The moat of a robotics company is a deep understanding of the industry it serves｜Taihe Opinion

Robots are infiltrating our lives at an increasing speed - delivering dishes in restaurants, providing services in buildings, moving goods in warehouses, and performing delicate operations in laboratories...

The heat in the capital market has also risen. According to the statistics of Taihe Capital, since 2021, the robot track has completed a total of 174 rounds of financing. Except for the undisclosed rounds of financing, there were 57 deals with a single amount of more than 100 million yuan, totaling more than 15 billion yuan. The robot customers that Taihe Capital participates in serve cover commercial services, warehousing, industry, medical care, life sciences and other segments, and accompany Purdue Technology, Jizhijia Geek+, Mecamand, Youai Zhihe and other leading companies The project has completed multiple rounds of financing. At present, nearly half of the robot head companies with a valuation exceeding 5 billion yuan are customers of Taihe Services.

In the process of serving top customers for a long time, we have witnessed the rise of a new wave of investment in robots, and also witnessed the complete process of many companies establishing core competitiveness and rapidly growing into top players in subdivided fields.

What this article hopes to discuss is what is the key to building a "moat" for robotics companies, and hopes to break some myths that "whenever you encounter a robotics company, you only ask about technical barriers". The following is the in-depth observation of Taihe team based on industry research and trading experience, and I will share it with you.

The core point of this article:

For robotics companies, "technical barriers" are not the core. As a typical mechanical engineering, the robotics industry does not have top-secret technical principles in most cases;
A truly solid moat comes from industry awareness, cost advantages and product matrix;
Mastering industry cognition is the most important thing, and profound industry cognition comes from the process of enterprises taking root in the scene and continuously accumulating data;
By seizing the time window and starting volume quickly, the company will have the opportunity to realize cost advantages;
Robots are a tool for production services. In order to resist the periodic risks of downstream customers, robot companies should choose a good track and continuously improve the product matrix;
The future head robot company should transform from an "equipment supplier" to a "solution and service provider".

"What are your company's technical barriers?"

If you want to select a list of "the most concerned issues for investors in the robot track", this is undoubtedly the number one issue. This question often appears in the actual roadshow scene of robotics companies. The founders often need to repeatedly explain to investors: what, where, and how high are the technical barriers of our company?

Technical barriers mean that the company has a technological advantage in a certain direction, and this advantage is difficult to be easily surpassed by peers. A company with strong technical barriers is more likely to occupy a leading position in the industry. Therefore, investors are particularly concerned about the company's technological leadership and use this as an investment basis.

But before answering this question, we might as well think about the question itself again: Is the technical barrier the most critical for robotics companies?

Is technology a moat for robotics companies?

Taihe believes that for robotics companies, technology itself cannot constitute a wide enough moat.

It is undeniable that technology is the foundation of every technology company. But robotics is not an industry that relies solely on high technology, but a very typical mechanical engineering, which has achieved continuous development for decades in the last century. In such an industry, there are very few top-secret technical principles or technical paths, that is, there are few principles or technologies that "Company A knows but Company B does not know".

If a company has unique technology and is strictly protected by patents, it will become an absolute monopoly in the industry. For example, Intuitive Surgical, a well-known American medical robot company, acquired its largest competitor, Computer Motion, in 2003, and has since become the absolute leader in the surgical robot segment, completely monopolizing the market with more than 4,000 patents.

However, according to our observations, most of the mature market segments in the robotics industry are either multi-giant monopoly or decentralized competition, and there is very little absolute monopoly like intuitive surgery. Take the familiar multi-joint robot market as an example: there are many players in the track, including foreign giants FANUC, abb, Yaskawa, etc., as well as domestic robot companies like Estun. Estun has experienced more than ten years of development. After development, he basically mastered the core technology, and then continued to win orders in the market.

This shows that the technical principle itself cannot constitute a solid barrier, and a temporary technological lead does not mean sustained commercial success, and latecomers can still get a share of the market.

