The workshop is empty, the assembly line is running normally, and the AI-driven robots take over the entire production line. This is a miniature landscape under the blueprint of “Industry 4.0”, a reality, or a reality that is becoming.
Since China became the world’s largest AI industrial robot market in 2013, the domestic perception of the AI industrial robot market has changed from wait-and-see to consensus, and has been pushed to a climax in the past two years.
When the waves are rising, it is also the time when the bubble is at its greatest. Thus, a contradiction arises:
On the one hand, rigid demand has increased, the market has continued to expand, and new entrants have been pouring in. Products are in short supply, so that manufacturers have recently raised prices.
On the other hand, domestic AI industrial robot companies continue to lose money, the domestic market has not yet dominated, and the international market is difficult to break through.
The current domestic AI industrial robot track is in the middle of the contradiction between ice and fire.
A Song of Fire: Go with the flow, be ambitious
Productivity is the soul of manufacturing, and manufacturing is the foundation of national competition.
The demographic dividend has receded, and the demand for manufacturing has not diminished. The most critical task of industrial production is to reconstruct productivity. Safe, stable, flexible, and flexible robots have become the winners and losers in solving difficult problems.
The global AI industrial robot market has always been dominated by overseas giants, especially represented by the “Four Families” of Fanuc, ABB, Yaskawa, and KUKA.
China’s AI industrial robot industry started at least 20 years behind, but in the shameful and courageous forge ahead, from an admirer to a peer, and even swayed a son, turning the KUKA robot, one of the “Four Families”, into a own wholly owned subsidiary, which is the symbol of German Industry 4.0.
China’s AI industrial robot sales have ranked first in the world for nine consecutive years. Recently, a set of data released by the International Federation of Robots is more intuitive. In 2021, the installed volume of AI industrial robots in China will reach 243,300 units, a year-on-year increase of 44%, and half of the world’s AI industrial robots will be assembled.
China’s AI industrial robot industry is no longer the same. As of today, the scope of application has covered 60 major industries and 168 medium-level industries in the national economy, and the output has increased tenfold compared to 2015.
The number of robot industrial parks and robot companies has increased sharply. According to the data from Tianyan Check, there are more than 114,000 AI industrial robot-related companies in my country. Among them, there will be more than 46,000 newly registered companies in 2021, with a growth rate of 72.97%.
Capital has poured in. Last year, there were 130 financings in the field of robotics and intelligent manufacturing, of which 54 financings exceeded 100 million yuan. The total financing for the year exceeded 23.2 billion yuan, a year-on-year increase of 47%.
The optimism of capital has directly pushed the entire industry chain to a high valuation, and there are many “lucky people” who have received multiple rounds of financing soon after its establishment.
In terms of R&D investment, listed companies in the AI industrial robot industry chain have increased investment in R&D. According to incomplete statistics, in the first half of this year, nearly 60% of the 31 listed companies in the industry chain achieved a year-on-year increase in R&D expenditure.
In the field of AI industrial robots, foreign manufacturers are the forerunners, with deep roots in mainstream markets such as photovoltaics, lithium batteries, and electronics.
As a latecomer, local robotics companies are also gradually growing in research and development, and progress has been made in the research and development of core components such as precision reducers, intelligent controllers, and real-time operating systems. Some sub-sectors have come to the forefront of the industry, and in mainstream circuits such as photovoltaics and lithium batteries, they also dare to compete with international first- and second-tier brands in many bidding projects, such as Eston, Eft and so on.
In terms of AI industrial robot manufacturing, in addition to forming its own industrial robot “Four Domestic Little Dragons”: SIASUN, Eston, Eft, and Guangzhou CNC. There are also more and more AI industrial robot start-ups such as Luoshi Robot, Li Qun Automation, Jizhijia, Hairou Innovation, Jieka Robot, and Han’s Robot. They compete with each other in terms of software and hardware capabilities from ontology manufacturing to system integration. .
Under the high prosperity of the industry, a large number of start-up companies choose to subdivide products or applications, and some of the choices are large enough, and companies that are leading enough also stand out one after another.
The development process of China’s AI industrial robots is a high-melody “song of fire”: rising against the trend, savage growth, and ambition.
