Lead screws are still waiting for Tesla: the mass-production contrast in the humanoid robot supply chain

[[**Tesla's Optimus starts production ramp-up before its design is finalized, boosting expectations for China's supply chain, but lead screws are still waiting for mass-production demand; motors and LiDAR are already seeing large-scale shipments, while lead screws are constrained by an unconverged joint approach and cost.**

**Key points**

**1. Tesla's expectations have already been transmitted, but lead screw orders are still in the early stage of mass production**

Optimus's hand design is not finalized yet production ramp-up has still begun, and reaching 20,000 units per week will take at least a year; Xinjian Transmission has won a mass-production order from anonymous Customer A, but its related revenue in 2026H1 grew only 3.36%.

**2. Motors and LiDAR are the first to achieve large-scale shipments**

Kinco shipped 113,600 frameless torque motors in 2026H1, up 422.9%; Hesai and RoboSense shipped 260,700 and 282,600 robot LiDAR units respectively, up 165.3% and 510.4%.

**3. Rotary joints dominate and high costs suppress demand for lead screws**

In 2025, rotary joints accounted for about 94%, while linear joints accounted for only 6%; in 2026H1, humanoid shipments exceeded 22,000 units, with entertainment performances and science-education data collection accounting for more than 60%. Some companies say the cost of an all-linear design would at least double.

**4. Capacity is being expanded first, while lead screw prices and mass-production capability still need to stabilize**

This article is from the WeChat public account: Channel 42, author: x, editor: James

On Tesla's third-quarter earnings call, an analyst pressed: since Optimus's hand design is not yet fully finalized, why start the production ramp-up early?

Musk explained that Tesla is advancing robot manufacturing, supply chain construction, hardware design fine-tuning and AI software development in parallel. He then described the approach with a piece of casino slang: "We're definitely, as they say in Vegas, betting on the come."

Betting on a future that has not yet materialized—that is very Musk.

Starting early production now, reaching mass production of 20,000 units per week will still take at least a year. But this bet has already been transmitted to China's supply chain.

In mid-September, several media outlets cited people in the industry chain as saying that Tesla's robot team went to supply chain companies in the Yangtze River Delta for factory audits, involving companies such as Tuopu Group, Sanhua Intelligent Controls and Joyson Electronics.

In October, the reply to the ChiNext IPO review inquiry disclosed by Xinjian Transmission provided another angle: an anonymous leading customer, Company A, has already placed a mass-production order with it for humanoid robot lead screw products. However, the document did not publicly confirm the identity of Company A.

Over the past two years, the market has been keen to break down Optimus's bill of materials in order to identify component targets that might emerge. One focus is the planetary roller screw. It is the core transmission component in linear actuators, capable of converting the motor's rotational motion into linear push-pull motion. By distributing the load across multiple rollers, it can achieve high load capacity and stiffness within a small structural space, making it suitable for some robot joints that require greater thrust. Around this technical route, domestic companies have begun to expand capacity, and sizable long-term demand forecasts have appeared in the market.

But if we look away from Tesla and then look at the financial reports of China's robot supply chain, a contrast emerges: some components already show relatively clear shipment and order signals, while some products that the market has high hopes for are still waiting for downstream demand to materialize.

The components that Tesla has made popular are not necessarily the components that are currently the first to form large-scale businesses.

The contrast in financial reports

First, look at motors.

In the first half of 2026, Kinco's frameless torque motor shipments were 113,600 units, up 422.90% year on year; in the same period, the company's robot industry revenue was about 243 million yuan, up 62.31% year on year, accounting for 54.36% of operating revenue.

Leadshine disclosed that its order backlog for frameless torque motors exceeds 1 million units, and that it continues to advance automated production line construction. It should be noted that order backlog is not completed delivery volume, and Leadshine's robot business revenue cannot be directly used to replace the company's total revenue.

LiDAR has likewise shown large-scale signals. In the first half of 2026, Hesai's robot LiDAR shipments were 260,653 units, up 165.3% year on year; RoboSense's LiDAR sales for "robotics and other applications" in the same period were 282,600 units, up 510.4% year on year.

These figures are not all contributed by humanoid robots. Motors, motion control and LiDAR can serve mobile robots, robot dogs, industrial robots and many other products, and do not have to rely entirely on the mass-production pace of a single humanoid robot manufacturer.

The situation for lead screws is different.

In 2025, Xinjian Transmission's revenue from lead screws and related electromechanical actuator assemblies in the embodied intelligent robot field was 31.1059 million yuan, up 346.36% year on year; in the first half of 2026, it was 16.2105 million yuan, up 3.36% year on year.

