just five in the United States.
The second stage aims to establish a secure industrial supply chain, integrate robotics throughout the economy, and achieve global technological leadership. This also seems to be largely on track with a 2024 analysis by the China Academy of Information and Communications Technology (CAICT) reporting that Chinese firms already “cover the entire industrial chain” from components to systems integration.
The national strategy has triggered a cascade of regional initiatives. In March, Shenzhen released an ambitious 2027 action plan targeting 100-plus billion yuan ($13.7-plus billion) in robotics-related industries while expanding its embodied intelligence cluster to include over 1,200 companies. Beijing quickly followed by launching a dedicated humanoid robot data training center with project funding subsidies of up to 24 million yuan per company for qualified R&D initiatives. Within weeks, at least ten other provinces introduced similar policies supporting humanoid robot development, uniformly describing the sector as “a key driver for high-quality economic growth and industrial upgrading.”
This coordinated push reflects how Chinese policymakers view humanoid robotics not merely as a technological showcase but as a strategic solution to the country’s most pressing structural challenges. With youth unemployment hovering around 17 percent and the working-age population projected to contract significantly by 2030, Beijing is urgently seeking new engines of productivity and economic growth. Humanoid robotics offers a compelling two-pronged approach: simultaneously driving industrial modernization while addressing emerging labor shortages in elder care, logistics, manufacturing, and service sectors—areas critical to China’s economic stability as its demographic crisis deepens.
The Competitive Edge: Why China Could Lead in Real-World Robotics
With the increased focus on embodied intelligence, a critical question emerges: will China become the first country to deploy commercially viable humanoid robots at scale? Xu Huazhe, an assistant professor at Tsinghua University’s Institute of Interdisciplinary Information Sciences, argues this is increasingly likely. In a December 2024 interview, he pointed to China’s distinctive advantage: not merely in technological development, but in the capacity to rapidly deploy and iterate these technologies in real-world environments.
China’s potential edge begins with its unmatched manufacturing ecosystem. The country’s robust industrial base allows robotics firms to draw directly from adjacent sectors with relevant expertise. Over the past year, major players like battery giant CATL and thermal management specialist Sanhua — both central to China’s electric vehicle revolution — have expanded into humanoid robotics, becoming key nodes in the emerging supply chain. New entities are similarly leveraging traditional manufacturing strengths. The Beijing Humanoid Robot Innovation Center, established in 2023 by UBTech and various partners, is developing core technologies through strategic alliances with specialized component producers like Jingcheng Machinery Electric, Tsino Dynatron, and RobStride. This interconnected ecosystem enables vertical integration at unprecedented speed and scale — exemplified by Unitree Robotics, which now independently develops and manufactures crucial components including servo motors, reducers, and controllers that would typically require years of supply chain development elsewhere.
A second crucial advantage lies in data access. Humanoid robots demand enormous volumes of multimodal training material — from visual inputs to body movements and environmental interactions. While high-quality datasets remain limited globally, Chinese firms benefit from both a comparatively permissive regulatory environment and a government that serves simultaneously as funder, customer, and data provider. Research by economists Noam Yuchtman and Yang Yifan demonstrated how Chinese AI firms systematically gain access to rich behavioral datasets through government procurement contracts that their Western counterparts cannot easily obtain. While data quantity alone doesn’t guarantee technological superiority, access to diverse real-world information and feedback environments demonstrably accelerates development cycles and deployment readiness.
The final catalyst in China’s embodied intelligence push is concrete market demand. The country’s elder care robotics market reached 7.9 billion yuan in 2024 and is projected to grow at a 15 percent annual rate, approaching 16 billion yuan by 2029. Simultaneously, industrial adoption shows remarkable momentum — a recent industry survey found that 99 percent of Chinese industrial robot users anticipate near-term demand for humanoid models, primarily for quality control and monitoring applications. This immediate factory-floor demand creates a particularly valuable dynamic: once deployed, these robots generate additional real-world operational data, initiating a virtuous cycle of continuous improvement that further accelerates development.
