A lidar company used to sell a sensor to a car maker and wait for the car to launch. The wait could take years. These days it can ship the same sensing kit to a warehouse robot, a delivery fleet or a machine being trained to fold laundry — and see the result in the same quarter.
That change is written into the numbers Hesai Group filed on 10 November 2026. The company still makes most of its money from lasers and optics. But the fastest-moving lines in the report are the ones that have nothing to do with passenger cars.
What the filing says
Hesai (Nasdaq: HSAI; HKEX: 2525) reported third-quarter net revenues of RMB 1.0683 billion (US$159.3 million as stated by the company, ≈ HK$1.24B), up about 34% year on year. Net income was RMB 80.8 million (US$12.1 million, ≈ HK$94M). That was the sixth consecutive quarter of GAAP profitability — a streak that matters in a sector where most lidar makers have never had one.
Lidar shipments reached 729,673 units in the quarter, up 65% from a year earlier. Of those, 638,100 were ADAS lidar for driver-assistance systems and 91,573 were robot lidar. The robot side grew more slowly in percentage terms but is smaller in base; cumulatively, robot lidar shipments have more than doubled year on year over the first nine months.
The more striking figure is in the components. Hesai shipped 23,405 robotic actuation modules in the quarter, up 97% from the previous three months. Cumulative shipments stood at 35,295 by the end of September. These are the joints and motors that make a humanoid (人形机器人) move, and the company says they typically account for more than half of such a machine's bill of materials. Full-body modules have entered mass production alongside robotic hands.
The core lidar business remained the cash engine: it reported operating income of RMB 145 million (≈ US$20M / HK$155M) in the quarter, up 88% year on year. Overall gross margin was 41%, down slightly from 42.1% a year earlier, which the company attributes to a higher share of lower-margin products.
The part that is not cars
Hesai groups its newer activities under Strategic Growth Initiatives (SGI, 战略增长业务): the actuation modules, a 3D data engine called Kosmo, and associated software. SGI brought in RMB 90.51 million (≈ US$13M / HK$97M) in the quarter, more than double the previous quarter, though it still recorded an operating loss of roughly RMB 82.9 million (≈ US$12M / HK$89M) as the company invests ahead of revenue.
The nine-month SGI total has already passed the company's original full-year target of RMB 100 million (≈ US$14M / HK$107M). Management reaffirmed guidance of RMB 200 million to RMB 300 million (≈ US$28M–42M / HK$214M–321M) for 2026. Its chief financial officer, Andrew Fan, said on the earnings call that robot-related products — robot lidar, Kosmo and actuation modules taken together — contributed more than half of third-quarter revenue. That is a broader bucket than the SGI segment, since robot lidar revenue sits inside the lidar division.
Kosmo was launched globally in October. It turns real spaces into editable, AI-ready 3D assets; a companion 8K spatial camera and a cloud platform were released with it. Hesai says early customers include Pixar's RenderMan, and that deployments span robotics training, film, cultural heritage and spatial navigation. A partnership with JD.com announced in September links Hesai's full-colour lidar to the retailer's unmanned logistics vehicles and robot fleet plans.
Why a laser company is building robot joints
On paper this looks like diversification. In practice it is closer to a supplier following its skill set into a new market.
A lidar maker already knows how to design optics, run a high-precision production line and cut unit costs at volume. Those are the same capabilities needed to manufacture actuators and spatial sensors for robots. Chief executive Yifan "David" Li has framed the shift as moving from a component vendor to a full-stack infrastructure platform for robotics and what the company calls physical AI.
There is also a defensive logic. Hesai's automotive business is exposed to car production plans, and several Chinese automakers lowered their 2026 sales targets this year. The company says it won 43% of China's long-range ADAS lidar market in September, its twentieth consecutive month at the top, according to research firm Gasgoo, and a separate Yole Group report put its share of the 2025 global robotaxi lidar market at 62%. Strong as those positions are, neither is a growth story if the underlying vehicle market softens. Robot customers, by contrast, are buying early and in small batches — but they are buying.
One caution on the headline growth: the comparable 2025 quarter included about RMB 185 million (≈ US$26M / HK$198M) of one-off gains from selling an equity stake and transferring intellectual property. Strip that out and underlying GAAP net income was up roughly 14% year on year. The company's own figures are the right ones to quote; the adjusted number is the right one to reason from.
Key takeaways
- Hesai's third-quarter revenue was RMB 1.0683 billion (US$159.3 million / ≈ HK$1.24B), up about 34% year on year, with a sixth straight profitable quarter.
- Robot-actuation-module shipments jumped 97% quarter on quarter to 23,405 units; cumulative shipments reached 35,295.
- Robot-related products together contributed more than half of quarterly revenue, according to the company's chief financial officer — but the SGI segment itself is still loss-making.
- The automotive lidar business remains profitable and dominant in China, yet management is repositioning the company as a robotics and physical-AI infrastructure supplier.
Honest limitations
All figures here come from Hesai's unaudited quarterly release, its HKEX filing and reported earnings-call comments; quarterly numbers are unaudited and can be restated. The "more than half of revenue" statement is the CFO's, using a definition of "robot-related" that includes lidar revenue booked in a different segment; there is no published reconciliation. Market-share figures are third-party estimates from Gasgoo and Yole Group, not independent audits. Currency conversions are approximate, and the US dollar amounts are those used by the company rather than a single market rate. Forward-looking statements about 2027 shipments and SGI guidance are company targets, not achieved results.
What readers can do now
If you follow listed Chinese hardware companies, read the segment note before the headline: a profitable core business funding a loss-making new one is a different risk profile from a company whose entire revenue is growing. If you buy or specify robot components, watch the actuator shipment number rather than the demo — it is the closest publicly reported proxy for whether factory orders are real. And when a company says its addressable market has changed, check whether the old customers are still growing; for Hesai, that answer currently sits in the ADAS column.
