On this page
China is accelerating
In April 2025, something unprecedented happened in Beijing. Twenty humanoid robots lined up alongside 12,000 human runners for the world's first robot...
In April 2025, something unprecedented happened in Beijing. Twenty humanoid robots lined up alongside 12,000 human runners for the world's first robot half-marathon. Some robots stumbled. Others ran smoothly for hours, switching batteries mid-race. One small machine, barely three feet tall, waved enthusiastically at the crowds as it completed the full 21 kilometers.
A decade earlier, the United States led the world in humanoid robotics. Government programs poured resources into research. American universities and companies dominated the field. Then progress stalled. Funding dried up. The field entered what some called a "humanoid winter." While American innovation slowed, China was building.
Around 2021, interest reignited. Tesla unveiled its Optimus prototype. Figure AI launched in the United States. But while America was reawakening, China had already been at it for years.
On January 28, 2025, during the Spring Festival Gala, China's most-watched television broadcast with over a billion viewers, 16 Unitree H1 humanoid robots took the stage. Dressed in traditional floral padded jackets, they performed the Yangge folk dance alongside human dancers from the Xinjiang Art Institute. The robots spun handkerchiefs, kicked their legs, twisted their waists, and caught the handkerchiefs mid-air with mechanical precision. The performance, directed by filmmaker Zhang Yimou, was powered by Alibaba Cloud technology and represented what Unitree called "the world's first large-scale, fully AI-driven and fully automated cluster humanoid robot performance in history."
A National Mandate#
In March 2025, China's National Development and Reform Commission announced a state-backed venture capital fund expected to attract nearly 1 trillion yuan ($138 billion) over the next 20 years, focused on robotics, AI, and cutting-edge innovation. This follows years of strategic planning. In 2023, China's Ministry of Industry and Information Technology identified humanoid robotics as a "frontier technology" with a target of achieving mass production and secure supply chains by 2025. The stated goal: become the world leader in humanoid robotics by 2027.
China's push comes from demographic necessity. The one-child policy, implemented from 1979 to 2015, drastically reduced birth rates while life expectancy soared alongside spectacular economic growth. By 2035, more than 400 million Chinese citizens will be over 60 years old, representing roughly 30 percent of the population. The working-age population peaked in 2015 and has been shrinking by about 5 to 6 million annually. Projections suggest a shortage of 30 million workers across 10 key manufacturing sectors by the end of 2025.
The crisis compounds at the family level. Many young Chinese workers face the "4-2-1" problem: one child supporting two parents and four grandparents. Rural-to-urban migration has eroded traditional family care systems, leaving millions of elderly without adequate support. Pension systems in 11 of China's 31 provincial jurisdictions are already running deficits. The state-run Chinese Academy of Sciences predicts the pension system could run out of money by 2035.
China views humanoid robots as a partial answer to this demographic crisis, envisioning adaptive systems that can provide elder care, detect health changes, and offer basic support.
The Shenzhen Advantage#
Shenzhen, often called the "Silicon Valley of hardware," sits just across the border from Hong Kong. This sprawling city is home to major manufacturers including Foxconn, which assembles Apple's iPhones. Sources estimate that Shenzhen produces between 35 and 90 percent of the world's electronics, depending on how categories are defined. What is undisputed: Shenzhen's supply chain infrastructure is unparalleled.
The city's Huaqiangbei Electronics Market spans over 70 million square feet across 20 shopping malls, housing thousands of small vendors selling every conceivable component: sensors, motors, actuators, circuit boards, microchips, and cameras. Need a custom part? Shenzhen's factories can deliver it in days, not months. This speed enables rapid prototyping and iteration at a scale Western companies cannot match.
Chinese robotics firms like UBTech, Fourier Intelligence, and Unitree leverage this ecosystem for cost and speed advantages. The same infrastructure is why smartphone maker Xiaomi can pivot to electric vehicles, or why EV company XPENG can develop humanoid robots. The supply chain supports rapid pivots across hardware categories.
In 2025, investment in China's embodied intelligence and robotics sector surged. According to data from IT Juzi, there were 114 investment events in the first five months of 2025 alone, surpassing the 77 deals recorded in all of 2024. Funding reached 23.2 billion yuan (approximately $3.2 billion), exceeding the previous year's 20.9 billion yuan ($2.9 billion). Companies like Unitree Robotics raised 700 million yuan ($97 million), while Galbot secured 1.1 billion yuan ($153 million). Fourier Intelligence closed a Series E round of nearly 800 million yuan ($111 million) in January 2025. A Morgan Stanley analysis found that 56 percent of the 100 publicly traded companies worldwide confirmed to be developing humanoid robots are based in China.