Sigh of Ice: Big but not strong, it is too early to replace domestic products
To climb the pyramid, you have to walk slowly up from the bottom, but there is no other way.
Just like the growth of a bud, there is always a part that escapes and slants out. The good news of the continuous financing of domestic AI industrial robots is also mixed with many contradictions.
On the one hand, manufacturing customers in labor-intensive industries, tool manufacturing, sewing, electronic product assembly and other sub-scenarios cannot find suitable products, solutions and suppliers; on the other hand, domestic AI industrial robot manufacturers are pushing new and old players. In the automobile and other industries, the scene of welding, handling, loading and unloading and other scenes are fiercely fighting in the Red Sea, struggling in the dilemma of high cost and low gross profit.
Another worrying phenomenon is that some “false fires” are still burning in the red-hot track.
The field of AI industrial robots is indeed growing rapidly, but the two phenomena of “low-end production of high-end industries” and “overcapacity of low-end products” are no longer new. They are still like two mountains, heavily pressing on the shoulders of domestic manufacturers. .
In the technically difficult high-end manufacturing field, foreign manufacturers still firmly control the right to speak.
Taking the three core components of AI industrial robots as an example, the cost of the controller, servo motor and reducer accounts for about 70% of the cost of the whole machine. However, its core technology and related patents are basically monopolized by the four major families, especially the control The four major families have a share of more than 80% in the Chinese controller market.
It can be said that the achievements of domestic AI industrial robot companies catching up are more reflected in speed and quantity rather than quality.
Looking at the domestic industry chain, the crowded places are the downstream integration companies with the lowest technical barriers. According to the sample data of 1092 integrators according to the statistics of MIR DATABANK, local integrators account for more than 95%.
Taking Dongguan, Guangdong as an example, there are more than 200 enterprises related to the robot industry in this region, but most of them either purchase foreign equipment for integration, or purchase foreign core components for assembly, and less than one-third have intellectual property rights.
In addition, domestic integrators also have the characteristics of small scale and scattered competition.
In addition to the enterprises with a revenue of over 300 million yuan in the field of automotive welding integration, there are less than 100 companies with a revenue of over 100 million yuan, and most of them are hovering below 30 million yuan.
As for upstream core components and body manufacturing companies with higher gold content, not only are there fewer companies, but the core barriers need to be strengthened.
Against the background of rising raw material costs, prolonged supply cycles, and rising labor and logistics costs, the industry’s supply and demand imbalance is also occurring: on the market side, non-standard reuse rates are low, resources are scattered, and enterprise usage costs are generally high; System integrators suffer from high R&D costs, difficulty in import substitution, and difficulty in mass production.
Different links in the industrial chain also have obvious differentiation: upstream core components are mixed, midstream body profits generally decline, and most downstream system integrators increase revenue but not profit.
Looking back on our development history, everything from small components to large industrial application scenarios is obtained through imitation, and not many people really master the core technology.
At present, although domestic manufacturers have made “point” breakthroughs in the research and development of precision reducers, servo motors, etc., the core components of robots, there is still a big gap in performance, and it will take some time to catch up.
Some industry experts have publicly stated that many companies would rather risk maintenance troubles to buy second-hand imported products than to use local brands.
If domestic manufacturers want to win the favor of local customers, they can only be reluctantly added to the company’s consideration list only if they are similar in performance to the products of the four major families and at the same time try to compress the price.
Domestic manufacturers are “spitting blood and eating away”, and continuous losses have become the norm.
At the same time, unreasonable local support policies have also induced problems such as duplication of construction and vicious competition.
The uneven product quality of local enterprises will directly weaken users’ confidence in autonomous robots. Therefore, although the slogan of “domestic replacement” is loudly shouted, most customers would rather risk the risk of difficult maintenance in the later period and choose foreign brands. .
Song Tao, general manager of Astro Robotics, believes, “The industry has been shouting slogans for many years, but it has officially entered the era of great competition after the epidemic. At present, the development strategies of most domestic robot brands are similar, that is to drive costs through scale advantages. Structural advantages, and at the same time, it is close to hand-to-hand combat at the competitive level.”
In the fields of automobiles and 3C, foreign giants have been rooted for a long time. In terms of product stability, process package adaptability, and customer relationship stability, it is difficult for new players to compete with them.
