2026-11-15 22:06

Introduction: Associated client Sharpa drives first mass production, with cost-plus markup rising to 80%. On July 31 this year, Horus Technology amended a robotic component supply agreement, increasing the annual transaction cap for 2026 from RMB 100 million to RMB 300 million.

On July 31 this year, Horus Technology revised a robotic component supply agreement, raising the 2026 annual transaction cap from RMB 100 million to RMB 300 million.

The buyer is Sharpa, a company developing humanoid robots and dexterous hands. Under the new agreement, Horus can supply Sharpa up to RMB 221 million worth of robotic actuators this year, plus RMB 53 million in manufacturing and support services.

The parties also renegotiated pricing terms. The cost-plus margin for robotic actuators, originally set between 40% and 50%, has now been increased to 70%–80%. The markup for manufacturing and support services has also been raised to the same range.

Horus explained that as product development deepens, the company has begun independently developing more core components such as motors and reducers, significantly increasing its R&D burden. In Q2 alone, related R&D expenses rose by approximately 29.9% compared to Q1.

This company, which recently achieved scale profitability in the automotive LiDAR market, is now allocating resources toward robotic joint development.

Sharpa holds a special status.

In its announcement, Horus disclosed that three co-founders—Li Yifan, Sun Kai, and Xiang Shaoqing—hold majority voting rights in Sharpa through indirect shareholding. Per HKEX rules, the procurement relationship constitutes a continuous related-party transaction.

In the first half of this year, Horus’s newly established strategic growth business generated RMB 44.94 million in revenue, of which RMB 44.90 million came from sales of actuators and provision of manufacturing support services to Sharpa. In Q3, strategic growth business revenue reached RMB 90.51 million, while group-related party revenue amounted to RMB 90.32 million—again showing near-identical figures.

New business is growing rapidly, with revenue highly concentrated among a single customer linked to the founding team.

Meanwhile, profitability in Horus’s traditional LiDAR business is improving. In Q3, the business shipped 729,700 units, up 65.3% year-on-year, generating an operating profit of RMB 145 million, up 88% YoY. Meanwhile, losses from new ventures—including robotic actuators and spatial data—amounted to RMB 82.89 million, offsetting approximately 57% of the mature business’s operating profit.

After years of R&D, price reductions, and mass production, automotive LiDAR has finally become a stable source of operating profit. Subsequently, Horus entered another market requiring renewed investment in R&D and manufacturing capacity.

Suteng Juchuang is pursuing a similar strategy. In August this year, Suteng Juchuang announced completion of its first robotic joint module development, securing initial procurement demand for tens of thousands of units, with plans to begin mass delivery in Q4.

Previously, both companies sold LiDARs to automakers; now they are manufacturing robotic joints, tactile sensors, and motion actuation modules.

Professional suppliers have long existed in the robotic joint market. Zhaowei Electronics specializes in micro transmission and drive systems, while Green Harmonic operates harmonic reducers. Both companies are also developing integrated robotic joints.

Horus and Suteng Juchuang, bringing accumulated technological, financial, and manufacturing expertise from the automotive industry, are entering markets already familiar to these established players.

A robotics firm brought Horus its first actuator order

The relationship between Horus and Sharpa dates back to 2025.

In April of that year, Horus invested approximately RMB 100 million via a partnership entity to indirectly acquire about 10% equity in Sharpa. Four months later, the company transferred these shares to two independent third-party investors for a consideration of approximately RMB 276 million, recognizing a gain of around RMB 148 million after deducting relevant taxes and fees.

Beyond equity transactions, the two parties also engaged in technical collaboration.

In May 2025, Horus signed an intellectual property licensing and transfer agreement with Sharpa, transferring certain internally developed technical know-how for RMB 37 million, with the transaction completed in August.

In March this year, cooperation expanded into component manufacturing. Horus began supplying LiDARs, robotic actuators, and providing manufacturing, assembly, and testing support to Sharpa under a framework agreement.

One key product is Sharpa’s W02 dexterous hand, launched this year.

The W02 features 21 active degrees of freedom and weighs less than 750 grams. Its five fingers can perform grasping, rotation, and fine manipulation, with tactile sensors at the fingertips to detect contact location and force changes.