Collectively, China’s manufacturing prowess, permissive data environment, and tangible market demand create a foundation for embodied intelligence development and deployment unlike anywhere else in the world — potentially allowing Chinese firms to compress the timeline from prototype to mass commercialization in ways U.S. competitors may struggle to match.
Beyond Technology: A Different Understanding of Competition
For the U.S., China’s advances in AI and humanoid robotics have primarily been framed as direct competition — a technological arms race to achieve breakthrough capabilities first. This framing has driven Washington’s response, with the Department of Commerce implementing aggressive chip export controls and restricting American investment in certain Chinese AI infrastructure. While these measures have constrained Chinese firms’ ability to train large-scale AI models, they haven’t fundamentally altered Beijing’s trajectory, as evidenced by DeepSeek’s recent AI breakthroughs despite chip limitations.
The impact on humanoid robotics may be even less decisive, as these embodied technologies rely more on physical engineering and real-world deployment than pure computational power. This highlights a fundamental misalignment in how each nation approaches the competition. While American policymakers focus narrowly on maintaining leadership in frontier AI research and military applications, China appears to be playing a different game entirely.
From Beijing’s perspective, the objective extends well beyond building the most advanced humanoid robot. China’s approach focuses on practical, near-term applications that address specific societal challenges. In Beijing’s Zhongguancun International Innovation Center, the humanoid robot Kuavo already serves as a receptionist. In Fuzhou, “smart nannies” assist caregivers in a senior living facility. In Chengdu, authorities will soon launch a “wellness robot” program targeting preventative healthcare for elderly residents. These applications reflect China’s pragmatic strategy: deploying currently viable technology to address pressing demographic and economic needs rather than pursuing purely theoretical advances.
Equally significant is China’s parallel focus on establishing governance frameworks for these technologies. Science and Technology Minister Yin Hejun declared earlier this month that China must become a “leading force in shaping scientific advancement and governance frameworks,” echoing the 2023 development plan that positions governance as integral to embodied intelligence development. This isn’t mere rhetoric — China has implemented at least nine distinct policies regulating AI development since 2021, creating the world’s most comprehensive regulatory environment for autonomous systems.
The strategy extends internationally as well. China’s technical guidelines for humanoid robot safety standards have already influenced discussions at International Organization for Standardization (ISO) and International Electrotechnical Commission (IEC) working groups, positioning Chinese approaches at the center of emerging global frameworks. This raises a fundamental question: is humanoid robotics truly a winner-takes-all competition between Washington and Beijing, or does China’s multidimensional approach suggest a different understanding of the race?
Reframing the U.S. Response: From Containment to Engagement
It seems that there is a fundamental misalignment in how each nation conceptualizes the competition, and a more effective American response would require a strategic pivot. Rather than maintaining its defensive posture, the U.S. should leverage its strengths to compete more effectively in global markets. This could begin with establishing export initiatives supporting American robotics companies in emerging economies and reducing market access barriers through trade agreements. By strategically increasing exports of robotics technology and training systems to key markets like Brazil, India, and Southeast Asia — where industrial robot adoption has already increased nearly tenfold in the past 15 years — the U.S. could simultaneously advance its economic interests while countering China’s expanding technological influence across the Global South.
The U.S. must also increase its engagement in international standards bodies, where today’s technical specifications will determine tomorrow’s market access. American participation in the ISO/IEC Joint Technical Committee on Artificial Intelligence — where more than 60 countries develop AI standards — remains inconsistent compared to Chinese delegations. Similarly, the IEEE Robotics and Automation Society’s Standards Committee, which establishes protocols for emerging capabilities in humanoid robotics, has seen declining American industry participation even as Chinese representation grows. Strategic re-engagement with coordinated public-private participation would help ensure democratic values and open-market principles guide the global regulatory environment.
China’s robot marathon wasn’t merely a technological showcase—it was a declaration of strategic intent. While the United States focuses on containing China’s AI capabilities, Beijing is building an ecosystem for humanoid robots that spans from research labs to factory floors to international standards bodies. The true measure of success in this competition won’t be which country builds the most advanced robot first, but which most effectively integrates these technologies into society while establishing governance frameworks that balance innovation with human welfare. And this race — with far higher stakes than any marathon — has only just begun.