Enterprises that are trapped in the fierce price war and are unwilling to continue internal friction turn their attention from mature application scenarios such as welding, handling, and depalletizing, that is, from the advantageous sites of foreign manufacturers to weak subdivision scenarios. Do it through and achieve edge breakthrough.
This approach is indeed effective in some enterprises with strong technical skills, willing to endure hardships and sufficient funds.
For example, in the fields of sewing and flexible collaborative robots, domestic brands have been at the forefront of the industry, but similar breakthroughs are rare after all.
And we know that the focus is always on those bigger tracks. If China’s AI industrial robots want to truly rise, it is the only way to achieve domestic substitution on mainstream tracks such as automobile manufacturing and electronics.
How to change from quantity and speed type to quality and connotation type is the biggest problem in the current development of the local robot industry.
The phenomenon of low-level repeated construction and blind launch has attracted the attention of the Ministry of Industry and Information Technology, which is formulating industry access conditions and raising the entry threshold.
Behind the ice and fire
The market is chaotic, opportunities are everywhere, desires are spreading, and uncertainty is also increasing sharply. The joys, sorrows, expectations, and regrets of various shapes constitute a scene of fire and ice in the field of AI industrial robots.
In the long run, there is no doubt about the investment value in the field of AI industrial robots. It is understandable that large institutions and large capitals are the first to enter the market, and small institutions and small capitals follow the news.
Local overheating of supply and demand and capital is a normal phenomenon in a rapidly developing industry. From a good point of view, overheating also means more resources, which is conducive to the development of the industry.
But what needs to be vigilant is that when the bubble continues to expand beyond the rational value, it will bring harm to the AI industrial robot industry.
At present, the domestic robot demand is growing at a rate of over 40% every year, but why is China lacking first-class robot brands?
The most obvious difference between hot and cold in the field of AI industrial robots is that compared with the crowded market, the core technology circle is slightly deserted.
In all fairness, the core technology in the field of AI industrial robots is still in the hands of foreign manufacturers. Although domestic manufacturers have gradually entered the energy industry such as photovoltaics and lithium batteries, as well as mainstream markets such as the electronics industry, they are still in the process of cutting into the manufacturing process.
As a latecomer, it is a common business choice for domestic enterprises to cut into the market segment. The marginalization of the dominant market is “curvature to save the country”, but it also means that it requires patience to be “positive and rigid” in the mainstream market.
But not all companies can endure loneliness and have the self-consciousness of “sitting on the bench” for a long time.
The iteration cycle of AI industrial robots is very long. A mature product even takes several years of patient polishing. Coupled with the superposition of factors such as technological transformation, market demand changes, and time during the period, the uncertainty and risk are extremely high.
Some unlucky people, in the process of growth, are easily squeezed out of the market by companies that lack real materials and use the Internet to “subsidize money”, resulting in the phenomenon that bad money drives out good money.
Companies selling concepts have raised money, and companies with products and income are naturally less favored by capital. With the “joint effort” of all parties, the industry bubble is quickly pushed up.
The high valuation of the industry brings a “weird” dislocation, that is, high valuation often carries the invisible requirements of high growth, but the industry characteristics of AI industrial robots determine that it is not a high growth industry.
Therefore, when the company cannot reach the growth rate corresponding to its valuation, and the financing exceeds the rational threshold, it will lay the groundwork for the capital retreat in the future.
The efficiency of capital utilization is low, and the industry delays entering a stable period. Similar to the “progress” fueled by seedlings, it has been reflected in many industries.
The good news is that this year, many people in the industry have expressed the view that the AI industrial robot market is becoming more cautious.
Han Fengtao, CTO of Luoshi Robot, believes that the field of industrial robots has undergone two rounds of shuffling.
The first round of licensing took place in 2017-2018, and the first batch of companies established in 2014-2015 was cleaned up; the second round of licensing took place in the past two years.
From the perspective of financing, Han Fengtao believes that after several years of development in the industry, leading companies have basically reached the B round. As a capital, talent, and technology-intensive industry, investing in small companies at this time is extremely risky.
His views represent the current investment trend. The industrial robot market has gone through the early stage of investment, and capital has begun to favor companies with technical strength and a certain scale.
