Beyond humanoid robots: A replacement of "humans as the labor force" is taking place.
While many people still view humanoid robots as "stage effects on the Spring Festival Gala," capital has already started to place bets based on industrial logic. On one hand, on the Spring Festival Gala stage, robots tumbled, jumped, and performed in coordination, allowing the outside world to first intuitively see the possibility of "acting like a human." On the other hand, the robot leasing platform Qingtianzu announced on March 18 that it had completed the angel round and angel + round of financing, with a cumulative financing amount reaching hundreds of millions of yuan. The leading investors were Dayang Motor, Muhua Kechuang, and Minzhuo Electromechanical, and follow - on investors included Lehua Entertainment, Fuzhuo Investment, Mingjia Capital, Ruizhi Venture Capital, Tianji Investment, Jiaxing Nantou, and Zhixing Investment. The old shareholder Dafeng Industry continued to increase its subscription. One event took place under the spotlight on the stage, and the other on the financing table. They seem to be two different things, but they actually point to the same change: the competition in the robot industry is shifting from "whether it can be made" to "how to sell it, use it, and spread it first."
This also means that what the market is really betting on is not just machines that can walk, somersault, and perform actions, but a kind of "machine labor" that can enter existing factories, shopping malls, warehouses, and service scenarios like humans, be repeatedly dispatched, continuously charged, and replicated on a large scale. More directly, what is valued in humanoid robots is not just the replacement of a few workstations, but the opportunity to become the next - generation general - purpose production tools in the manufacturing industry. The warming - up of robot leasing platforms indicates that capital has begun to compete in advance for the commercial entrance of this transformation. The rapid rise of platforms like Qingtianzu in a short period is not because the profit model of robot leasing has been verified by the market, but because the entire industry is looking for the same answer: when traditional automation is becoming increasingly difficult to adapt to the changing market, can robots, for the first time, be used as a general - purpose labor force like humans?
01
Why is the industrial community no longer satisfied with robotic arms
but starts to bet on "machine labor"?
For many years in the past, the industrial community has basically relied on two methods to improve efficiency. The first is to transform factories. That is, continuously add equipment, robotic arms, conveyor belts, and visual inspection systems to the production line, break down the actions that humans can do into standard steps, and then let machines repeat the execution. The second is to transform people. More directly, it is to make workers do more work per unit time through longer working hours, finer division of labor, stricter assessment, and tighter process management. In essence, previous rounds of industrial upgrades have revolved around these two paths, and they have indeed pushed the efficiency of the manufacturing industry to a very high level.
However, the problem is that it is becoming increasingly difficult to move forward on both paths. First, look at the path of factory transformation. Today's most advanced automated production lines are of course very efficient. In many factories, there are few humans in sight, and what you can see are rows of robotic arms and highly automated equipment. However, this kind of automation has a strong prerequisite, that is, the environment must be stable, standardized, and structured enough. Parts should be placed in fixed positions, pallets should be placed at fixed angles, the rhythm of the conveyor belt should be precisely consistent, and the movement trajectories of the robotic arms are all pre - programmed. It is good at repeating the same thing in a certain environment. Once the environment, process, or product changes, this system becomes less effective.
This is also the biggest problem with traditional automation. It is not inefficient, but not general - purpose enough. For example, a production line was originally designed around a certain fixed product. Where the robotic arm welds, what the grasping angle is, and how the assembly rhythm runs are all fixed. Once the market changes, such as a product modification, the addition of new components, or the adoption of a new process, trouble comes. Enterprises cannot simply change a few parameters. Instead, they often have to find an automation supplier for re - evaluation, redesign the positions, rearrange the workstations, and re - debug the equipment. As time passes and costs increase, the flexibility of the production line is dragged down. In the past, when market changes were not so fast, this logic could still work. But today, as product iterations become more frequent and demands become more fragmented, the shortcoming of traditional automation, "high efficiency but slow to modify," is becoming more and more obvious.
Next, look at the path of transforming people. It has also reached its upper limit. In the past, the manufacturing industry could rely on a large number of low - cost laborers, divide the process more finely, extend working hours, and implement stricter management to squeeze out production capacity. But now this path is becoming increasingly unfeasible. On one hand, labor costs are rising, and the acceptance of young people to work in factories is also decreasing. Many manufacturing scenarios are not facing the problem of high or low wages, but the problem of difficulty in recruiting and retaining workers. On the other hand, the protection of labor rights and interests is becoming more and more clear. There is less and less room to continue to exchange efficiency by infinitely squeezing labor costs and extending working hours. In the final analysis, the reality that enterprises face today is that it is neither cost - effective, nor stable, nor sustainable to continue to rely on the "human - wave tactic" to squeeze out efficiency.
