Summary of Key Points
This news article, from the perspective of Shoucheng Holdings' Kang Yu, explores the industry value of the second World Humanoid Robot Games: it serves not only as a platform for robots to showcase their capabilities but also as a "stress test" to reveal industry-wide issues. The robotics industry is still in its infancy (2026 marks the beginning), and although autonomous capabilities have improved, robots are still far from being "practical" (i.e., capable of running stably for extended periods and integrating into daily life and work scenarios). Investors like Shoucheng do not focus on a single technological approach (humanoid or wheeled robots) but instead invest in the entire supply chain, including the robots themselves, components, data, and related models. They also place great importance on the teams and leading enterprises within the industry. Education and research (such as the participating high school students) are seen as the "seeds" for the industry's future growth. For robots to truly become a part of our lives, they must overcome numerous challenges related to reliability, data collection, mass production, and commercialization.
1. The Robot Games: Not About Speed, but a "Health Check" for the Industry
Kang Yu compares the games to a "reliability test under the spotlight"—failures such as robots falling, running out of power, or experiencing component malfunctions on the field are all issues that must be addressed before mass production and practical applications. For example:
- Some companies bring more than a dozen robots to the competition, indicating that consistency is a significant challenge (each robot must perform similarly during mass production, not just one being fast).
- Robots falling while turning represents real-world scenarios; if they are to be used in households, their stability during turns is crucial for safety.
- Robots running out of energy after a short distance highlights the need for improved battery life—after all, we don't want robots to need constant recharging after just half an hour of use.
These seemingly "unappealing" failures are actually more valuable than breaking records, as they help identify problems early on, preventing accidents once products hit the market.
2. The Robotics Industry is Still in its "Kindergarten Stage": 2026 is Just the Beginning
Kang Yu emphasizes that the industry is still in its early stages, meaning robots are not yet ready for widespread use. Although there has been progress in hardware (such as faster speeds and more durable batteries), achieving long-term, reliable operation is a long way off. For example, household robots need to be able to perform tasks like mopping floors or cooking every day without breaking down, and industrial robots must work continuously for 8 hours without failure. Current robots are more like laboratory prototypes, with a long way to go before becoming part of everyday life.
3. Investors Do Not Bet on Humanoid Robots Alone: They Invest in the Entire Supply Chain and Focus on Teams and Leading Enterprises
As an investor, Shoucheng does not specialize in humanoid robots alone; they invest in both bipedal and wheeled robots, as well as other forms of embodied intelligence. The reason is simple: embodied intelligence combines AI with the physical world, and robots don't necessarily have to resemble humans. However, humanoid robots have their advantages, as most labor-related data is generated by humans, making them more suitable for training with such data. Their investment strategy focuses on the entire supply chain, including:
- Upstream: components (high-cost and prone to issues, such as joints), materials, data, and computing power.
- Midstream: robot bodies (e.g., Yushu Technology) and models (e.g., Xinghai Tu).
- Downstream: practical applications.
They also focus on teams and leading enterprises that can drive the development of the entire industry, as the industry is still in a consolidation phase, and only companies with core capabilities will survive.
4. High School Students Debugging Robots in the Middle of the Night: The Future of the Industry
Kang Yu highlights a group of high school students participating in the competition. They are debugging robots in the venue until midnight, some waiting for code to compile while doing their homework, and immediately getting up when problems arise. She believes these students represent the industry's future; although their efforts may not have immediate commercial value, they will cultivate engineers and entrepreneurs. The students who debug robots today could become the driving force behind future industry advancements.
5. From the Competition Field to Everyday Life: Robots Must Overcome Multiple Barriers
For robots to become truly useful, several key issues need to be resolved:
- Reliability: Failures identified in the competition (falls, power outages) must be completely addressed.
- Data Challenges: There is a lack of data for embodied intelligence; many datasets need to be collected manually, unlike in autonomous driving, where data is collected naturally during use.
- Mass Production and Cost: Without solving consistency issues, mass production will be impossible, and costs will remain high.
- Commercialization: There is a lack of "killer applications" (such as essential services like WeChat or food delivery).
Kang Yu describes the industry as being in a period similar to the Spring and Autumn or Warring States periods, with no dominant players yet. However, each breakthrough in the competition is a step towards practical applications.
In summary, the robotics industry is full of potential, but it must first overcome practical challenges. There is still a long way to go from the experimental stage in the competition fields to widespread, practical use in our lives.