- Статьи
- Economy
- Robots are working hard: China has become the hegemon in the market of humanoid machines
Robots are working hard: China has become the hegemon in the market of humanoid machines
Shares of the Chinese company Unitree soared by 460% on the first day of trading on the Shanghai Stock Exchange. Starting at 150.8 yuan, the quotes reached 845 yuan, allowing the developer to raise more than $900 million. The success on the stock exchange is not accidental: by the end of 2025, the company shipped over 5.5 thousand humanoid robots to customers, taking the first place in the world with a market share of 32.4%, which quite clearly illustrates the growing gap between China and the rest of the world in this sector. Whether China will become the same dominant force in robotics as in green energy, rare earth metals mining and other industries, and what this may mean for the world, is in the Izvestia article.
Competition is above all
American corporations are constantly showing investors spectacular commercials with dancing prototypes of robots of various types. Meanwhile, Asian manufacturers have already switched to mass commercial assembly. Today, China accounts for more than 80% of the world's humanoid robot installations. Six of the world's top ten companies in the sector are Chinese.
China's dominance in this market was difficult to predict ten years ago. In 2016, the undisputed industry leaders were the American Boston Dynamics with its Atlas devices and the Japanese corporations Honda and Toyota. The Chinese developments of that time looked like clumsy copies, completely dependent on imported servos and harmonic reducers from Japan and Europe. To be fair, at that time the industry as a whole was still very raw and had an extremely limited scope of applications, and its market volume barely reached $100 million worldwide.
The leadership takeover scenario exactly repeated the strategy previously applied by Beijing in the market of electric vehicles and solar panels. The government has allocated huge subsidies to form a closed supply chain within the country. The focus was on the mass production of key components: powerful rare earth magnets, compact electric motors, lidars and vision systems. Considerable funds were allocated, and intense competition between different producers was encouraged through a system of national and regional incentives.
Western (mostly American, of course) engineers have been trying for years to create an ideal, versatile, and super-expensive machine for hundreds of thousands of dollars. Chinese developers have taken the path of cheapening, launching devices worth from $15 to $30 thousand on the market. This price made the payback of the machines understandable to the owners of warehouse complexes and factories. And the transition from piece-by-piece manual assembly to an automated conveyor allowed Chinese brands like Unitree, UBTECH and Fourier Intelligence to leave Western competitors far behind in terms of shipments.
The AI came to the rescue
However, the real explosion in the market occurred due to the integration of generative artificial intelligence (AI). Previously, robots were programmed mechanisms capable of moving only along strictly defined routes in a controlled environment. Any obstacle or change in lighting caused a malfunction.
Modern so-called VLA models have endowed machines with spatial understanding. Today, a humanoid is able to perceive voice commands in natural language, recognize objects around him and independently build a trajectory of movement. Neural network training has moved to the stage of embodied intelligence.
In the commercial sector, these devices already perform routine but in-demand tasks. They patrol hazardous chemical industries, inspect high-rise power transmission lines, sort cargo at logistics hubs, and control the quality of parts on carmaker assembly lines.
At the same time, the capabilities of robots are still limited. Machines have serious difficulties with fine motor skills and manipulations with soft objects. Actions that require tactile feedback—tying knots, working with thin wires, or cleaning chaotically scattered small objects — remain an impossible task for them. However, the existing functionality is already sufficient to replace line staff in heavy monotonous jobs in a number of sectors. This is also evident in the volume of sales. In 2025, they increased tenfold compared to the previous year. In the first half of 2026, growth was slightly lower, but it still amounted to an impressive 272% compared to the same period last year.
The successful IPO of Unitree coincided with a cooling period on the American stock market, where investors began to doubt the payback of the multibillion-dollar costs of building data centers for cloud neural networks, the monetization scheme of which is still unclear. The integration of the program code into tangible mechanisms offers investment funds a different rate of return.
It is difficult for the owner of a logistics center to calculate how much direct profit a subscription to a text neural network will bring him. And replacing two movers with one robot, which does not require breaks, hospital and union guarantees, is easily calculated within the framework of the balance sheet. Capital is migrating to the segment of embodied AI, seeing in it a solution to the problem of developed and now developing countries — demographic decline and shortage of workers.
Close up and get behind
China's seizure of the global robotics market is causing serious concern in the United States and Europe. Complex autonomous vehicles are considered dual-use technologies. Robobacks are already being actively tested by the armed forces of various states as mobile firing points or reconnaissance means on the battlefield.
Civilian applications carry no less risks, since a humanoid or four—legged robot is a walking data collection center equipped with dozens of cameras, microphones and lidars. Located on the territory of a European factory or an American port, such a machine collects huge arrays of telemetric information.
Given the current level of confrontation, American and European regulators are highly likely to apply the same restrictive measures to Chinese robots that were previously imposed on Huawei telecommunications equipment or DJI drones. The industry will inevitably face prohibitive duties, bans on use in the public sector, and strict import quotas. But it is not a fact that in this case, success will accompany the side trying to isolate itself.
The United States has spent huge resources to maintain control over the production of advanced microchips and the development of complex algorithms, which are the brains of the modern economy. They didn't exactly succeed in this. Meanwhile, China is going to secure a total monopoly on the production of "workers" — and it is much closer to this result. With an aging population and the need for total automation, Western countries' abandonment of cheap and functional Asian machines will result in a sharp increase in production costs and a further lag in the real sector. Especially considering that the Chinese are also extremely positive about AI, and Americans and especially Europeans are afraid of it.
Переведено сервисом «Яндекс Переводчик»