Rather than saying that technology is the "moat" of robotics companies, it is more accurate to say that it is a "stepping stone". Take Fanuc, the world leader in the multi-joint robot industry, as an example:

Technology enables significantly leading product performance. In the load scenario below 20KG, FANUC's welding robot can achieve the industry's largest coverage and the highest repeat positioning accuracy, which is about 50% higher than domestic first-tier manufacturers in terms of performance and efficiency;

Technology can also help companies reduce production costs. FANUC started from the lowest technical numerical control system of industrial robots. It has comprehensive technology and can optimize product design from the bottom. Compared with its peers, FANUC has invested the most and most comprehensively in the self-development of core components, and therefore has a 5%-10% leading edge in gross profit. In contrast, many system-integrated robot companies in China generally have poor profitability and are subject to fluctuations in the price and supply of upstream components.

It can be said that with the stepping stone of technology, the company has the possibility to enter the ranks of the top. Technology can ensure the leadership of the company's products, and ensure the realization of large-scale production and R&D collaboration. At the same time, the continuous landing and iteration of products can also feed back technology.

What is a "true" moat?

So, since technology is not the most critical element to help robot companies occupy the leading position in the industry, what can become the real moat of robot companies?

After observing and contacting most robot companies in the market, Taihe believes that the most important moat is a deep understanding of the industry it serves.

As mentioned above, there are very few core technologies in the robotics industry that "we have what others don't have", so the company's own barriers come from the part of "what others have and what we have". In this part, the most important thing is the industry knowledge accumulated by the company, including scene understanding and data accumulation.

Scenes

The application of robots is in specific scenarios. They may appear in restaurants, office buildings, warehouses, hospitals, laboratories... Pay attention to avoiding people in restaurants, and mainly responsible for moving goods in warehouses. There are thousands of differences between the times. Only by deeply understanding the pain points, needs and standards of the scene can the technology be effectively implemented.

data

With the introduction of intelligent perception technology and AI algorithms, robot companies have begun to capture, analyze and understand massive data in the scene, and then train, control and optimize algorithms, which has gradually become one of the key factors that determine the performance and efficiency of robots. Therefore, it is particularly important for a new generation of robotics companies to go deep into industries and scenarios, and accumulate data and cognition.

Let us use two examples to specifically explain the value and difficulty of industry awareness:

For example, Mega Robotics, a leading life science automation company in China, provides automated laboratory products and services. This is a very special scenario: the robot has to deal with various types of biochemical reactions. In addition to ensuring accuracy and beats, it must also be able to withstand highly corrosive gases, UV radiation sterilization, and so on.

In order for the product to meet the above requirements, Mega needs to be tested repeatedly in the scene: for example, due to the different volatility, for experiments that also require "5ml samples", the extraction volume of different experimental liquids is not the same, and the robot needs to be tested according to different different liquids. Mega’s final solution is to use sensors to first detect liquid conditions, and then calculate automatic volatility based on previously accumulated industry data; another example is that due to the special laboratory environment, robot coating materials need to meet the specified cleanliness. It took two years to find a coating material that meets the requirements, ensuring the highest D-level standard required by GMP (Good manufacturing practice).

In addition to hardware equipment, industry awareness is also reflected in software and algorithms. To meet the special scene needs of the laboratory, MegaFluent has developed an automated experiment system, MegaFluent, whose algorithm can assist in the analysis of experimental data and the optimization of the experimental process, thereby improving the efficiency of the experiment. Therefore, based on the scene data, continuously iterating the system algorithm and improving the ease of use, these industry cognitions that have been repeatedly trained and tested in the environment have become a higher competitive barrier for Mega Robots.

Purdue Technologies is another interesting example. Purdue provides mobile delivery services in commercial scenarios, and the most important scenario is catering—this is the highest-dimensional and most difficult scenario in commercial use. Due to the dense population of restaurants during the peak dining period and the different moving lines and passages of each restaurant, the restaurant scene is extremely non-standard, and the entropy complexity of the environment reaches 11.5, which is much higher than that of the factory environment.
Image source: Internet

How to allocate transmission routes in a dynamic and complex environment? How to react correctly when encountering different obstacles? How can the supporting liquid not be spilled when it crosses the threshold? These are very detailed and realistic issues.
Image source: Internet

Purdue also went deep into the scene and did a lot of tests, using a lot of scene data to train the algorithm - the more data accumulated, the better the performance and effect of the algorithm. In order to prevent the liquid from spilling over the ridge, Purdue also pioneered the application of the suspension system on the car to its own products, achieving extremely high stability and being able to support liquid dishes smoothly over the ridge without spilling.

