The most revealing thing about UBTech's new robot order is where the robots are going: car dealerships, not living rooms. On Nov. 11, the Chinese humanoid robot company said Yongda Auto, a major luxury car dealer, will buy 400 U1 Pro robots, with delivery planned within the year. It is the first bulk commercial order for the U1 series since its June launch. Earlier, in July, UBTech placed U1 robots in stores run by Bosi Group, a Mercedes-Benz partner in China, for display and simple customer guidance. On Nov. 9, it signed a three-year sales cooperation with China Telecom's Tianyi Terminal for 100,000 units.
The pattern is consistent: UBTech is using existing retail and telecom channels to reach consumers, not selling a robot that does housework. The U1 Pro has bionic silicone skin, 88 degrees of freedom, and more than 30 micro-expressions. It does not walk on its own, cannot do chores, and runs for only two to four hours on a charge for conversation. The price is 169,800 yuan. It is an emotional companion and a display object, not a domestic worker.
The founder's timeline
Founder and CEO Zhou Jian has given a timetable that is more useful than the order headline. He has said that family emotional companionship is three to five years away, while household chores are seven to ten years away. At the 2026 APEC meetings, he said general-purpose humanoid robots with strong generalization still need time, given current hardware and large-model capabilities. That is a notable admission from the head of a company selling a companion robot now.
The gap between promise and delivery is already visible. At the June launch, UBTech said it had 13,361 orders across channels, and Zhou promised 10,000 deliveries this year. By the August earnings call, CFO Zhang Ju had cut the expected delivery to 1,500 to 2,000 units, about 80% lower. The company blamed capacity. The U1's head packs 19 degrees of freedom and 2,000 to 3,000 parts into a small space, with more than 30 micro-expressions. The silicone skin has a low yield rate, eyebrows and eyelashes are inserted by hand, and the micro servo motors are non-standard. The supply chain is not mature. A Guangxi factory was delayed by two to three months because of natural factors, and UBTech scrambled to add capacity in Shenzhen.
That is a manufacturing story as much as an AI story. A robot can be impressive in a showroom and still be hard to build at volume. The company's own numbers show the difference between orders and delivered units.
The business still runs on industry
In the first half of 2026, UBTech's revenue was 1.269 billion yuan. Full-size embodied humanoid robots brought in 590 million yuan, but about 569 million yuan of that came from custom robots and solutions for other industries. Customers include Airbus, BYD, Geely, and Foxconn. The Walker series has entered car factories for transport, inspection, and assembly. Those projects are real, but they are long to recognize, hard to replicate, and slow to sign. If UBTech remains mainly a project integrator, it will struggle to earn a high valuation multiple.
The market has noticed. UBTech's market value was about HK$33.1 billion on Nov. 19. Its shares rose on the order announcement, then gave back the gain and continued lower. Compared with Unitree, a private robot company with similar first-half revenue, UBTech's market value is only about 14% of the other company's. Investors are not treating the dealership order as proof of a home robot market.
UBTech's AI work is serious on paper. It has open-sourced Thinker, a 100-billion-parameter embodied AI foundation model, and claims first place on nine international benchmarks. It also has Thinker-VLA for millimeter-level loading, generalized grasping, and two-arm coordinated box handling, plus Thinker-WM for industrial scenes. These are aimed mainly at industrial customers. Home generalization remains unsolved.
The honest read is this: three to five years is plausible for a narrow, expensive companion robot in some homes. It is not a timeline for a useful domestic helper. The U1 orders validate a channel strategy, not a home-robot business. Watch deliveries, returns, and actual daily use. Those will tell us more than another batch order.