Summary of Key Points
YuShu Technology’s IPO has sparked a capital frenzy (nearly ten million people applied, with a win rate of just 0.018%), resulting in a market value of 61 billion yuan, representing a 16-fold increase in valuation in less than two years. However, the actual market demand for these robots amounts to only 1.838 billion yuan, creating a stark contrast between the two figures. The humanoid robotics industry is still in its early stages: capital is investing in future possibilities, while real demand primarily comes from schools for research and education purposes as well as state-owned enterprises (SOEs) for industrial development. Current applications are mainly focused on low-complexity tasks, indicating a long way to go before widespread commercialization.
Detailed Analysis
1. Capital Is Boiling Hot, but the Real Market Is Still Focusing on Fundamentals
The overwhelming interest in YuShu’s IPO (9.78 million applicants and 8,000 times excess subscription) reflects the market’s enthusiasm for humanoid robots. However, actual purchase data shows that from after the Spring Festival of 2025 to June 2026, there were only 302 projects involving a total of 1.838 billion yuan—a pittance compared to the company’s market value of several hundred billion yuan.
Why such a disparity? The market growth is driven by a few large-scale projects; for example, a smart robotics training facility in Suzhou accounted for 60% of the monthly sales, and a major order from a Sichuan SOE for 236 million yuan was a one-time investment in infrastructure. Policy support also played a role, with the Ministry of Industry and Information Technology issuing guidelines in June 2026, leading to record numbers of projects and transaction amounts that month. Overall, the market has not yet reached a state of widespread enthusiasm; orders are sporadic, suggesting that the industry is still in its exploratory phase.
2. Who Is Buying Robots? Schools and SOEs Lead the Way, but Their Needs Differ
The purchasing market can be divided into two groups:
- Schools: They account for 68% of projects but make up small purchases (ranging from tens of thousands to several hundred thousand yuan) for laboratory use or as teaching tools. Most of these projects (193 in total) are for research and educational purposes.
- SOEs: They spend 74% of the total funds, with individual orders often reaching tens of millions or even hundreds of millions of yuan,主要用于 building industrial bases and data collection centers. A few companies (2 projects) accounted for another 7% of the total spending, but they purchased comprehensive solutions that include both robots and related infrastructure.
In short, schools use robots to teach students, while SOEs use them to build industrial ecosystems.
3. What Can Robots Do? For Now, They Are Mainly Used as Teaching Tools and Data Collectors
The current uses of robots are limited:
- Research and Education: 42% of sold robots are used in teaching, such as in school laboratories.
- Limited Real-World Applications: Some robots are used for data collection (testing them in various environments to train their “intelligence”) or for demonstration and inspection tasks (e.g., checking equipment in factories). However, these tasks are relatively simple, and complex industrial logistics applications account for only 4%.
This is due to the robots’ limited generalization capabilities—their ability to adapt to unfamiliar situations is still weak, similar to early-stage home computers. YuShu’s CEO has rightly stated that the industry has not yet reached a commercialization stage.
4. Industrial Distribution: Concentrated in the East, with Larger Orders in the West
Regional distribution patterns are evident:
- Eastern Coast: Regions like Guangdong, Zhejiang, and Jiangsu have a higher number of projects due to their strong manufacturing and research institutions, which are conducive to testing and development.
- Western Regions: Although there are fewer projects (e.g., Sichuan with 16 projects but 417 million yuan spent), the amount is significant (22.7% of the total). This money is used for building training facilities and data collection centers.
Local governments are actively investing in these areas to attract companies, creating “robotics training camps” to accumulate data and develop technology, aiming to secure a share of the future market. Market growth is driven by investment rather than consumer demand.
5. The Paradox of YuShu’s Success: Why Is It Still Called an Early Stage?
YuShu sold 5,500 robots in 2025 and earned 600 million yuan, making it one of the few companies in the world to profit from humanoid robots. Yet its CEO insists that the industry is still in its early stages. This seems contradictory, but it’s not:
- YuShu’s profits may come from sales to schools and SOEs (e.g., educational versions of robots), but these purchases are more related to infrastructure investments rather than large-scale commercial applications.
- Fundamental issues with robots remain unresolved, such as limited generalization capabilities and high costs, which lower market acceptance. For instance, robots cannot operate as flexibly as humans on factory assembly lines, so they are currently used mainly in laboratory and pilot scenarios.
YuShu’s cautious approach serves to remind the market that the current excitement is based on expectations, not reality; the industry will need another two to three years to reach a critical point.
Conclusion
The story of humanoid robots represents the clash between “future imagination” and “real-world implementation.” Capital sees the potential of these technologies in a decade’s time, while the real market is focused on building the foundation: establishing infrastructure, developing technology, and accumulating data. This gap is not a bubble but a necessary part of the growth process for emerging industries—just like the early internet era, where infrastructure development preceded widespread adoption. Everyone is waiting for that critical point, but we may only have two to three years left before it arrives.