The company's sales revenue from anonymous Customer A also changed significantly: 16.5875 million yuan in 2025, 1.5248 million yuan in the first half of 2026, and 224,400 yuan in July-August. The company explained that after multiple rounds of technical solution iterations between the two sides, the technical route and process parameters of the main products were gradually finalized, procurement demand during the earlier R&D trial-production stage declined somewhat, and cooperation began to enter the early stage of mass production.

Company A has already placed mass-production orders, but there is still a time lag between order execution, customer production ramp-up, delivery and revenue recognition.

Why can some components show large-scale signals first, while lead screws are still waiting?

One reason is that mature manufacturing systems and diverse customers already exist behind motors, motion control and LiDAR. Robots change product size, performance requirements and operating environments, but suppliers do not have to build an entire manufacturing capability from scratch.

Another reason goes back to the robot itself: what joint solution different manufacturers choose determines which components can win orders.

Tesla can drive expectations, but not more demand

To understand this demand difference, first look at how the joint solution for the robot itself is chosen.

Tesla can amplify market expectations for one technical route, but it cannot decide demand for the entire Chinese robot market.

From the perspective of joint motion forms, rotary joints still dominate. Guanyan Tianxia estimates that in 2025 global demand for humanoid robot joint modules, rotary joints will account for about 94% and linear joints about 6%.

The difference between the two solutions is not just that one uses motor rotation and the other converts rotation into linear motion. The mechanical structure, control scheme and supply chain accumulation of rotary joints are relatively mature, making it easier to develop and debug on existing architectures. By comparison, linear drive has advantages in some high-thrust, high-stiffness scenarios, but cost, structural layout and control requirements also affect the overall machine design.

The performance advantages of lead screws need to be weighed through overall machine design before they can translate into actual demand.

Looking at actual scenarios, the demand rhythm for lead screws is likewise slow. Current humanoid robot shipments are still mainly for display, scientific research and data collection, while industrial scenarios with high loads and long periods of continuous operation account for a limited share. The Embodied Intelligence Development Report (2026) released by institutions including the China Academy of Information and Communications Technology shows that in the first half of 2026, global humanoid robot shipments exceeded 22,000 units, of which entertainment and commercial performances plus science and education data collection together accounted for more than 60%, guided tours and shopping guides about 19%, intelligent manufacturing about 13%, and warehousing and logistics about 5%.

The humanoid robot market has not yet formed a unified demand structure dominated by industrial operations. Scenarios have not converged, so routes have not been finalized. Different robot manufacturers, based on their own cost structures and target customers, have made different trade-offs regarding load, cost, space and motion performance.

Unitree Robotics stated in public materials that its joint drive mainly adopts a high-bandwidth force-controlled quasi-direct-drive planetary rotary joint route characterized by rotary drive; Kepler K2 adopts a hybrid architecture combining planetary roller screw linear actuators with rotary actuators.

Zhang Haixing, founder of Matrix Superintelligence, who once worked at Tesla, also mentioned in an October interview with Southern Weekly that its second-generation robot had almost entirely used linear joints, but was not released externally because costs were too high; the third-generation MATRIX-3 switched to a hybrid "rotary + linear" drive. Zhang Haixing said the third-generation product reduced lead screw usage by about 40% compared with the previous solution. He also said that if all rotary joints were replaced with linear joints, the cost of the whole machine would at least double.

This is only the product experience of a single company, but what can be glimpsed is that even if the performance advantages of linear joints are recognized, cost will force robot manufacturers to recalculate.

The trade-offs of a single company may seem minor, but the independent decisions of many robot manufacturers become the biggest variable in the supply chain. It is difficult for suppliers to lock down processes in advance around a stable solution as they did in the automotive era. This differs somewhat from the logic of the partially scaled supply chains that have already formed in the automotive industry.

Tesla's influence in the new energy vehicle sector comes not only from the technical choice of a certain component, but also from vehicle sales, a mature manufacturing system and long-term supply chain collaboration. After a technology is adopted at scale, suppliers can invest in equipment and improve processes around relatively stable product demand, and then reduce costs through scaled production.

But in the humanoid robot sector, the body design, application scenarios and technical routes for core components are still being explored simultaneously.

China's supply chain dividend also makes daily iteration possible for robot manufacturers. According to a Southern Daily report, the "Robot Valley" in Nanshan District, Shenzhen, has gathered more than 200 robot industry chain companies, from precision machining, motor drives and sensors to complete machine assembly, and many suppliers are less than half an hour's drive apart. Zhao Tongyang, founder of Zhongqing Robotics, said companies can draw designs in the morning, make prototypes in the afternoon and iterate the next day, compressing the R&D cycle from "month-level" to "day-level."

