Robotics business surges 510%, RoboSense (2498.HK) accelerates its positioning in physical AI

On July 9, RoboSense (2498.HK) $ROBOSENSE (02498.HK)$ released its sales figures for the first half of 2026: total LiDAR product sales reached 719,200 units, up 170% year-over-year. Of this, robotics segment sales totaled 282,600 units, surging 510% year-over-year; ADAS segment sales reached 436,600 units, up 98% year-over-year.

These figures reflect a clear shift in the company's growth structure: the ADAS business continues to serve as the foundational scale driver, while the robotics business has become a high-elasticity growth contributor. Meanwhile, the company’s self-developed digital chip architecture integrates automotive, robotics, and future physical AI applications into a unified technological and manufacturing framework. Smart vehicles and robots are currently the two types of physical AI terminals that have entered large-scale commercial deployment first. As both core segments scale simultaneously, RoboSense’s growth narrative is expanding from a single automotive cycle to the broader physical AI application market.

Leading across all scenarios in the robotics race, early positioning builds long-term value

Robotics is the most resilient segment amid this structural shift. In the first half of the year, RoboSense shipped 282,600 LiDAR units for robotics applications, a 510% year-over-year increase, making it a key driver of the company's sales growth. Compared to automotive markets, robotic end products are more fragmented and operate in more complex environments, demanding higher performance from sensors in terms of miniaturization, reliability, power consumption, cost efficiency, and integration speed. RoboSense’s ability to achieve scaled deliveries in this market demonstrates that its product capabilities have successfully expanded beyond automotive applications into a broader range of intelligent devices.

The E-platform all-solid-state digital LiDAR is the core product enabling RoboSense’s robotics business to scale. The E1, the platform’s first product, is the world’s first mass-produced all-solid-state LiDAR. It significantly outperforms consumer-grade perception hardware in stability and consistency, and is widely adaptable across diverse scenarios including home appliances, autonomous mobility platforms, and humanoid embodied intelligence systems. As of April 2026, cumulative deliveries of this series have surpassed 300,000 units, setting a new industry record for mass production.

Leveraging standardized hardware and a unified chip architecture, the company has established coverage across multiple specialized segments. In the consumer and commercial equipment space, RoboSense has long supplied leading manufacturers such as Segway and Kuma, and in 2026 launched a strategic partnership with Roborock. In the high-end intelligent equipment segment, the company works closely with top-tier humanoid and quadruped robotics firms including Agility Robotics, Unitree, and Galaxy Robotics. To date, RoboSense has served over 3,400 robotics companies globally, placing it among the industry leaders in ecosystem scale.

According to multiple third-party research firms, RoboSense has ranked first globally in 3D LiDAR shipments for robotics for several consecutive quarters, leading in five key sub-segments: robotic lawn mowers, commercial cleaning and sanitation robots, Chinese humanoid robots, Chinese autonomous delivery vehicles, and Chinese embodied intelligence robots. In a rapidly evolving robotics market, early client acquisition and broad scenario coverage often generate strong compounding effects, which in turn drive continuous improvements in product definition, delivery experience, and economies of scale.

Beyond LiDAR, RoboSense is also advancing multimodal perception solutions such as LiDAR-vision fusion. As robots demand higher capabilities in environmental recognition, close-range obstacle avoidance, and task execution, the integration of LiDAR with vision systems is poised to become a critical extension of the company’s perception product portfolio.

Robust automotive order backlog, with premiumization and globalization strengthening the core business

The automotive ADAS business serves as RoboSense’s 'ballast.' In the premium passenger vehicle segment priced above RMB 500,000, RoboSense holds a commanding 61% market share in front-facing primary LiDAR. As of March 31, 2026, the company had secured定点 (nomination) for 177 vehicle models from 36 automakers, with an order backlog exceeding 9 million units. The value of the automotive business extends beyond current sales and order reserves—it also lies in the automotive-grade reliability, manufacturing consistency, and supply chain management capabilities developed through long-term series production engagements.