This has pushed the manufacturing industry into a new contradiction: what is really lacking now is not a dedicated device with higher speed and greater power, but a general - purpose executor that can directly enter the existing environment like a human, adapt to different processes, and take over different positions when necessary. Because the real world is not designed for robotic arms, but for humans. The width of factory passages, the height of workstations, the size of tools, and the layout of operating consoles in factories, the elevators, doorknobs, stairs, and counters in shopping malls, and the shelves, turnover boxes, and barcode scanners in warehouses are all built according to the human body structure and movement logic. The more something is like a human, the easier it is to seamlessly integrate into this entire existing environment without having to completely rebuild the environment first.
This is why the industrial community is betting on humanoid robots today. It is not because the fact that they "look like humans" is cool in itself, but because they are most likely to be compatible with the physical world that humans have already built. They can work at workstations designed for humans, pick up tools designed for humans, and enter the production and service environments originally built around humans. In other words, the value of humanoid robots lies not in being anthropomorphic, but in being compatible. Robotic arms can of course work, and they are very efficient in fixed processes. However, their capabilities are largely locked by fixed bases, fixed trajectories, and fixed tasks. What is expected of humanoid robots is to move from one workstation to another, switch from one task to another, and migrate from one scenario to another without having to rebuild the environment every time.
Therefore, the reason why humanoid robots have suddenly become the direction pursued by the global manufacturing industry and the capital market at the same time is not core technology show - off, nor simply because they can walk, jump, and perform actions. It is because they have, for the first time, let the industrial community see a possibility: in the future, what may be replaced by machines is not just a certain fixed process, but the role of "humans as general - purpose labor force" itself. In the past, automation meant one machine replacing one action. What humanoid robots really want to do is for one machine to replace a type of work that a human can cover. The significance is completely different. The former solves the problem of local efficiency, while the latter aims at the most scarce and difficult - to - standardize part of the capabilities in the entire manufacturing system, the general - purpose labor force.
This is also the most fundamental logic behind this wave of humanoid robot fever. What the industrial community really wants is not a stronger robotic arm, but a machine labor force that can be dispatched, allocated, transferred, and reused like a human. Once this is established, when enterprises face market changes, they don't have to start from factory transformation every time. Instead, they can dispatch robots like they dispatch workers. They can go for assembly today, handling tomorrow, and quality inspection the day after tomorrow. The environment doesn't need to be greatly modified, the process doesn't need to be redone, and the equipment doesn't need to be rebuilt. In the final analysis, what humanoid robots really want to replace is not a certain process, but the role of "humans as general - purpose labor force." This is the fundamental reason why they are being bet on, and this is where the real significance of this competition lies.
02
Robots haven't really achieved commercial success yet.
Why do capital investors pursue leasing platforms first?
If the previous part answered why the industrial community needs a "general - purpose machine labor force," then the next question is: how should this machine labor force be sold, enter the scenarios, and start running. For today's humanoid robot industry, the technological imagination space has been greatly explored. What really blocks commercialization is not "whether there is a prospect," but "who is willing to pay the bill first." It is precisely at this point that what is first pursued by capital is not the large - scale sales of complete machines, but robot leasing platforms like Qingtianzu.
The reason is not complicated. The reason why leasing platforms are popular first is not because the commercialization of robots is mature. On the contrary, it is because it is still immature. Today's humanoid robots and various service robots are still in the early stage of the industry as a whole. The price of complete machines is not low, and the ability boundaries are still unstable. There are significant differences in actual usability among different brands and models. More importantly, many so - called implementation scenarios, although they seem lively now, when it comes to actual calculations, they still stay at the levels of "seemingly useful" such as display, reception, guidance, and event promotion. There are not many real - demand scenarios that can form stable repeat purchases and long - term payments. To put it simply, the supply side wants to sell, but the demand side still dares not buy directly.
This is a very typical early - stage state of the industry. Robot manufacturers of course hope to sell their equipment because only through sales can the investments in R & D, manufacturing, and channels be amortized. However, from the perspective of customers, especially potential users such as shopping malls, restaurants, retail stores, and brand events, the question is not "how cool the robots are," but "is it worth my while to spend hundreds of thousands or more to buy it back at once." After all, the capabilities of robots are still far from the level of "being able to work stably like an employee after being bought." Many scenarios have not yet verified a high enough input - output ratio. For most enterprises, it's okay to give it a try, but it's difficult to make a large - scale purchase.
At this time, the leasing model naturally emerges. In essence, it builds a bridge between the supply and demand sides in the immature stage of the industry. For customers, leasing breaks down the original large - scale one - time procurement cost into short - term service fees calculated on a daily, weekly, or project - by - project basis. Enterprises don't have to spend a large amount of money to buy equipment and bear the depreciation risk, nor do they have to include robots in long - term fixed - asset management from the beginning. Instead, they can first use them in short - term projects such as shopping mall activities, brand launches, and store promotions to test the effects, see the reactions of consumers, and determine whether it is worth further investment. In other words, the leasing model reduces not the price itself, but the trial - and - error threshold.