This iteration efficiency, for precision transmission parts such as lead screws, is both an opportunity and an uncertainty.

When robot manufacturers are still trying different joint architectures, suppliers that can respond quickly and cooperate with revisions have a chance to participate in more product development first.

But motors, LiDAR and some precision manufacturing services may accumulate shipments through multiple customers and different robot forms, while demand for lead screws remains stuck on specific technical routes. Once the solution is adjusted, changes will be transmitted all the way up the supply chain. A supply chain staff member said of the modification and finalization of the lead screw in Optimus's hand: "Even the wiring in the motor has to be redone."

Demand for planetary roller screws still depends on humanoid robot body shipments, the share of models using linear drive, and the actual number of lead screws used per robot.

Micro ball screws depend on dexterous hands. GGII estimates that in 2025 China's dexterous hand market sales will be about 19,200 units, and are expected to reach 70,200 units in 2026. But the technical routes have likewise not converged, with linkage, tendon, direct drive and hybrid solutions all in use.

Technical value needs to pass through body architecture, actual shipments and per-unit usage before it can be transmitted to suppliers. Therefore, the long-term market space for lead screws can be large, but current revenue is still limited.

The distance from prototyping to mass production

The constraint on the demand side is only half the story. The other half is hidden on the factory floor.

A supply chain practitioner said that as robots go from 1 unit to 10 units, 100 units, and 1,000 units, the manufacturing system also undergoes a series of changes.

A single robot can be assembled by hand by a few engineers, and if problems arise, it can be readjusted; at 10 units, production processes need to be sorted out, and SOPs (standard operating procedures) and SIPs (standard inspection procedures) need to be established; at 100 units, traceability of core components, incoming material inspection, and production testing tools need to be improved; and as production further expands to 1,000 units, the production method may need to shift from work islands to assembly lines, with clear division of labor at workstations and inspection mechanisms established between processes.

He also mentioned that robot products iterate very quickly. To compress prototyping time, companies sometimes reduce reliability verification, with the result that problems such as defective incoming materials are only exposed once they reach the production line.

The more frequently products are revised, the more suppliers need to respond quickly; but if verification processes are compressed, design defects and quality problems may emerge in concentrated fashion in subsequent production. Mass production is not about making more prototypes, but about maintaining consistency and stability of product quality in continuous delivery.

For precision transmission parts such as lead screws, this is especially important. Suppliers need to prove not only that a given lead screw can meet precision requirements, but also whether products from different batches can remain consistent, whether the yield rate in mass production can be stable, and whether the product is reliable under repeated loads and long-term operation.

The same logic applies to cost.

Data disclosed by Xinjian Transmission shows that the average selling price of its planetary roller screws fell from 1,652.31 yuan in 2024 to 758.83 yuan in 2025, and rebounded to 781.29 yuan in the first half of 2026. The pricing system for this business is still changing. The company explicitly explained that as customer verification advances and order scale increases, product prices and gross margins will face new changes.

This means that cost reduction after mass production is not just about making more of the same product, but may also involve redesigning products, changing processes, and finding a new balance among performance, reliability and cost.

Industry figures also mentioned that if a robot is to drop from 200,000 yuan to 100,000 yuan, it cannot rely solely on supplier price cuts or competitive negotiations; rather, it requires value analysis and value engineering to re-examine product structure, component selection and design solutions.

In other words, if the robot maker adjusts the joint architecture, the equipment and processes originally built around a certain type of lead screw may not be directly transferable to other products.

Back to Xinjian Transmission: in 2025, its designed production capacity for lead screws and related electromechanical actuator assemblies was 1.5 million units, with output of about 1.09 million units and sales of about 980,000 units, for an overall capacity utilization rate of 72.7%. Its IPO fundraising project also plans to build capacity for both humanoid robot and automotive planetary roller screws at the same time. Wuzhou Xinchun and Beite Technology are also advancing related projects.

Companies expanding production in advance may help them win future customers, but it also means they must make judgments about future demand.

As robot delivery volumes increase from dozens of units to hundreds and then thousands, requirements for suppliers will also change. In the early stage, rapid service and responsiveness are valued more; once customers begin stable batch delivery, they will pay more attention to suppliers' quality systems, equipment capabilities, production processes, process recommendations and continuous delivery capability.

For lead screw suppliers, isn't pre-planned capacity also a kind of bet?

Lead screws have already waited two years for Tesla, but the story they are waiting for is not over yet. AI industry signal channel: Frontier Technology. This content comes from the internet and the original link; the views represent only the author and do not represent Huxiu's position. For copyright issues, please contact hezuo@huxiu.com, and we will verify and handle it promptly. Everyone who is changing and wants to change the world is on the Huxiu APP[ ]