The robots on this company's shelves do not look like the ones in the headlines. There are no towering humanoids (人形机器人) walking factory floors — just small devices that turn a dumb curtain or door lock into something that listens. Yet when the company rang the bell in Hong Kong, it was pitched as a bet on the same idea driving the robotics boom: intelligence moving into physical things.
A listing built on small machines
Woan Robot (卧安机器人), the Shenzhen company behind the SwitchBot brand, began trading on the Main Board of the Hong Kong Stock Exchange on 30 December 2025 under the ticker 06600.HK. The deal:
- Final offer price of HK$73.80 per share, within a marketed range of HK$63–81.
- 22,222,300 H-shares sold globally.
- Gross proceeds of about HK$1.64 billion (≈ US$211 million), with net proceeds around HK$1.54 billion.
- Hong Kong public offering subscribed roughly 254.5 times; the international tranche about 8.9 times.
The stock earned the label "first AI embodied home-robot stock" (AI具身家庭机器人第一股) — a descriptor the company and local press used to frame what makes a curtain-robot maker a capital-markets story.
What Woan actually sells
Woan is a provider of AI embodied home-robot systems (具身智能家用机器人系统). Rather than asking households to replace whole appliances, it sells retro-fit devices that add intelligence to what is already there:
- Enhanced execution robots — a "finger" robot that presses buttons, plus door-lock and curtain robots.
- Mobile and motion robots for cleaning and carrying.
- Perception and decision systems — smart hubs, sensors and cameras that coordinate the others.
The logic is pragmatic. A finger robot that flicks a light switch is cheaper and less disruptive than a new smart lamp. That positioning helped Woan build a broad matrix: seven categories and 47 standard product units, with the SwitchBot app at more than 3.5 million registered users and over 10.7 million connected devices by 2025. The origin story is a 35-gram "finger" device, the SwitchBot Bot, which raised about US$70,000 in a 2017 crowdfunding campaign and proved there was demand for retro-fitting rather than replacing.
Crucially, these are shipped, paid-for products — not conference demos. That distinction separates Woan from the humanoid developers that dominate robotics headlines but sell almost nothing to consumers today.
The market position it claims
Independent research cited in the prospectus (Frost & Sullivan) places Woan as the largest global AI embodied home-robot system provider by 2024 retail sales, with an 11.9% share, and the only company with full-category coverage in the segment. Those are single-sourced advisor figures, not regulator data, but they set the narrative the IPO rode.
The financials show a company that crossed a threshold in 2025. Revenue climbed from RMB 274.6 million (≈ US$38 million) in 2022 to RMB 609.9 million (≈ US$84 million) in 2024, and hit RMB 396 million (≈ US$55 million) in the first half of 2025, up 44.1% year on year. After three years of losses, it turned narrowly profitable in 1H 2025 with about RMB 28 million (≈ US$3.9 million) in net income.
Why a home-robot maker belongs in an AI column
The "embodied AI" (具身智能) story is usually told through factory arms and warehouse humanoids. Woan tells it through the home — a slower, duller, but far larger market. An aging Japan, where Woan earns two-thirds of its revenue, is a natural early adopter of assistive home robotics, and the company's retro-fit philosophy suits households unwilling to rip out working appliances.
That is the bet underwriters sold: not a single hero robot, but a system of small, intelligent actors that compound into a smart home. If the H1 humanoid lands, Woan would stretch from "add intelligence to what you own" to "do the chores you hate" — a much bigger addressable problem, and a much harder engineering one.
Who backed it, and what the money buys
Cornerstone investors committed over HK$700 million (≈ US$90 million), with Hillhouse (高瓴) the largest at about US$30 million, alongside Cithara, Infini Capital, China Orient, Wind Sabre, Yield Royal and Sage. The roster of early backers also includes "DJI's godfather" Li Zexiang (李泽湘), a noted robotics investor who sits as a non-executive director.
Per the allotment results, about 66.5% of net proceeds goes to R&D — enhancing robot localization, AI machine-vision control and edge computing, plus work on a VLA-model-based embodied housework robot. The rest funds commercial expansion and potential M&A.
The humanoid question
Woan has signaled it will not stay small. It plans a first humanoid (人形机器人) home robot, the H1, aimed at high-frequency chores like sorting laundry and clearing dishes — a move that would put it in the same conversation as the headline-grabbing humanoid developers, but aimed at the home rather than the warehouse.
Honest limitations
- Raise and subscription figures come from the company's allotment results and exchange-adjacent advisories; they are pre-greenshoe unless stated.
- Market-share ranks are from a single commissioned advisor (Frost & Sullivan) cited in the prospectus, not an independent regulator.
- Revenue is heavily concentrated: Japan supplied 67.7% of sales in 1H 2025, with Europe 17.2% and North America 11.7% — a channel and geography risk.
- The H1 humanoid is a planned product, not a shipped one; timelines for home humanoids are historically slippery.
What readers can do now
- Separate "embodied AI" hype from revenue: Woan ships paid products today, unlike most humanoid pure-plays — judge peers on sales, not demos.
- Watch the Japan concentration; a single-market dependency is a real risk even for a profitable quarter.
- If you track robotics IPOs, compare proceeds allocation — a high R&D share (here ~66.5%) signals a platform bet, not a dividend story.
