TL;DR
China’s smartphone supply chain is pivoting to humanoid robot manufacturing, with companies like Lingyi iTech, Lens Technology, and AAC Technologies repurposing precision component factories for robotics production. Morgan Stanley doubled its 2026 forecast to 28,000 units as Foxconn prepares robot manufacturing lines in Vietnam and UBTech’s Walker S2 enters mass production with orders exceeding 800 million yuan.
In April, a humanoid robot called the Honor D1 completed the Beijing half-marathon in 48 minutes and 19 seconds under remote control and 50 minutes and 26 seconds in fully autonomous mode, beating every other robot in the 112-team field. The machine that won the race was built using liquid-cooling technology originally developed for Honor’s smartphone division, with structural components supplied by Lens Technology and AAC Technologies, two of the largest precision parts manufacturers in China’s mobile phone supply chain. The company that assembled those components into a bipedal robot capable of running 21 kilometres without overheating is the same company that, until recently, was known primarily for putting screens on phones.
This is the story of China’s humanoid robot industry in 2026. It is not being built from scratch. It is being assembled from the supply chain that already exists.
The pivot
Lingyi iTech, a Shenzhen-based manufacturer that supplies precision components to Apple, Samsung, and Xiaomi, announced in early 2026 that it was going “all in” on embodied intelligence. The company is building a super factory in Beijing targeting production of 10,000 humanoid robots per year by the end of 2026 and 500,000 per year by 2030. Lens Technology, which makes cover glass for iPhones and Samsung Galaxy devices, supplied structural components for the Honor D1. AAC Technologies, the world’s largest manufacturer of miniature acoustic components for smartphones, provided additional precision parts. These are not robotics startups. They are the companies that built the physical infrastructure of the smartphone era, and they are repurposing that infrastructure for the next product category.
The logic is industrial, not speculative. China’s smartphone market shipped roughly 280 million units in 2025, but growth has stalled. The components that go into a phone, precision motors, sensors, thermal management systems, lightweight structural materials, battery cells, are the same components that go into a humanoid robot. The factories that manufacture them at scale already exist. The engineers who designed them are already employed. The pivot is not from phones to robots. It is from one form factor to another, using the same supply chain.
The factory floor
UBTech, one of China’s largest humanoid robot companies, partnered with Foxconn in 2025 to deploy its Walker S1 robots on iPhone assembly lines. The Walker S2, the next-generation model, entered mass production in early 2026 with orders exceeding 800 million yuan. Foxconn itself is now planning humanoid robot manufacturing lines in Vietnam, with production trials scheduled for September 2026 and official production beginning in November. The company that built more iPhones than any other factory on earth is preparing to build the robots that will assemble the next generation of consumer electronics.
Tesla has described its Shanghai Gigafactory as a “golden key” to mass-producing its Optimus humanoid robot, citing China’s dominance in the components humanoid robots require. But Tesla’s ambitions in China are running into a market that is already further along than most Western observers realise. Morgan Stanley doubled its forecast for China’s humanoid robot sales in 2026 to 28,000 units, a 133 per cent year-on-year increase, and expects material costs to fall 16 per cent as supply chain efficiencies from smartphone manufacturing carry over into robotics production.
The numbers
Shenzhen’s robotics industry output hit a record 242.6 billion yuan in 2025, up 20 per cent year on year. The city now accounts for 43 per cent of China’s total service robot output. Production costs for humanoid robots are falling approximately 20 per cent annually, a rate that mirrors the early years of smartphone manufacturing when component costs dropped as volumes scaled and suppliers competed on efficiency rather than novelty.
Morgan Stanley projects the global humanoid robot market will reach $5 trillion by 2050, with 25.4 million robots working globally by 2036. Tesla has raised its 2026 capital expenditure to $25 billion, with a significant portion directed at Optimus production and AI chip fabrication. But Tesla is planning hundreds of units in 2026. Chinese manufacturers are planning tens of thousands. The difference is not ambition. It is supply chain readiness.
The competition
The race to commercialise humanoid robots is global, but the manufacturing advantage is concentrated. Siemens and Humanoid deployed an Nvidia-powered humanoid robot in a live factory trial in Erlangen, Germany, in January 2026, demonstrating that the technology works in real production environments. Meta acquired Assured Robot Intelligence to build what it calls the Android of humanoid robots, an open platform for robot intelligence that other manufacturers can use. Neura Robotics in Germany has attracted attention from investors betting on European alternatives to Chinese and American robotics companies.
But manufacturing at scale is a different problem from manufacturing a prototype. China controls an estimated 90 per cent of the global humanoid robot market not because its robots are more advanced but because its supply chain is more mature. The same factories that drove smartphone component costs down by orders of magnitude over a decade are now applying the same process to actuators, sensors, and precision motors for robots. The same logistics networks that ship billions of phone components across Shenzhen, Dongguan, and the Pearl River Delta are being reconfigured to ship robot parts.
The transition
The half-marathon in Beijing was a publicity event, but the industrial logic behind it is serious. Honor did not build the D1 by starting a robotics division from nothing. It took the thermal management system it had developed to prevent flagship smartphones from overheating during gaming sessions and adapted it to prevent a bipedal robot from overheating during a 21-kilometre run. The engineering challenge is different in scale but not in kind. Both problems require moving heat away from high-performance processors in a compact form factor under sustained load.
This is the pattern that makes China’s humanoid robot industry structurally different from its Western competitors. The United States has better AI models. Europe has more advanced industrial automation in specific applications. But China has the supply chain, and in hardware manufacturing, the supply chain is the moat. Lingyi iTech’s target of 500,000 robots per year by 2030 sounds ambitious until you consider that the company already manufactures hundreds of millions of precision components annually for smartphones. The factory that built your phone is now building the robot that will build the next one. The transition is not a leap. It is a product line extension, executed at the scale that only China’s manufacturing ecosystem can deliver.