For manufacturers, this model also has practical significance. What the robot industry fears most now is not that the technology cannot be developed, but that the equipment cannot enter real scenarios after being made. No matter how stable the operation in the laboratory or how dazzling the actions at the press conference, without continuous scenario verification and user feedback, it is difficult for product capabilities to truly iterate. The role of the leasing platform is to shorten the path from "release to use" as much as possible, so that robots can start moving, running, and entering different types of commercial scenarios. Even if it is just light - weight tasks such as reception, guidance, interaction, and event execution at the beginning, it is more meaningful than leaving the equipment in the exhibition hall. Because only by truly entering the scenario can manufacturers know where the problems of the robots lie, what kind of capabilities customers are willing to pay for, and in which direction the subsequent products should be improved.
From the perspective of capital, the most attractive thing about platforms like Qingtianzu is actually not the short - term rental income itself, but the opportunity to occupy two very important entrances: a traffic entrance and a dispatching entrance. The so - called traffic entrance means who becomes the first stop for enterprises to contact robots. When merchants want to try robots, they may not necessarily go directly to specific complete - machine manufacturers, but may first turn to a platform that can provide a variety of equipment and service forms. The dispatching entrance is even more important. Once a platform has mastered enough equipment resources, customer demands, and scenario data, it is not just an "intermediary," but may gradually become a distribution center for robots to enter the real commercial world. Whoever occupies this entrance first will have a better chance of determining which equipment is recommended first, which scenarios are educated first, and which services form standards first.
The financing rhythm of Qingtianzu can actually illustrate this point very well. The company has completed three rounds of financing in less than three months, with a cumulative financing amount reaching hundreds of millions of yuan. Moreover, the structure of the investors is very diverse, including not only ordinary venture capital funds, but also manufacturing enterprises, industrial capital, and different types of financial investors. Such a combination itself shows that the market is not investing in a verified profit statement, but in a position. What everyone is competing for is not who has made how much money from robot leasing today, but who can first make the process of "how robots are used by enterprises" run smoothly and who can first occupy the forefront of future commercialization.
However, this area must not be over - hyped. Because the leasing model is not naturally established, and it is not guaranteed that whoever shouts it out first will be able to make it work. Although it seems light on the surface, it is actually heavy. A robot leasing platform is not a pure Internet platform. It has to bear a whole set of offline operation costs: equipment procurement or allocation, depreciation and amortization, warehousing, distribution, on - site deployment, debugging, training, maintenance, after - sales service, and even human - machine cooperation guarantee in different scenarios. None of these costs are light. Especially in the stage when the capabilities of robots are still unstable, and the equipment failure rate and maintenance requirements are still relatively high, the platform not only has to rent out the robots, but also ensure that they can work normally on - site. Otherwise, customers will not use them again.
What's more troublesome is that this business has very high requirements for order density and equipment utilization rate. If the platform has a lot of machines, but the orders are mainly concentrated in holidays, event seasons, or a few popular scenarios, and a large number of equipment are idle on ordinary days, the depreciation and capital occupation will quickly drag down the platform. On the contrary, if the customer scenarios are scattered, the project cycles are short, and the income per order is limited, while the distribution and maintenance costs are high, then a single business may not necessarily make money. In the final analysis, for a leasing platform to be established, the prerequisite is not that "the market thinks robots have a prospect," but that it can really increase the equipment turnover rate, reduce the performance cost, and develop repeat - purchase scenarios. This is far from as easy as the financing news makes it seem.
Therefore, the popularity of platforms like Qingtianzu is more like a product of the commercial exploration stage of the robot industry. The reason why they are pursued is not because the industry has found a mature answer, but because everyone has realized that technology alone is not enough. What is really important next is who can first put robots into real scenarios and who can first develop the usage path, service process, and customer awareness.
03
What is really being competed for in this race is not the robots themselves
but the organizational power of future labor force
If we look at humanoid robots and robot leasing platforms together, what is really being competed for in this race is no longer "who can build a more human - like robot first" or "who can get more financing first," but the organizational power of the labor force in the future manufacturing industry. Because once humanoid robots are successful, what they change is not just the replacement relationship of a few positions, but the entire way in which enterprises organize production, allocate labor force, and respond to market changes.
In the past, the core ability of the manufacturing industry was ultimately to organize a large number of skilled workers. How factories arranged shifts, how processes were split, how rhythms were set, how training was carried out, and how teams were managed were all essentially centered around the variable of "humans." Whoever could recruit more workers, train them more skillfully, and coordinate workers and assembly lines more closely would be more competitive. That's why in the past, the focus of competition in the manufacturing industry was often on factory expansion, production line construction, worker training, and process discipline. In the final analysis, it was an industrial system based on the supply of human labor.
However, if humanoid robots really start to enter the real production system, this logic will gradually change. In the future, the core ability of the manufacturing industry may no longer be just to organize a large number of skilled workers, but to organize a large number of machine labor forces. In the past, enterprises managed the number, attendance rate, and proficiency of workers. In the future, enterprises may be more concerned about how many robots there are, how many scenarios they can cover.