A mechanical hand approaching human size must accommodate dozens of motion units, imposing high demands on motor size, reduction mechanisms, weight, and thermal management. Horus’s actuator modules must deliver power output and motion control within extremely limited space.

Sharpa’s other product, the D01 full-body robot, includes 66 degrees of freedom, including the dexterous hand, and similarly employs Horus actuator modules.

Sharpa provides specific product and technical requirements, while Horus develops actuators tailored to finger joints, arms, and other motion mechanisms, delivering products for integration and validation by Sharpa.

In Q3 this year, Horus delivered 23,400 actuator modules, up ~97% quarter-on-quarter. Cumulative deliveries for the first nine months reached 35,300 units, with monthly production capacity increasing to 10,000 units.

These deliveries correspond to actuator modules—each humanoid robot typically requires dozens of motion units. Horus’s actuator business remains in early-stage mass production.

Sharpa plays a crucial role at this stage. Robot manufacturers require suppliers to collaborate closely during product development, repeatedly adjusting joint dimensions, torque, control interfaces, and mechanical structures. Horus gradually assumes greater responsibilities in component design and manufacturing.

In July, as Horus increased internal development of motors, reducers, and other components within the actuator, it raised the cost-plus margin in the agreement. The announcement disclosed that this adjustment was based on comparable company transfer pricing analysis conducted by an independent accounting firm.

The cost-plus margin is calculated on the agreed cost base and differs from product gross margin or group operating profit. As Horus assumes more R&D and manufacturing work, the pricing adjustment compensates for the added technological and production investments.

This collaborative model enables Horus to secure development scenarios, first-order volume, and manufacturing experience—while tightly aligning its new business with Sharpa’s product timeline.

The W02 dexterous hand uses a specific motion mechanism, joint dimensions, and control scheme. Another robotics company might adopt a different number of degrees of freedom, alternative drive methods, or even entirely different motors and reducers.

The modules Horus developed for Sharpa contain extensive engineering work tailored specifically to this robot. How much of the original design can be reused when supplying other brands depends on the customer’s mechanical structure and control system.

LiDAR firms enter the precision transmission market

Horus’s capability in robotic actuators partly stems from its prior LiDAR business.

Automotive LiDAR involves chip design, electronic control, precision alignment, structural machining, and large-scale consistency production. Some products also incorporate motors, rotating or scanning mechanisms.

Over recent years, Horus advanced LiDAR mass production through dedicated chips, optical design, and process improvements. In Q3 2026, the company’s LiDAR business generated revenue of RMB 978 million, up 22.9% YoY, while operating profit surged 88%.

Calculating revenue per unit (revenue divided by shipment volume), each LiDAR corresponds to approximately RMB 1,340—down from about RMB 1,802 a year ago. This metric includes both automotive and robotic LiDARs, reflecting overall product portfolio shifts.

With expanding output and declining manufacturing costs, Horus has accumulated mature industrial product development and mass production capabilities in the LiDAR business.

However, robotic joints require deeper expertise in mechanical transmission.

An actuator must transmit motor power to the joint while maintaining precision across varying loads and motion speeds. Gear backlash, transmission efficiency, and mechanical lifespan all impact product performance.

Horus now faces well-established professional component firms with long-standing operations in these areas.

Zhaowei Electronics has long specialized in micro transmission and drive systems. In H1 2026, the company had already developed multiple dexterous hand products for robotics, offering both high-degree-of-freedom solutions with individual drives and low-cost alternatives using linkages and under-driven mechanisms to reduce motor count.

Full-drive dexterous hands feature independent drive units for each finger joint, suitable for complex tasks. Under-driven designs use mechanical linkages to coordinate multiple joint movements, reducing the number of motors and controllers required.

Zhaowei Electronics can select among micro motors, reducer mechanisms, and transmission structures based on different product requirements.

Green Harmonic, meanwhile, has long focused on harmonic reducers.

Its technological expertise centers on flexible gear materials, tooth profile design, ultra-precision machining, and transmission reliability. In H1 2026, Green Harmonic continued R&D on integrated electromechanical modules combining motors and reducers, including compact solutions using frameless torque motors, planetary reducers, and encoders.

