Humanoid Robots Are a Components Trade Before They Are an AI Trade

The economic unit is the joint

The most important supply-chain object in a humanoid may be the actuator rather than the complete robot. A rotary joint can combine a frameless torque motor, harmonic or RV reducer, bearings, encoder, and force or torque sensor. Linear movement can add planetary roller screws. Dexterous hands require another group of small motors, gears, encoders, bearings, and sensors. Cameras, lidar, inertial sensors, tactile sensors, battery cells, and power electronics then surround that motion system.

This structure matters because robot production multiplies component demand. Selling one more humanoid does not create demand for one motor or one bearing. It creates demand across many joints, with different specifications for hips, knees, shoulders, elbows, hands, and other moving assemblies. If industry volumes eventually move from thousands of robots into hundreds of thousands, the component ramp can be much steeper than the finished-unit count suggests.

Chinese public companies already appear throughout this stack. Leader Harmonious Drive Systems, also known as Leaderdrive, trades in Shanghai as 688017 and produces harmonic reducers, a high-precision transmission component used inside robot joints. Zhejiang Shuanghuan Driveline, 002472 in Shenzhen, is positioned in RV reducers. Leadshine Technology, 002979, participates in frameless torque motors, while Shanghai MOONS' Electric, 603728, and Shenzhen Zhaowei Machinery & Electronics, 003021, sit in small motor and motion-control categories relevant to dexterous hands and compact actuators.[2]

The sensor layer creates a separate path into the theme. MERICS identifies Hikvision, Luster LightTech, and TZTEK in Chinese machine vision, while Hesai and RoboSense participate in lidar. These companies matter because increasingly capable AI does not eliminate the need for physical perception. It increases the usefulness of better cameras, depth sensing, force measurement, and feedback at each joint. A robot that can reason but cannot reliably measure contact force, limb position, balance, or its surroundings remains difficult to deploy outside controlled demonstrations.

Power creates another crossover from China's existing industrial base. EVE Energy, CATL, Sunwoda, and Jiangsu Azure are among the battery companies identified within the emerging humanoid supply chain.[2] Their existing manufacturing scale is important, but it also illustrates why supplier exposure must be measured rather than assumed. Humanoid robots could become a major new battery application while remaining economically small for a giant battery producer. The same robot program could be far more significant to a specialized reducer, sensor, or miniature motor supplier.

The less obvious exposure sits upstream of the motor

Magnets push the network another relationship upstream. High-performance motors depend on permanent magnets, which connects humanoid production to China's rare-earth processing and magnet manufacturing ecosystem. JL MAG Rare-Earth, listed as 300748 in Shenzhen and 06680 in Hong Kong, is already moving beyond supplying magnetic material. The company says it has created a dedicated humanoid robot motor rotor business, developed automated production lines for those rotors, and achieved small-batch deliveries to customers.[3]

That is a more revealing signal than another robot prototype. JL MAG is attempting to move from supplying a material into supplying a precision component closer to final assembly. If successful, more of the value associated with a motor can remain with the magnet and rotor producer rather than passing to another intermediary. It also shows how the humanoid supply chain can pull established EV and industrial suppliers into new product categories without requiring them to become robot manufacturers themselves.

This is where company-level relationship analysis becomes important. The relevant questions are not merely whether a company has humanoid exposure, but how asymmetric that exposure is. A major robot OEM could represent a small experimental customer for a large battery producer while becoming a meaningful source of incremental demand for a smaller reducer or motor company. Several competing robot makers could also depend on the same upstream component supplier, creating concentration that would not be visible from conventional sector classifications.

Qualification can reinforce that effect. Motors, reducers, bearings, screws, sensors, and machined parts have to work together inside tightly integrated joints. Once a robot design has been tuned around a particular component's dimensions, backlash, torque, weight, durability, and control characteristics, switching suppliers can require more than finding another factory with nominally similar capacity. That makes precision manufacturing capability and customer qualification potentially more important than headline production capacity.

Conclusion

Sources

How to Cite This

For research inquiries or data access: press@altsets.com

Go Deeper