RoboSense: Company Profile
RoboSense posts first quarterly profit on 303K robotics LiDAR shipments, signaling transition from loss-making supplier to vertically integrated sensor platform with sustained profitability path.
- 303,000 Robotics LiDAR units shipped FY2025 Company-stated; +1,142% YoY; independent verification pending
- RMB 104M First-ever quarterly profit, Q4 2025 Company press release, March 2026
- 41.5% Robotics segment gross margin, Q2 2025 On RMB 150M revenue, ~34,400 units
- 185,500+ Robotics LiDAR units shipped Q1 2026 Company-stated; +1,459% YoY
- HQ
- Shenzhen, China
- Founded
- 2014
- Segments
- Infrastructure
- Products
- Digital LiDAR portfolio·In-house chip portfolio·LiDAR perception solutions·Active Camera AC2·VTLA-3D large model·Multi-DOF dexterous hand
- Competitors
- Hesai Technology·Innoviz Technologies·Ouster
RoboSense Hits First Quarterly Profit as Robotics LiDAR Shipments Surge Past 300,000 Units
RoboSense (HKEX: 2498) posted its first-ever quarterly profit of RMB 104M in Q4 2025, a milestone that reframes the company's investment narrative from "loss-making LiDAR supplier" to a vertically integrated sensor platform with a credible path to sustained profitability. The catalyst: a robotics segment that shipped 303,000 LiDAR units in full-year 2025, up 1,142% year-over-year, driven by demand from autonomous lawnmowers, delivery robots, cleaning systems, and early humanoid platforms. Q1 2026 data suggests the trajectory is accelerating, not plateauing.
Business Overview
Founded and headquartered in China, RoboSense listed on the Hong Kong Stock Exchange in January 2024 (ticker: 2498), having previously raised approximately $207.7M in pre-IPO financing. The company operates across two primary revenue streams: automotive ADAS LiDAR supplied to 350+ OEMs and Tier 1s, and a fast-scaling robotics segment now serving 3,400+ customers across logistics, consumer, industrial, and public-service verticals.
Q4 2025's RMB 104M profit signals that robotics volume is beginning to absorb that fixed cost base.
On a last-twelve-months basis through September 2025, the company carried a net loss of approximately RMB 383M against RMB 639M in R&D expenditure — a cost structure that reflects deliberate investment in chip design, manufacturing scale, and AI software rather than operational inefficiency. Q4 2025's RMB 104M profit signals that robotics volume is beginning to absorb that fixed cost base.
Technology Stack
RoboSense's core differentiation is vertical integration across the sensor stack. The company designs four categories of proprietary semiconductors in-house: RISC-V-based data-processing SoCs, digital large-area SPAD-SoCs, 2D addressable VCSELs, and 2D MEMS devices. This chip portfolio reduces bill-of-materials costs, tightens performance specifications, and limits exposure to third-party supply chain disruptions — structural advantages that are difficult for fabless competitors to replicate quickly.
Manufacturing is centralized at the Mars Intelligent Manufacturing Base, a ~48,000 m² multi-factory complex operating at 95% automation. The facility holds automotive-grade certifications including ISO 26262, ISO 21434, IATF 16949, and ASPICE CL2 — qualification barriers that matter in OEM procurement cycles and are increasingly relevant as robotics customers demand supply chain rigor.
Beyond sensors, RoboSense is expanding into adjacent layers: the Active Camera series (AC2, launched 2025), the VTLA-3D AI planning model for long-horizon sequential action execution, and a multi-DOF dexterous hand targeting embodied intelligence applications. Both the VTLA-3D model and dexterous hand remain at prototype stage with no disclosed paying customers or unit volumes. The camera line is in limited deployment. These expansions introduce execution risk and potential channel conflict with integration partners.
| Product | Platform | Deployment Status | Key Metric |
|---|---|---|---|
| Digital LiDAR portfolio | Sensor | FIELDED | 303,000 robotics units FY2025 |
| In-house chip portfolio | Semiconductor | FIELDED | 4 chip categories, 95% automated production |
| LiDAR perception software | Software | FIELDED | Attach rate undisclosed |
| Active Camera AC2 | Sensor | LIMITED | Launched 2025, limited validation |
| VTLA-3D planning model | Software | PROTOTYPE | >10 sequential actions; no revenue disclosed |
| Multi-DOF dexterous hand | Handheld | PROTOTYPE | No volume or customer data |
Market Position
MODERATE CONFIDENCE: RoboSense claims the No. 1 global position in robotics LiDAR by unit volume for 2025, based on company-stated figures without independent third-party verification. The 303,000 robotics units shipped in FY2025 and 185,500 units in Q1 2026 alone (+1,459% YoY) are directionally consistent with a dominant share position in a market where most competitors do not publish comparable granular data.
Named 2025 deployment partners include Coco Robotics (last-mile delivery), Rhino Robotics, Kua Robotics, EasyGo Smart Driving, New Stone Robotics, and LionsBot, alongside undisclosed bulk-production agreements with European and North American warehouse automation operators. The 3,400+ robotics customer count reduces single-customer concentration risk, though order volume distribution across that base is not disclosed.
The robotics segment's 41.5% gross margin in Q2 2025 — on RMB 150M revenue and ~34,400 units — compares favorably to the consolidated 27.7% gross margin, confirming that robotics is the higher-margin growth vector. Automotive ADAS, while providing volume and certification credibility, faces more intense price pressure.
Outlook
The near-term catalyst set is well-defined. The FY2025 audited annual report, expected H1 2026, will either confirm or qualify the 303,000-unit robotics figure and provide the first disclosed look at software revenue contribution — currently the largest unquantified variable in the bull case. A sustained breakeven trajectory depends on robotics mix continuing to shift consolidated gross margins toward the 30–35% range.
Geopolitical and regulatory exposure is a material constraint on international scaling. As a China-headquartered sensor and AI company, RoboSense faces potential data-handling scrutiny and local content requirements in EU and North American markets — risks that are structural rather than operational and unlikely to be resolved by product improvements alone.
The strategic expansion into cameras, AI planning models, and dexterous hands is coherent in vision but unproven in execution. If VTLA-3D and the dexterous hand remain cost centers through 2027, the RMB 639M annual R&D run rate will be difficult to justify on sensor hardware margins alone. Conversely, a named high-volume humanoid platform win would substantially de-risk the thesis.