These two companies have evolved from selling standalone motors and reducers to offering complete joint modules.

Horus entering the actuator market now competes directly with them for the same customers and procurement budgets.

Horus can leverage existing electronic control, structural design, and manufacturing management capabilities, but achieving high-precision reducers, transmission mechanisms, and long-term motion reliability requires dedicated R&D and process mastery.

This year, Horus continues increasing R&D investment in actuator self-development, beginning to build more core components like motors and reducers.

Suteng Juchuang’s approach differs slightly.

In its H1 2026 report, the company disclosed that its first robotic joint module leveraged miniaturized motor technology originally developed for LiDARs.

The existing team already possessed experience in small-scale motor design and control. The new joint module further improved torque density, control accuracy, and lifespan. This year, Suteng Juchuang secured initial procurement orders for tens of thousands of robotic joint modules.

Suteng Juchuang also extended its existing optical and semiconductor technologies into two new types of robotic products.

One is a spatial camera, using proprietary SPAD chips to capture color images and depth information for robot navigation, environmental perception, and data acquisition.

The other is a tactile sensor. By adapting its MEMS technology, Suteng Juchuang measures mechanical forces when a robot’s hand contacts external objects.

Spatial cameras and LiDARs both revolve around optical sensing and depth information; tactile sensors introduce new requirements for mechanical measurement; robotic joints further enter the domain of mechanical transmission and motion control.

Suteng Juchuang aims to expand its former model of selling single LiDARs to robot customers into a suite of key components.

Another shift in this competition: component suppliers are striving to sell complete modules.

In the past, robot manufacturers could separately purchase motors, reducers, encoders, and controllers, then assemble and integrate them themselves. Now, Zhaowei Electronics, Green Harmonic, Horus, and Suteng Juchuang are all developing higher-integration actuator products.

Complete modules reduce some R&D and assembly work for OEMs, while suppliers simultaneously manage performance and cost across multiple stages.

Automotive LiDAR firms bring electronic technology and scale manufacturing experience, while traditional precision transmission firms hold deeper expertise in motors, reducers, and mechanical reliability.

They are now competing for the same product: integrated motion actuation modules that can be directly installed on robotic joints.

Who spends first? Who profits?

Horus and Suteng Juchuang entered the robotics industry with different starting points.

Horus achieved early development and mass delivery through Sharpa. Their collaboration is close-knit, involving substantial customization, and features a cost-plus pricing arrangement.

Suteng Juchuang, by contrast, has a broad base of robotic customers.

In H1 2026, Suteng Juchuang sold 282,600 units of robotic and other application LiDARs, up 510.4% YoY. The company disclosed that its products now serve over 3,400 global robot clients across lawn mowing, logistics handling, commercial cleaning, and embodied intelligence markets.

These customers provide ready-made sales channels. Suteng Juchuang can introduce spatial cameras, tactile sensors, and joint modules to existing clients, though actuators are typically selected by mechanical and motion control teams, with procurement standards differing from those for LiDARs.

A lawn mowing robot manufacturer may need low-cost environmental perception devices but rarely high-degree-of-freedom joints. A humanoid robot company that has already purchased LiDARs may still independently choose motors and reducers through its mechanical R&D team.

Horus and Sharpa’s collaboration takes a deeper product development approach.

Horus handles actuator design, manufacturing, and support services, while Sharpa provides specific robot products, technical requirements, and purchase orders. Through cost-plus pricing, the arrangement incorporates additional R&D and manufacturing efforts into the contract price.

This type of relationship differs significantly from selling standardized components in open markets.

Deep customization allows suppliers to participate in customer product design, securing longer-term engineering collaboration opportunities and generating revenue through development services and manufacturing contracts.

Standardized modules are better suited for repeated sales across diverse customers. Suppliers must bear R&D costs upfront and rely on customer volume and order scale to amortize costs.

Both models can be profitable—the difference lies in who bears the R&D cost, whether the product can be reused, and how much customers are willing to pay.

The Horus-Sharpa agreement covers product supply, manufacturing support, and R&D activities. As self-developed components increase, the cost-plus margin rose from 40%–50% to 70%–80%.

This allows Horus to reflect more engineering and manufacturing investment in initial orders through contract pricing.

