Lingyi iTech's Beijing Super Factory Targets 500,000 Humanoid Robot Units by 2030

2026 is seen as the inaugural year of physical AI, where embodied intelligence moves beyond stage and scenic performances to take on real work in factories, inspection, retail stores, and other industrial and service scenarios.

However, as the industry matures, challenges such as many humanoid robot prototypes but difficult mass production, high manufacturing costs, and low supply chain scalability have become prominent. Recently, reporters visited the Beijing Lingyi Embodied Intelligence Super Factory as part of a new industrialization media tour organized by the Ministry of Industry and Information Technology. This factory is the first in the Beijing-Tianjin-Hebei region capable of ramping annual capacity from 10,000 to 500,000 units, covering full-chain manufacturing services from core components and modules to complete assembly and testing, filling the gap in large-scale embodied intelligence manufacturing in the region.

Xia Wei, production head of Lingyi iTech's Beijing Super Factory, said the factory's annual capacity is 10,000 units in 2026 and is planned to reach 500,000 units by 2030. The factory is divided into multiple functional areas, including component assembly, module assembly, module testing, and complete machine assembly. The component assembly line is a flexible line that can simultaneously produce different models of encoders, drives, sensors, and other core components required by more than 30 robot company customers. The module assembly line uses force and vision dual-guided precision assembly to assemble scattered parts into functional modules such as hands, upper limbs, and lower limbs, with a changeover time of less than 15 minutes. The complete machine assembly line has 10 adjustable assembly stations and an automatic conveyor line, where workers complete takt-based assembly of lower limbs, torso, head, and arms, followed by software deployment, functional testing, complete machine testing, and packaging, with a total manufacturing time of about 40 hours.

The factory's parent company, Lingyi iTech, is a listed manufacturer with nearly 20 years of experience in precision CNC machining for consumer electronics. Xia Wei said that after humanoid robots move from R&D to manufacturing, much of the process involves CNC and mold processing, and Lingyi has the capability for large-scale precision parts manufacturing. "Previously, the precision parts we made were phone-sized; now we are just scaling them up," he said. Many humanoid robot companies focus on prototypes during R&D, perhaps hand-building 1 to 3 units in their own R&D bases, but when entering mass production, investing in a factory solely for that capacity is relatively costly, while dedicated contract manufacturing factories offer better cost advantages.

One of the core reasons for the high cost of humanoid robots is that the supply chain has not yet been scaled, with core components still in customized, small-batch production, unable to amortize high fixed costs and R&D investments through large-scale standardized production, leading to the common industry pain point of easy R&D but difficult mass production. To achieve mass production, engineering reliability and consistency are hard thresholds. Changing the supplier of a single motor can cause fluctuations in overall performance; poor tolerance control of a structural part can significantly reduce assembly efficiency.

Lingyi iTech has micron-level precision and million-level consistency capabilities in consumer electronics, which can be transferred to robot manufacturing. Xia Wei introduced that the factory intervenes from the design stage, providing manufacturability optimization and large-scale assembly testing, including precision testing, electromechanical testing, environmental simulation testing, safety compliance testing, and reliability verification. The torque of each screw, calibration data of each joint module, and motion curve of each complete machine are bound to a unique product SN code. If a quality defect is detected, through cloud data closed-loop and AI root cause analysis, quality closed-loop optimization can be completed within 24 hours.

Counterpoint Research shows that global humanoid robot shipments exceeded 22,000 units in the first half of 2026, a year-on-year increase of nearly 300%, and full-year global humanoid robot shipments are expected to exceed 50,000 units.