As L2+ advanced driver assistance systems continue to penetrate mainstream markets, the role of LiDAR in providing safety redundancy and robust perception in complex scenarios becomes increasingly evident. RoboSense’s digital LiDAR combination—featuring the EM4 main LiDAR and the E1 blind-spot LiDAR—delivers long-range detection, full-scenario blind-spot coverage, and strong interference resistance, making it suitable for complex driving conditions across all scenarios. This solution has already gained recognition from multiple leading autonomous driving companies.

The automotive business provides certainty, while the robotics segment offers growth elasticity; together, these two lines are reshaping how the market perceives RoboSense’s LiDAR growth trajectory. Historically, LiDAR company valuations were anchored primarily to automotive OEM cycles, vehicle SOP timelines, and per-vehicle sensor take rates. With non-automotive contributions rising, RoboSense’s growth drivers are now expanding from a single automotive cycle to encompass both intelligent vehicles and robotics end-markets. This evolving shipment mix is enhancing the company’s long-term growth elasticity.

An in-house developed chip ecosystem forms the core internal driver, and cross-scenario reuse unlocks growth momentum.

Dual external scenario demand represents a shared industry opportunity. RoboSense is achieving high growth simultaneously in both automotive and robotics segments, primarily due to its full-stack, in-house developed digital chip ecosystem.

In April 2026, the company launched its 'Genesis' digital architecture platform along with two in-house developed chips—'Phoenix' and 'Peacock.' The Phoenix chip supports 2,160 lines, has passed AEC-Q100 automotive-grade certification, and enables ultra-long-range detection up to 600 meters alongside precise identification of small objects. A 4-megapixel high-definition LiDAR solution based on this chip has already secured定点 (design win) from a leading automaker. The Peacock chip meets VGA-level 3D imaging standards, featuring an ultra-wide field of view of 180°×135° and near-field detection capability under 5 centimeters, addressing robots’ blind spots and enabling full-domain perception while also serving automotive blind-spot coverage applications. Mass production is expected to ramp up significantly in Q3.

The value of this chip ecosystem lies in its cross-platform reusability. Automotive applications validate high reliability and large-scale manufacturing capabilities, while robotics amplifies the benefits of generalization, miniaturization, and rapid adaptability—both synergizing under a unified digital technology framework. Combined with experience from over one million automotive-grade units produced, the company has established a complete closed loop spanning foundational chips, product platforms, manufacturing systems, and supply chain capabilities. Guosheng Securities notes that this full-stack technological upgrade will reinforce RoboSense’s long-term product value and market positioning in the era of physical AI, while also driving continuous cost structure optimization and gross margin improvement.

From a market perspective, intelligent driving and robotics are jointly expanding the demand frontier for the LiDAR industry. According to CIC data, the global ADAS LiDAR market is projected to reach USD 19.7 billion by 2030, growing at a CAGR of 59.8% from 2025 to 2030. Meanwhile, adoption rates of perception hardware in robotics continue to rise, pushing the total global LiDAR solutions market to USD 41.3 billion by 2030, with a CAGR of 63.7%. In this context, RoboSense—backed by its digital chip foundation and dual-track focus on automotive and robotics—is well-positioned to capture incremental demand driven by the expansion of physical AI terminals.

Conclusion

2026 marks a pivotal year for RoboSense as it iterates its operational structure and upgrades its growth logic. First-half sales data further confirms the shift in its business mix: automotive operations solidify its scale base, while robotics opens a high-elasticity growth runway—all underpinned by its in-house digital chip ecosystem, which enables product and manufacturing capability reuse across a broader range of physical AI terminals.

In the long run, as applications such as intelligent driving, robotaxis, autonomous delivery, commercial cleaning and sanitation, and humanoid and quadruped robots continue to evolve, LiDAR—as the core hardware for 3D spatial perception—still has substantial room for penetration. RoboSense will center its strategy around LiDAR, continuously strengthening its in-house chip development and digital architecture capabilities to solidify its automotive foundation, expand its robotics advantage, and extend its long-term growth boundaries in the physical AI era through diversified innovative businesses. Leveraging its automotive-grade mass production experience, global customer base, and digital architecture capabilities, RoboSense is transitioning from a leader in automotive LiDAR toward a cross-scenario perception platform for the physical AI era.

22K Views