The alumni network of Xi'an Jiaotong University is probably buzzing with excitement. Today, Zhi Tao, an alumna of Xi'an Jiaotong University, officially sounded the gong for the listing of Yunji, the company she founded, on the Hong Kong stock exchange, with a market capitalization of approximately RMB 9.8 billion. At the age of 53 this year, Zhi Tao's success is a testament to the fact that it's never too late to start a business or take a company public.

Through the story of Yunji, let's delve into the current and future scarce opportunities in the robotics industry.

- 01 -

Anyone who has ever stayed in a hotel might be familiar with Yunji and its service robots. When you order takeout, instead of having to go downstairs to pick it up, a Yunji robot now brings it right to your door, taking on those repetitive tasks that you'd rather avoid.

Although the public's perception of robotics entrepreneurship began to crystallize during the 2025 Spring Festival Gala, the first wave of enthusiasm for robotics actually kicked off around 2014. Even the well-known robotics company Unitree Technology, which has gained internet fame, was established in 2016.

Many people tend to dismiss robots as mere gimmicks, believing they are only capable of performing entertainment tasks like dancing and running. They argue that robots struggle to meet genuine needs or convince people to part with their money for them.

However, this is far from the truth. The commercialization of robots has been underway for the past decade, and people have consistently been willing to pay for them. Of course, the robotics industry encompasses a wide range of categories, with some robots achieving success, others failing, and some still in the developmental stage.

For instance, service robots like those from Yunji have found success, especially in hotel settings.

According to Frost & Sullivan, in 2024, Yunji Technology claimed the top spot in the robotic intelligent agent market for hotel scenarios, boasting a 13.9% share of the domestic market. This surpassed the combined market share of the companies ranked second to fifth.

This raises a puzzling question: Why do robots thrive in hotel scenarios but not in others?

Which scenarios are less promising? Take front-desk robots as an example.

Between 2014 and 2015, a group of entrepreneurs placed their bets on front-desk (guidance) robots rather than hotel robots. However, to this day, no particularly large enterprise has emerged in this particular sector.

So, which scenarios hold promise?

Firstly, scenarios related to 'delivery' tend to achieve commercial success.

Yunji's products are essentially hotel delivery robots. What do they deliver? Takeout, guest supplies (such as slippers and toothbrushes), emergency medications, and so on.

Although Yunji has not yet turned a profit, its scalable revenue clearly demonstrates the genuine demand for its services.

From 2022 to 2024, Yunji's revenue was RMB 161 million, RMB 145 million, and RMB 245 million, respectively, totaling RMB 550 million. During the same period, the company's net losses were RMB 365 million, RMB 265 million, and RMB 185 million, respectively.

Another type of delivery robot that has achieved commercial success is the industrial delivery (heavy-duty) robot.

There are numerous examples, such as SEER Robotics, Mech-Mind Robotics, and Stan Robotics, all of which are unicorn companies. These robots navigate through various factory floors, freeing workers from strenuous, dirty, and tedious tasks.

Secondly, apart from delivery, another category of robots that has achieved commercial success is floor-cleaning robots.

Notable players in this field include Dreame, Ecovacs, and Narwal Robotics.

In the second quarter of 2025, the top five global floor-cleaning robot manufacturers were all Chinese companies: Roborock, Ecovacs, Dreame, Xiaomi, and Narwal Robotics. Together, these five companies accounted for nearly 70% of the global market share.

Another area is educational robots. In fiscal year 2024, UBTECH's educational intelligent robots and solutions generated approximately RMB 363 million in revenue, accounting for 27.8% of its total revenue.

- 02 -

We can observe that in the Robots 1.0 era, the core characteristic of commercialization was assisting humans in performing manual labor. The more a job is disliked by humans, the greater the potential for robots to excel.

Take steel mill robots as an example. In steel mills, there is a fixed job role - bottom-blowing furnace cleaning workers. Bottom-blowing furnaces are steel-making equipment that require regular cleaning (to remove slag from the furnace mouth). Having humans perform this task is truly arduous.

Standing next to a bottom-blowing furnace can potentially cause heatstroke due to temperatures exceeding 50 degrees Celsius. Even with heat-resistant workwear and protective gloves, workers still experience extreme discomfort.

However, it is precisely in these types of job roles that robots have the greatest potential, especially for dangerous, uncomfortable, or beyond-human-limit tasks. Scenarios like these are particularly suitable for the Robots 1.0 era (2015 - 2025).

Listening to this now, it may seem straightforward. However, when standing in the midst of a major technological shift and facing immense uncertainty about the future, things are not so clear-cut.

Companies like Yunji, which have unwaveringly stuck to the hotel scenario for a decade, require tremendous determination. First, you must have faith; second, you must persevere. Otherwise, you may start to wonder: Could switching to front-desk guidance robots be an option?

Well, the Robots 1.0 era is now a thing of the past, as we have seen giants emerge in all the scenarios we just mentioned: either these giants have gone public or become unicorns.

Moreover, with today's rapidly advancing technology, the infrastructure is no longer at the level it was a decade ago. Given the strides in AI technology, using today's technology to do what was done a decade ago (such as delivery and floor cleaning) seems somewhat wasteful.

Previously, Pencil News interviewed Yu Guangdong, the founding partner of Boiling Point Capital (Feidian Capital). He mentioned that if a team is still working on 1.0-era technologies today, regardless of how advantageous their conditions are, he would not invest.

So, what are the worthwhile investment opportunities in the Robots 2.0 entrepreneurship era over the next decade? Yu Guangdong said he is interested in robots with AI brains.

What does it mean to have an AI brain? In the 1.0 era, we saw many robots performing tasks like painting walls, cleaning windows, scrubbing boats, cleaning pools, and weeding. If these robots do not possess 'human-like intelligence,' they can essentially be defined as 'brainless'.

In the 2.0 era, the combination of 'AI Brain + Robot Execution' may present significant opportunities. Whether these opportunities will continue to be in the delivery and floor-cleaning scenarios of the 1.0 era is difficult to determine.

What we can determine now is that 'AI brains' can already empower 1.0-era scenarios.

For example, Yunji's products have already incorporated partial 'AI brains,' enabling intelligent recognition of guests' voice, text, and image requests, dynamic coordination of robots, employees, and equipment, and automatic task assignment and full-link tracking.

Previously, Pencil News interviewed Dreame Technology. On the one hand, large models have been integrated into their floor-cleaning robots; additionally, they have explored extended functions similar to those of a 'cleaning aunt,' such as folding clothes.

Perhaps we can attempt to summarize the new opportunities in Robots 2.0 based on the exploration of Unitree Technology.

According to Reportify data, Unitree Technology's current core product is quadruped robots, accounting for approximately 65% of the company's revenue. Eighty percent of these quadruped robots are used in research, education, and consumer sectors, while 20% are used in industrial sectors, such as inspection and firefighting.

From these scenarios, it can be seen that the difference in job roles between 2.0 and 1.0 robots is not yet significant.

The reason behind this is that today's robot AI brains are far from advanced. Even with the rapid progress in AI large models over the past three years, AI large models were not invented specifically for robots. This 'adaptation' process will take five years or even longer.

This prolonged adaptation process also implies more infinite possibilities.

This article does not constitute any investment advice.