Humanoid Robot Integrated Joint Modules: The Real Test of Engineering and Mass Production Seen from CIIF Exhibits

This article is reposted from Juncai Magnetic Applications, with copyright belonging to the original author; the original link is provided at the end under "Sources".

Humanoid Robot Integrated Joint Modules: The Real Test of Engineering and Mass Production Seen from CIIF Exhibits

For humanoid robots to truly enter factories, the key lies not in a few steps on stage, but in whether each joint can withstand daily wear and tear, instantaneous impacts, and long-term precision degradation.

At the 26th China International Industry Fair (CIIF), this industry shift was directly confirmed. In the robotics zone, while whole-machine demonstrations remained lively, the focus of supply chain engineers and procurement managers shifted to the integrated joint modules hidden inside the machines. Compared to the previous two years, exhibitors at this year's fair are no longer keen on showing off humanoid robot parameters such as degrees of freedom and walking stability. Instead, they emphasize "what tasks it can perform and how long it takes to recoup the investment."

This is not merely a shift from showing off to pragmatism, but a change in the underlying narrative logic of the industry. Capital market funds are flowing from whole-machine hardware to dexterous hands, force-controlled sensors, centralized control systems, and process software. These are the core links that enable robots to perform practical tasks. The entire industrial humanoid robot track is experiencing a "disenchantment": the industry has crossed the prototype validation stage of "whether it can move" and entered a critical period of engineering and mass production, with stability, fatigue life, and batch consistency at its core.

1. Two Types of Joints, Two Different Strategies

The joint industry chain can be divided into three layers: the upstream consists of basic materials and core components, including special alloys, rare earth magnetic materials, precision bearings, micro motors, reducers, and encoders; the midstream is the joint and module integration link, where integrated module manufacturers need to master mechanical transmission, electronic control development, sensor calibration, and automated assembly processes simultaneously; the downstream comprises humanoid robot OEM enterprises and various application scenarios.

The selection logic for trunk heavy-load joints and end-effector finger micro-joints is completely different. Trunk large joints mostly use frameless torque motors, paired with harmonic, RV, or cycloid reducers. Dexterous finger joints rely on coreless motors, micro planetary gearboxes, and micro sensors to complete power output in extremely small spaces.

The current supply chain shows structural differentiation: high-end micro coreless motors and micro encoders are still dominated by overseas brands, while domestic enterprises are accelerating breakthroughs in precision gear processing, structural parts, and mid-to-low-end sensors, promoting domestic substitution.

At this year's CIIF, multiple manufacturers exhibited force-controlled integrated joint modules and lightweight harmonic reducer products. The force-controlled integrated joint module exhibited by Midea Welling can perceive micro torque changes of 0.02N·m, focusing on high-precision force control capabilities. Liangzhi Joint also brought a planetary integrated joint module, integrating the motor, reducer, and encoder into one, reducing the integration difficulty for OEMs. Many enterprises have already planned automated production lines, taking the ability to deliver tens of thousands of units as their core competitiveness. The competition in the trunk large joint track has shifted from parameter comparison to the comparison of scaled delivery capabilities.

At the same time, the industrial value of dexterous hands cannot be underestimated. According to calculations by Guojin Securities, the single-machine value of dexterous hands can account for up to 32% of the BOM (Bill of Materials) cost of the whole machine, making it a high-value, high-barrier segmented track in the humanoid robot industry chain. The core carrier of the dexterous hand is precisely the finger joints.

Finger joints are constrained by narrow installation spaces, requiring high integration of micro motors, micro reducers, and sensing elements, posing strict requirements for lightweight design, power density, and low backlash. The industry generally favors the technical route of built-in integrated micro-drive technology. However, the miniaturization processing of micro reducers, batch manufacturing consistency, and scaled cost control remain core engineering problems hindering the mass production of solutions.

Zhaowei Electromechanical is a representative domestic manufacturer in this track. Relying on long-term accumulation in micro transmission, its ZWHAND series includes multiple dexterous hand products such as C06, B06, and B21, developing modules like finger joint motors, thumb parallel linear motors, and palm linear drive motors. The technical route abandons traditional tendon-driven mechanisms, integrating motors, transmission, sensing, and structural parts entirely within the palm and fingers. This "joint built-in power unit" technology achieves independent motion control of finger joints, balancing dexterous operation and service life.

