The Decade of Humanoid Robots: Why Does It Belong to China?

```

2025 is the inaugural year when humanoid robots transition from demonstrations to large-scale deployment. This year, approximately 15,000 humanoid robots were installed worldwide, 85% of which were delivered by Chinese companies. Meanwhile, the annual installation of industrial robots in China is close to 300,000 units, about nine times that of the United States.

This is not an accidental eruption, but the result of years of systematic accumulation. According to Wind Trading Desk and Barclays analyst Zornitsa Todorova’s macro research, she makes a direct assessment: The decade of robots belongs to China.

This conclusion is backed by China’s near-vertical integration of the entire industrial chain—over 90% of refined magnetic rare earths, 45% of battery exports, 22% of actuator exports, and about 70% of global robot patent applications are concentrated in China. This means that it is almost impossible for Western countries to bypass China to independently establish local industrial chains in the short term.

Based on production forecasts, using the expansion curve of Chinese electric vehicles as a reference, by 2035 the annual deployment upper limit of humanoid robots in China could reach about 11 million units, with a global total of about 13 million units.

If the cumulative inventory reaches 24 million, this is equivalent to increasing China’s effective labor supply by about 3.8%, offsetting nearly 60% of the predicted decrease in labor force. This is an optimistic scenario, but not unfounded—it almost replicates the path of electric vehicles from policy support to explosive growth.

Patent Accumulation: China’s Innovative Engine Shifted Gears Around 2015

Analysis of over 300,000 robot-related patents since 2000 using the European Patent Office PATSTAT database shows that China accounts for about 70% of global robot patent applications, while the United States has only about 4%.

In 2023 alone, China added over 30,000 new robot patent inventions, nearly 30 times that of the United States in the same period.

The source structure is particularly noteworthy.

One third of Chinese robot patents come from universities and national research institutes, compared to about 10% in the U.S. National strategies such as "Made in China 2025" explicitly position universities as the core of technology R&D, establishing an integrated mechanism of industry-academia-research development.

From the perspective of patent content, the two countries’ technological focuses are distinctly different: Chinese patents center on kinematics and physical structures (hardware), while the U.S. focuses more on control and autonomy (software and perception).

For short-term large-scale manufacturing, China’s hardware dominance more directly determines who controls production; but at the level of system intelligence, the U.S. technology base forms differentiated competition.

In the humanoid robot segment, China holds about 75% of global related patents, with innovation accelerating around 2015 and a clear gap opening by 2018. UBTech leads globally with 812 cumulative patents, far surpassing Boston Dynamics’ 119.

Policy Is the Architecture, Subsidies Are Just One Link

China’s robot policy is not a project list, but an “all-stack” industrial system design.

From "Made in China 2025" in 2015 to the 14th and 15th Five-Year Plans, and the 2023 “Robotics+” action plan, government documents have defined robotics and embodied intelligence as strategic foundational industries and promoted coordinated advances at all levels—raw materials, energy, components, production, and deployment—shaping both supply and demand.

The "Robotics+" plan identifies 10 priority deployment industries, including manufacturing, healthcare, agriculture, elderly care, construction, etc., marking the shift of policy focus from "building supply" to the second phase of "creating demand".

The effect is already shown in the numbers: as of 2024, China’s industrial robot stock surpassed 2 million units, reaching the "Made in China 2025" goal two years ahead of schedule; the goals in the 14th Five-Year Plan for “annual revenue growth of 20%” and “doubling robot density” have also been met ahead of time.

Supply Chain Control: A Deeper Moat Than Policy

China's core advantage in humanoid robots is not just being the first to mass-produce, but vertical integration of the entire supply chain.

Raw materials: Over 90% of the world's refined rare earth magnets (key for robot electric joints) come from China.

China is also the world’s largest processor of refined cobalt, lithium, and copper, and ranks second in nickel processing. In parts: China accounts for 22% of global actuator exports, 45% of battery exports, and is second in semiconductor exports worldwide.

