Morgan Stanley: Automotive lidar under pressure as embodied intelligence opens data-collection opportunities

How much pressure is the automotive business under?

Morgan Stanley expects Hesai’s and RoboSense’s core results for the third quarter of 2026 to meet lowered market expectations, but sees continued pressure on automotive sales volumes and margins in the third and fourth quarters. Weak automotive demand, price cuts and rising material costs have weighed on auto-parts companies’ results.
Competition in lidar for advanced driver assistance is intense, requiring suppliers to balance volume against price. Automakers’ dual-supplier strategies also weaken pricing power; even when shipments rise, profits may not grow in tandem.
The bank expects Hesai’s third-quarter lidar shipment growth to slow both year over year and quarter over quarter, but sees gross margin remaining at around 40%, enough to cover operating expenses and deliver an operating profit.

RoboSense has reported third-quarter lidar shipments of 596 thousand units, up 221% year over year and 53% quarter over quarter. However, Morgan Stanley expects price cuts and provision costs related to a Geely Auto recall to reduce its gross margin sequentially in the quarter.
For Hesai, sales at a major electric-vehicle customer have fallen short of expectations, and its share of purchases by traditional automaker customers may decline. The bank has therefore cut its earnings forecasts for 2026 through 2028 by 40%, 34% and 26%, respectively.
Because margins on lidar for advanced driver assistance are below expectations, the bank has widened its net loss forecasts for RoboSense for 2026 and 2027 by 11% and 2%, respectively, and cut its 2028 earnings forecast by 10%. It still expects the company to turn profitable in 2028, citing an improved customer mix, cost savings from in-house chips and expansion of its robotics business.

Morgan Stanley believes embodied intelligence needs large amounts of real-world experience data to perform tasks reliably. Lidar can be used alone or with cameras to collect environmental information and provide data for training.
Hesai’s opportunity extends from selling lidar to robot makers to spatial-data platforms and training-data collection: Kosmo One combines lidar, imagery and inertial sensors to record environments, while Kosmo World converts the scans into three-dimensional scenes usable for simulation.
RoboSense is expanding into vision-based tactile and MEMS tactile sensors. These sensors can combine tactile information with visual and robot-motion data to help train grasping and fine manipulation, extending data coverage from identifying an object’s location to how to manipulate it.

Morgan Stanley believes price reductions driven by technological advances could broaden lidar adoption, while robotics could become a second growth engine with higher margins. Hesai’s scale and manufacturing capabilities and RoboSense’s cost savings from in-house chips underpin the bank’s positive view of their respective profitability.
But investment in new products also carries costs: Hesai’s loss-making growth initiatives are diluting profits from its core lidar business, while RoboSense faces greater cash-burn pressure. Camera-dominant alternatives and restrictions involving China and the United States also create uncertainty for long-term adoption and overseas customer expansion.
The bank maintains its overweight ratings on both companies, while cutting its price target for Hesai from 27 US dollars to 22 US dollars and its target for RoboSense from 22 Hong Kong dollars to 20 Hong Kong dollars.