2. Supply Chain Enters the Window for Sample Volume Expansion

Signals from multiple industry chains show that the joint module supply chain is transitioning from the prototype trial production stage to the preparation for scaled delivery.

On October 8, Changying Precision disclosed that from January to August 2026, the delivery volume of precision components for humanoid robots exceeded 1.1 million units. In the second half of the year, the order growth rate increased, and upstream component sample orders continued to be released. On October 10, a supply chain review by 36Kr pointed out that dexterous finger joints belong to an independent segmented track. The core barrier is the integrated integration of micro gearboxes, micro motors, and electronic control sensing. Solution providers with system integration capabilities are more likely to enter the supply chain of leading OEMs.

There are many players in the trunk large joint sector with mature routes, and the main line of competition is cost reduction. In contrast, participants in the end-effector micro finger joint sector are scattered, and most enterprises are still in the sample or small-batch trial stage. The track barriers lie not only in part processing; micro-gear micron-level processing, heat treatment, module assembly, and sensor calibration will all affect the lifespan and consistency of the final product. Enterprises with both transmission design and module integration capabilities are more likely to pass multiple rounds of whole-machine validation.

According to IDC data, in the first half of 2026, global humanoid robot shipments reached nearly 25,000 units, a year-on-year increase of 432.1%; China's market shipments exceeded 19,000 units, accounting for about 77.9% of the global total.

The dexterous hand track has formed a parallel pattern of full-drive and under-drive, with no substitution relationship between the two. Full-drive has high degrees of freedom and strong fine operation capabilities, but higher hardware costs. Under-drive relies on a small number of drive units to achieve coordinated grasping, with lower material costs, suitable for standardized industrial grasping, and easier to scale in the short term. OEMs will flexibly choose the degree of freedom scheme based on their application scenarios, cost budgets, and commercialization pace.

3. Mass Production is Not Merely Replicating Samples

While the track is expanding, engineering challenges remain prominent.

Whether it is trunk large joints or end-effector micro finger joints, mass production consistency, fatigue life, and scaled costs are common industry problems. It is easy to achieve ideal performance indicators in a laboratory environment, but under working conditions such as high and low-temperature alternation and continuous impact loads, joints are prone to precision degradation and insufficient lifespan.

The core contradiction in the current industry chain can be summarized as: sample performance can meet the standards, but the scaled yield rate is insufficient; single-point component breakthroughs have been achieved, but the joint validation cycle for the entire system is lengthy.

The dexterous hands on the exhibition booths can perfectly execute preset actions, but in real production lines, screw positions may deviate, incoming materials have tolerances, and lighting fluctuates. If the hardware cannot handle these variables, it is the biggest gap standing in the way of industrialization. In the short term, orders for various joint modules are still mainly pilot projects. Large-scale commercial landing requires continuous, long-term working condition validation.

In the humanoid robot integrated joint module track, there will be no single technical solution that dominates the market. Competition in the trunk joint track is tending towards a red ocean, while the end-effector dexterous hand micro joint track is still in the supply chain finalization stage. Different OEMs have not yet formed unified standards for their needs regarding degrees of freedom, weight, and cost. This leaves differentiated industrial opportunities for solution providers deeply engaged in micro transmission and micro drive.

In summary, the track has crossed the principle validation stage and entered the critical period of engineering and mass production. The short-term competition focus is concentrated on production line yield control, continuous cost reduction capabilities, and long-term reliability in industrial scenarios. The mid-to-long-term industrial landscape will depend on the localization level of upstream components, midstream module integration capabilities, and the commercial landing pace of downstream whole machines.

The reliability iteration and scaled manufacturing capabilities of integrated joint solutions will directly determine the landing speed of the entire humanoid robot industry.

Sources

• Sohu (2026-10-17)
• Original link: Humanoid Robot Integrated Joint Modules: The Real Test of Engineering and Mass Production Seen from CIIF Exhibits
• https://m.sohu.com/a/1088558308_104421?scm=10001.325_13-325_13.0.0-0-0-0-0.5_1334