As embodied intelligence moves toward mass production, why could data collection become a critical bottleneck?
Industrial Securities believes there is still a gap between the mass production of humanoid robots and the maturity of embodied intelligence, and that real-world data could become a scarce resource for breakthroughs in physical AI. An industry-chain estimate of a data-collection market exceeding 60 billion yuan depends on two assumptions: the scale of the robot hardware market in 2030 and the share of costs attributable to data collection.
Industrial Securities expects data collection to rise from a supporting function for robots to a critical bottleneck. Its commercial opportunity depends both on the scale of the robotics industry and on demand for real-world data collection and data services.
If robots are being mass-produced, why is more data still needed?
Industrial Securities stresses that mass production of humanoid robots does not mean embodied intelligence has matured: being able to produce the hardware and achieving breakthroughs in physical AI capabilities are two different processes. The firm regards data as the scarcest foundational resource and data collection as the source of that resource, so collection could rise from a supporting function to a critical bottleneck.
The time frame for this assessment comes from Wang Xingxing, founder of Unitree Robotics: signs of a breakthrough in physical AI could appear in as little as 1 to 2 years, or as long as 2 to 3 years. This refers to early signs of a capability breakthrough. The rationale for demand for data collection is to supply data for this process, rather than merely to expand in step with robot shipments.
How is the market opportunity of more than 60 billion yuan calculated?
According to industry-chain estimates, the global humanoid robot hardware market will be worth approximately 250 billion yuan in 2030. Assuming data-collection hardware and data services account for 25% of total costs, the corresponding data-collection market is estimated at more than 60 billion yuan. This estimate depends on both expansion of the robotics market and the 25% cost share; it cannot be assessed solely by looking at growth in robot numbers.
The material also presents an upside scenario: if demand for upstream components is greater, the overall opportunity could increase further. This should be understood separately from the baseline estimate of more than 60 billion yuan, which already includes both collection hardware and data services.
How are collection methods and pricing models changing?
Collection is moving from enclosed venues to factories, chain retail stores and homes. Distributed collection allows operators to “collect data while working.” The change involves more than adding collection locations: it brings data directly from real working environments. On this basis, the material stresses that the scarcity and irreplaceability of such data are becoming more pronounced.
Data service providers mainly charge by the hour, with package prices negotiable for large purchases. Data can be sold to third parties, including robot manufacturers, or used by the collectors themselves. Therefore, whether expanded data-collection activity translates into greater revenue from data services sold externally depends on two further factors: how customers purchase and how the data is used.