The contracts followed the capital. UBTech secured a 250 million yuan ($35 million) contract for its Walker S2 robots, bringing the Walker series to nearly 4 billion yuan ($556 million) in total contracts. Unitree and Fourier Intelligence jointly won a 124 million yuan ($17 million) contract from a China Mobile subsidiary. Unitree's CEO announced that the company had achieved $140 million in annual revenue as of June 2025. Huawei and UBTech signed a comprehensive cooperation agreement to develop joint solutions for "humanoid robots plus smart factories," leveraging Huawei's Ascend chips, cloud infrastructure, and large AI models.
Deployment at Speed#
While Western companies focus on perfecting technology in controlled environments, Chinese firms deploy rapidly in the real world. Unitree Robotics sells its G1 humanoid robot for approximately $16,000, dramatically undercutting Western competitors. UBTech's Walker S1 robots, launched in late 2024, have been deployed for training in automotive factories including Dongfeng, BYD, and Foxconn. By the end of 2024, UBTech had achieved the highest number of deployed humanoid robots for training in automotive factories worldwide. Fourier Intelligence introduced the GR-2 in late 2025, featuring 53 degrees of freedom and enhanced battery life, while also launching the N1, its first open-source humanoid robot designed to accelerate global adoption.
The strategy mirrors China's approach to electric vehicles and drones: subsidize development, build domestic demand, iterate rapidly through deployment, and eventually export globally. Today, Chinese humanoid robots are marketed for entertainment and light industrial tasks. Tomorrow, they will compete globally across manufacturing, logistics, and care.
The April 2025 Beijing marathon crystallized this approach. More than 20 Chinese companies participated, each testing their robots over 21 kilometers of outdoor terrain with slopes, turns, and unpredictable conditions. Some robots failed. Others performed impressively. The winning robot, Tiangong Ultra, developed by the Beijing Innovation Center of Humanoid Robotics, completed the course with only three battery changes. According to the team's chief technology officer, "No other robotics firms in the West have matched Tiangong's sporting achievements."
Whether or not that claim holds up, the signal is real: China is testing publicly, failing publicly, and iterating fast.
Setting the Rules#
In November 2025, China's Ministry of Industry and Information Technology announced the formation of a Standardisation Technical Committee for Humanoid Robots. The 65-member body reads like a who's who of Chinese robotics: Unitree founder Wang Xingxing and AgiBot co-founder Peng Zhihui were appointed deputy chairs. The roster includes executives from UBTech, Fourier Intelligence, Galbot, and LimX Dynamics, alongside representatives from Huawei, ZTE, Baidu, and XPENG. Scholars from Tsinghua University and Shanghai Jiao Tong University provide research backing.
The committee will set national standards for everything from battery safety to data interfaces, potentially giving Chinese manufacturers a unified framework that international competitors currently lack. In the West, organizations like the IEEE (Institute of Electrical and Electronics Engineers) have only recently published roadmaps to begin the standardization process. An ISO (International Organization for Standardization) working group met in Barcelona in late 2025 to advance safety standards, a process insiders admit is lagging behind China's deployment schedule.
Whoever sets the standards shapes the market. China's top-down approach, bringing competitors like Unitree and AgiBot onto the same committee, forces the collaboration necessary to build shared industrial baselines. Beijing is betting that unified standards will accelerate domestic production while creating technical barriers for foreign competitors.
What This Means#
American and European companies face a direct choice: match China's pace and investment, or differentiate on capabilities that China cannot easily replicate. For the United States, that likely means leveraging advantages in AI research, software platforms, and premium market segments. For Europe, the options are narrower.
Dr. Dietmar Ley, Chairman of VDMA (German Mechanical Engineering Industry Association) Robotics + Automation, warned in March 2025 that "Europe must not lag behind in this critical area. It is essential that European humanoid technology moves beyond the labs and into scalable, competitively priced production."
China's coordinated strategy, combining state funding, manufacturing infrastructure, aggressive pricing, rapid deployment, and now standards-setting, is something few nations can match. And beneath the geopolitical race lies a structural question that the next chapter addresses: will the humanoid industry be dominated by vertically integrated companies that control everything from chips to software to service, or will it evolve into an open ecosystem where thousands of companies build on shared platforms? The answer shapes not just who wins, but how the entire industry organizes itself.