The implication: every humanoid robot assembled in the U.S. or Europe may still depend on Chinese-made key parts.

Strategic vulnerability lies not in end products but in components, raw materials, and software—this is also the real obstacle the West faces in trying to build "de-risked" supply chains.

Changes in Labor Structure Are Creating Long-term Demand Support

As demographic structures continue to change, with slowing labor supply growth and increasing aging, the supply-demand balance in some sectors is facing adjustment pressure.

Even with continuous productivity improvements, this effect may not be fully offset.

This means robots are shifting from mere efficiency tools to important supplements that help relieve labor mismatches and support industrial operation. Especially in standardized, highly repetitive scenarios, long-term demand is gradually taking shape.

Refer to the EV Curve: Annual Deployments Could Reach 13 Million Units by 2035

The projection for mass production of humanoid robots uses the Chinese EV penetration path as a reference, for specific reasons:

The two share identical core components (semiconductors, actuators, gearboxes, batteries), both rely on large-scale integration and scale learning, were initially highly concentrated in China, and both enjoyed powerful policy support.

The logic: starting from about 15,000 units in 2025 and the industry consensus anchor of 60,000 units in 2026, following the accelerated EV penetration curve which led to about 11 million annual sales in year ten, the estimate is that China could deploy about 11 million units in 2035.

If China maintains its 85% global share, the worldwide deployment that year could reach about 13 million units, with 2 million outside China, mainly in the United States.

This is an upper-bound scenario, not a baseline forecast.

But there is basis in real-world company plans: Tesla has announced it will convert its Fremont factory for humanoid robot production; Boston Dynamics aims to produce 30,000 Atlas units annually by 2028; Figure AI has set a target of 100,000 units for the same period.

Top companies together could reach about 200,000 units outside China between 2028-2029. If exponential growth starts in 2030, the ten-year goal of 2 million units is not out of reach.

According to Wallstreetcn, Morgan Stanley points out that since early 2026, global humanoid robot VC investment has already surpassed the total for all of 2025, with such rapid growth drawing industry attention—of which about 46% of annual VC funds came from China, making it the world’s most active capital battleground.

The firm particularly notes that China is perfectly replicating its EV success story, and by laying out the whole industry chain and its amazing iteration speed, is expected by 2030 to push its share of global manufacturing to 16.5%.

It is worth noting why industrial robots and smartphones are not good references: industrial robots automate processes; humanoid robots replace workers.

The ceiling for the former is determined by the number of automatable tasks, but for the latter, the potential market is the entire labor force—a scale not comparable to the former.

The West Is Not Out, But Must Redefine Its Competitive Approach

China’s advantage is currently huge, but this does not mean there is no room for the West—just that the logic of competition is entirely different.

U.S. manufacturing accounts for only about 9% of GDP, so near-term demand in industrial humanoids cannot match China. For the next 5–7 years, U.S. deployments are more likely in service industries, where commercial maturity won’t rise markedly before the 2030s. U.S. labor force is also expected to still grow to 141 million by 2035, so there is no structural urgency like in China.

Safety has also become a key reason for higher Western barriers for robot adoption.

This leaves a window for Western manufacturers. Major Western players include Tesla, Figure AI, Agility (U.S.), Hexagon AB (Sweden), and Boston Dynamics (Korea).

But the ecosystem remains highly fragmented—a large number of firms are small, software-oriented, and weak in manufacturing. Over the next 3–5 years, industry consolidation and M&A are highly likely. Only by reaching critical scale can they truly compete with China in cost and supply chain.

Risk Disclosure and DisclaimerThe market has risks; investment requires caution. This article does not constitute personal investment advice, nor does it take into account individual users’ specific investment objectives, financial situations, or needs. You should consider whether any opinions, views, or conclusions in this article suit your particular circumstances. You are solely responsible for any investments based on this information. ```