Deep industry cognition comes from the process of enterprises taking root in the scene and continuously accumulating data. Behind these seemingly minor adjustments is a huge improvement in customer experience. Only by fully understanding the scenario you choose can you build a real moat for robotics companies.

The Two Biggest Keys to Enhancing Competitive Advantage

The advantage of industry cognition is brought by experience, and experience is the product of time and quantity. Quantitative changes accumulate and will eventually bring about qualitative changes. On the basis of industry cognition, we believe that there are two other points, which are the trump cards for robot companies to remain invincible in market competition.

1. Seize the time window and establish a cost advantage

Whether a company can seize the time window, take advantage of the first-mover advantage, quickly seize the market and expand its scale is the key to establishing a cost advantage for an enterprise.

We have observed that at the initial stage of the emergence of a certain new technology, companies that have mastered the technology earlier and established industry awareness will be able to establish greater cost advantages, and when latecomers enter this track, they will more fearful. Taking the field of AMR (autonomous mobile robot) as an example, both Geek+ and Purdue have rapidly expanded their scale on the basis of establishing industry cognition advantages, becoming the fastest in their respective sub-sectors to achieve shipments exceeding 10,000 units The company has formed an upstream procurement advantage and reduced the BOM cost (Bill of Materials, that is, the cost of the final product is determined according to the cost of each self-made or purchased part in the BOM).

In this way, on the basis of "we have what others don't have, and we are superior when others have it", the company has further realized "we are cheap when others are superior", and can provide customers with high-quality services with more price competitiveness.

Of course, the reduction in BOM cost brought about by the scale effect of starting volume is only part of the cost advantage. On the basis of quantity, the company can do the integration and research and development of parts and components. This process can further realize versatility, standardization and modularization, and has stronger economic effects for enterprises. With scale, enterprises can continue to extend to upstream core components and enhance control over the supply chain. In this way, on the one hand, it can ensure the consistency of product quality, and on the other hand, it can also achieve better cost control.

2. Improve the product matrix and become an anti-cyclical enterprise

For customer companies, robots are tools used for production services, which are related to the capital expenditure of customers, so the robot industry has a certain periodicity. If the customer's industry is in a period of rapid development, then the shipments of the robot company will also greatly increase during this period; correspondingly, if the growth rate of the customer's industry slows down, the revenue of the robot company may also be affected.

In order to resist this cycle, head robot companies with universal underlying technologies should continue to develop different products, which will be suitable for different scenarios and have different functions. A company with a complete product matrix can rely on continuous growth to resist cycles.

For example, the Mecamand robot used 3D vision technology to quickly occupy a leading position in the field of visual positioning, and then cut into the field of visual inspection without stopping, opening up a broader market for the company; the same is true for the Luoshi robot, which established itself in the small six-axis market. After building word of mouth and brand, successfully developed a new generation of flexible force-controlled robots and realized mass production - new products can be used in general industrial, commercial and medical scenarios.

Rich products not only allow the company to have a larger market space, but also enhance the company's ability to resist downstream cycles.

Of course, the choice of track is also very important. We recommend that robotics companies choose downstream industries with large market scale, fast growth, and high cognitive requirements as cut-outs, such as the current commercial, lithium battery, semiconductor and other fields. Such a downstream industry can help robotics companies quickly increase volume and form cost advantages, while accumulating industry knowledge, iterating technology and expanding outward.

Where is the "second curve"?

It is conceivable that after the leading companies have established a moat, the industry competition pattern will gradually become clear, and the growth and profit margins of the entire industry will tend to stabilize. In the case of such a clear pattern, how the head company draws the second growth curve is the key to distinguishing "strong companies" from "great companies" in the future.

Taihe believes that transitioning from a pure "equipment supplier" to a "solution and service provider" will be an effective path for the transformation and upgrading of robotics companies. In this way, its business model will also change from a one-time charging model for equipment/projects to a recurring charging model for services, which will greatly improve customer retention and stickiness.

In fact, many companies have already started to move in this direction:

Geek+, through data accumulation of different types of e-commerce customer scenarios, has accumulated rich experience in warehouse operations, such as