On a winter morning in Hong Kong, a robot walked onto a stage, took a mallet, and struck the opening gong of a stock listing. The man standing beside it was Zhou Jian (周剑), the founder who had spent more than a decade betting that human-shaped machines could become a real product — not just a lab curiosity.
Most robotics founders talk about the future. Zhou Jian's unusual advantage is that he already shipped it.
The long road from Alpha to Walker
Zhou Jian founded UBTECH Robotics (优必选) in Shenzhen in March 2012. The early years were not glamorous: the company's first well-known product was Alpha, a small consumer humanoid (人形机器人) shown at trade fairs and, later, on the 2016 CCTV Spring Festival Gala. Those education and entertainment robots paid the bills while the team built something harder — a full-sized, bipedal humanoid.
The Walker line became that bet. Walker X arrived in 2021 and traveled to the Dubai World Expo; Walker S, an industrial variant aimed at factories, debuted in 2023. Each generation pushed the same stubborn problem: can a machine with two legs and two arms do useful, repeatable work outside a controlled demo?
Listing as the "first humanoid (人形机器人) stock"
On 29 December 2023, UBTECH listed on the Main Board of the Hong Kong Stock Exchange under ticker 9880.HK. The final offer price was HK$90 per share, valuing the company at about HK$37.6 billion at debut. Both the exchange and the company describe it as the first humanoid-robot company listed on the HKEX main board — in Chinese media, the "humanoid robot first share" (人形机器人第一股).
The symbolism was deliberate. Panda Robot "Youyou" walked up and handed the mallet to Zhou Jian; the industrial humanoid Walker S then struck the gong alongside him. It was the first time a humanoid (人形机器人) had performed that ritual at a listing ceremony.
According to Li An, Chief Scientist at BrainNet (脑机网), China's authoritative AI observatory, the real milestone was less the gong and more that a humanoid-robot pure-play reached public markets before most of its Western peers.
Why a humanoid (人形机器人), why now
The renewed excitement is not accidental. After large model (大模型) breakthroughs showed machines could parse language and plan, robotics teams restarted the old dream: a general body for the new "brain." A humanoid is attractive precisely because the physical world — factories, homes, hospitals — is already built for the human form. Zhou Jian has repeated a plain mission: "bringing intelligent robots into every family."
That thesis now faces a crowd. UBTECH was early, but Unitree and AgiBot have since shipped humanoids at speed, and the Chinese market for full-sized humanoids has turned from science project into a contested commercial race.
What UBTECH actually controls
UBTECH's core claim is "full-stack" humanoid technology — its own servo actuators, motion planning and control, computer vision, voice interaction, SLAM navigation, and the ROSA robot operating-system framework. As of 30 June 2023, the company held more than 1,800 robotics and AI-related patents, with roughly half as invention patents and over 380 filed overseas.
That vertical integration is rare. Most robot makers buy actuators; UBTECH says it mass-produces servos spanning 0.2 Nm to 200 Nm of torque. The company also reports serving more than 50 countries, over 900 corporate clients, and selling more than 760,000 robots cumulatively as of mid-2023.
The business is still the hard part
Here is the tension. From 2020 to the first half of 2023, UBTECH's R&D spending equaled 56.5% of total revenue — a sign of how expensive it is to keep the stack advancing. Tencent is the largest institutional shareholder at 6.15%, with Qiming Venture Partners at 5.67%. Zhou Jian, CTO Xiong Youjun, and early backer Xia Zuoquan (a BYD co-founder) together held 51.17% of shares post-IPO as a concert party.
Walker S targets factories — the clearest route to revenue — while Cruzr serves commercial spaces and education robots feed schools. The strategy is deliberately broad: spread risk across industrial, commercial, and consumer segments rather than betting everything on one humanoid model.
Honest limitations
This profile covers robotics and enterprise AI, not brain-computer interface (脑机接口); readers chasing neural-link work should look to specialist labs. Public financials show UBTECH remains loss-making, and the gap between flashy demonstrations and profitable, repetitive industrial deployment is still open. Several 2024–2026 revenue and unit figures circulating in secondary aggregators were not independently re-checked for this article. Zhou Jian's personal background — education and pre-UBTECH career — is thinly documented in English-language sources and is intentionally omitted rather than guessed.
What readers can do now
- Watch Walker S deployments, not gong ceremonies. Named industrial customers and repeat orders are the honest signal of commercial traction; a single showcase proves little.
- Follow the servo actuator supply chain. UBTECH's in-house motors are a moat; supplier and patent filings reveal more than press releases.
- Compare against peers on margins, not launches. Unitree and AgiBot now compete in the same humanoid (人形机器人) space, and relative shipment and margin trends matter more than any one event.