For independent robot clients, procurement terms may differ. Customers can solicit quotes from Zhaowei Electronics, Green Harmonic, and other actuator suppliers, or choose to source components separately based on their own technical capabilities—or opt for integrated modules.

The larger the procurement volume, the more likely customers will request reduced customization, standardized module specifications, and lower unit prices.

Robot joint market demand remains fragmented.

Take Yuqi Tech’s G1 as an example: depending on configuration, the robot uses between 23 and 43 joint motors. Other humanoid robots employ varying numbers and structures of motion units in legs, waist, arms, and dexterous hands.

Within the same robot, joint torque, size, speed, and load vary significantly. High-load leg joints differ substantially from miniature actuators in dexterous hands, requiring distinct designs.

Different OEMs’ mechanical structures and control interfaces further diversify product offerings.

If a supplier can develop a set of universal motors, reducers, drive circuits, and control interfaces, it could leverage a limited core component set to cover more robot models.

Zhaowei Electronics has already met diverse needs through multiple dexterous hand solutions. Green Harmonic enters joint systems via reducers and electromechanical integrated modules. Horus and Suteng Juchuang are also striving for similar platform-level capabilities.

For Horus, the Sharpa order already covers actuator development, manufacturing, and actual delivery. Next-phase revenue growth will depend on Sharpa’s full-robot sales progress and the procurement scale from other robot clients.

For Suteng Juchuang, over 3,400 robot clients form a sales network, but the required components vary significantly across applications. The company must promote spatial cameras, tactile sensors, and joint modules separately.

Both companies must sustain ongoing R&D investment.

In Q3, Horus’s strategic growth business generated RMB 90.51 million in revenue, with total operating costs and expenses reaching approximately RMB 173 million, resulting in an operating loss of RMB 82.89 million.

In the same quarter, Horus’s LiDAR business generated an operating profit of RMB 145 million, up 88% YoY. Over half of the mature business’s profit was consumed by new ventures.

Suteng Juchuang spent approximately RMB 347 million on R&D in H1 2026, with gross profit reaching about RMB 222 million. New-generation LiDARs, spatial cameras, tactile sensors, and actuators are all under continuous investment.

Both companies are entering new markets via different strategies: Horus shoulders more complex product development through deep collaboration with a key client, while Suteng Juchuang leverages existing technology and customer base to promote new components.

At the same time, companies like Zhaowei Electronics and Green Harmonic are enhancing product integration to capture greater procurement spending from robot OEMs.

This competition will ultimately reshape the supply structure for robotic joints. Which components come from specialized vendors, which modules are developed in-house by OEMs, and which R&D costs are borne by customers through contracts—all will influence the revenue and profit potential for suppliers.

Conclusion

Three years ago, Horus and Suteng Juchuang were still vying for more mass-production vehicle models in automotive LiDAR.

In 2026, both companies have established quarterly shipment scales of hundreds of thousands of units. Horus’s mature LiDAR business is now contributing operating profit, while Suteng Juchuang continues increasing deliveries of automotive and robotic LiDARs.

They have also redirected part of their technology, capital, and R&D resources into the robotic joint market.

Horus achieved first mass production of actuators via Sharpa; Suteng Juchuang leverages micro-motors, MEMS, and optoelectronic technologies to develop joints, tactile sensors, and spatial cameras.

They now face established suppliers like Zhaowei Electronics and Green Harmonic, which have long operated in precision transmission. Robot OEMs are also raising demands for component cost, performance, and delivery reliability.

Some of the engineering and manufacturing capabilities Horus has accumulated in LiDAR can be applied to new products, but core competition in joints still revolves around motors, reducers, motion control, and mechanical reliability.

Horus has secured a mass production project supported by a deep-collaboration client; Suteng Juchuang has started promoting more components to its existing robot clients.

The future performance of both companies will hinge on how many customers they penetrate, how much R&D work each project entails, and the actual scale of delivery.

Automotive LiDAR has become a business driven by scale manufacturing. The robotic joint market is forming a new supply chain division, and LiDAR vendors aim to capture a larger share.

Their competition has expanded beyond automotive ADAS sensors to encompass the entire robotic motion execution system.

Source: Automotive Financial Reports