During the 2025 World Robot Conference, I listened to the speech of Wang Xingxing, the CEO of Yushu Technology, and his media interview, where he discussed various aspects of the robotics industry. The audience seemed to gain a lot from it.
He mentioned that the robotics industry has been booming in the first half of this year, thanks to market demand and policy support. Whether it is manufacturers of complete machines or parts, almost every company has seen a growth of 50% to 100%, which is quite rare in the entire industry. In simple terms, it is the demand for robots that is driving the entire industry forward.
Regarding AI, Wang Xingxing believes that both the AI and robotics fields need global collaboration for development. Over the past few years, no large company has been able to hold enough resources to remain the leader in the AI field indefinitely. Historical evidence shows that new breakthroughs in AI often come unexpectedly, and they are often driven by innovative young minds. Many times, it is the combined efforts of various companies and universities that push this field forward.
He pointed out a major issue in the robotics industry today: an overemphasis on data, which leads to neglecting the models. The biggest problem lies in the models themselves; the current robotic model architectures are neither user-friendly nor standardized. Even if there is a lot of good data available for training, it cannot be effectively utilized.
When discussing the AI capabilities of humanoid robots, Wang Xingxing made an analogy, saying it is similar to the one to three years before ChatGPT was released. In the entire industry, everyone sees a similar direction and technological path, but no one can bring it to fruition. In the past, voice AI took over a decade to develop, and everyone thought it was clumsy and not useful until ChatGPT emerged, marking a critical turning point. Currently, humanoid robots have not yet reached that pivotal moment. If one day a robot can be taken to an unfamiliar venue and can understand a command like “hand that bottle of water to that audience member,” then we would be approaching the ChatGPT moment for robots.
So when might this moment arrive? Wang Xingxing believes that if all goes well, it could be in one to three years, or at most three to five years.
In the next two to five years, intelligent robot technology needs to focus on several directions: developing a unified, end-to-end intelligent robot model; creating cheaper and more durable hardware; achieving ultra-large-scale production; and developing low-cost, large-scale computing power.
Interestingly, Wang Xingxing mentioned that he was previously very opposed to developing humanoid robots, and Yushu Technology had not ventured into this area until 2022. However, with the rapid advancements in AI technologies like ChatGPT, and the keen interest from figures like Elon Musk in cutting-edge technologies, clients began approaching Yushu, directly offering to pay for humanoid robots, which prompted Yushu to start developing them.
He hopes that the robots produced by the company will be versatile humanoid robots capable of working in various environments, whether in factories, performance venues, or homes, rather than being limited to specific scenarios. Currently, most employees at Yushu, including himself, are busy training robots to work in diverse settings.
Regarding the company’s IPO, Wang Xingxing stated that they are progressing through a standard listing process. He likened the IPO to a college entrance examination, as it represents a phase in the company’s journey towards more mature management and operations, as well as a reflection on the past nine years.
Finally, Wang Xingxing believes that in the coming years, the annual shipment volume of humanoid robots in the entire industry could double without issue. If there are significant technological breakthroughs, it is possible that in the next two to three years, a year could see sales of hundreds of thousands or even millions of units.
The excitement in the robotics industry is indeed quite evident, as seen from the growth of manufacturers indicating strong market demand. Moreover, not only is there potential for a quantitative explosion, but there is also significant technological momentum pushing forward. Humanoid robots, once overlooked, are now in high demand, and with many companies and research institutions working together, it is likely that in a few years, we will frequently encounter humanoid robots capable of performing various tasks. Just imagine having a robot at home to help with chores or more flexible robots assisting in production at factories; life will undoubtedly become more convenient. All these developments are driven by market forces, technological advancements, and the dedication of many researchers. I believe it won’t be long before robots bring significant changes to our lives.