System integrators with poor resource integration capabilities, poor understanding of the industry, and disadvantaged capital scale, as well as companies with unreal technical capabilities and insufficient delivery capabilities, fell one after another in the test of the market. “The domestic robot companies that can basically be named now have begun to take shape and have accumulated a certain market influence in their respective fields.”
The emergence of vision, force perception, and application vertical companies supporting robots also confirms the rational trend of the industry. Because when the robot does not have the above capabilities, its vertical industry is impossible.
However, the cooling of the market is gradually transmitted to the primary market, and it will still take a year or two for the valuation to return to a reasonable value.
Institutions are no longer as aggressive as they used to be. For enterprises, the Matthew effect will intensify, and more resources will flow to leading enterprises, making it more difficult for enterprises at the waist and below.
As for the most pressing issue at the moment, the shortage of raw materials and parts supply caused by the epidemic and international factors has forced many factories to extend the delivery cycle by 1-3 times, while most of the new production capacity is still in the construction stage, and orders continue to be Under the growth, AI industrial robots have been in short supply in China.
“Machine substitution” is the realistic logic of the transformation of the manufacturing industry. Without robots, it is difficult for the manufacturing industry to progress further.
Capital is not familiar with industrial robots
The current boom in the field of AI industrial robots is mostly driven by capital.
However, the construction period of the AI industrial robot track is long, and it is not suitable for the short and fast investment model of the Internet. Adhering to the Internet thinking, it is impossible to invest in leading companies by investing money alone.
Google has spent many years in the field of robotics, and has experienced a process of first acquisition and then shrinking.
In 2013, Google started a series of high-profile acquisitions, bringing seven robotics startups under its umbrella in just half a year.
At that time, Google had a good hand: “Father of Android” Andy Rubin, humanoid robot expert Rosenberg, the famous Boston Dynamics, Japanese biped robot Schaft, Bot&Dolly, which developed industrial robotic arms, etc.
The media described with some exaggeration, “Maybe in a few years, you can see the scene of Google robots driving Google cars to deliver goods to users.”
But in the next five years, executives left, the project was dissolved and reorganized several times, and the ace companies sold themselves. Boston Dynamics, which was originally acquired for $3 billion, was sold to Softbank at a price of $165 million. is directly closed.
In 2021, Google will set up an independent subsidiary of Intrinsic, aiming at AI industrial robots and turning its breakthrough to the development of software tools that it is good at.
Google’s frustration with hardware exploration shows that the field of robotics is indeed a capital-intensive, talent-intensive, and technology-intensive industry, but money and resources do not necessarily lead to top companies. This industry has its own development. law.
The first threshold in the field of AI industrial robots, complex industrial know-how.
Previously, Li Yuhao, CEO of Maiyan Intelligence, once complained about Li Yuhao: an engineer, a number of listed company executives, and a robot “Sisyphus” who strives to “make mistakes” | Leifeng.com “The robot industry is too difficult x3, second only to Cars are the second hardest thing in the world.”
There are many sub-sectors in the manufacturing industry, and the scenarios and needs are relatively scattered. China has a complete range of industries. Each type of industrial enterprise has its own unique industrial know-how and industry knowledge. It is difficult to achieve migration between different industries.
If there is no deep understanding of the manufacturing industry, how can we talk about the upgrading of production lines?
The second threshold, complex customer needs.
The mature process of technical products often requires repeated iterations. It takes time to explore needs, improve industry awareness, and polish products. It takes time to go to the scene, work with customers, and when necessary, work with workers in front-line factories.
Luoshi Robot, founded in 2014, is one of the first companies to enter the field of AI industrial robots. Even so, its CTO Han Fengtao bluntly said that he has only slowly understood the needs of customers in the past year or two.
Even within a single track, customer needs are diverse. Taking knife manufacturing in sub-scenarios as an example, customers must not only sharpen one knife, but also several types of knives, which requires enterprises to update their knowledge in a timely manner and be equipped with the ability to automatically change production.
In addition to the larger and medium-sized markets, the main body of China’s industry is a large number of small-scale, multi-industry long-tail markets. Such enterprises have poor ability to undertake new technologies, and their needs are even more varied. If AI industrial robots want to seek in-depth development, they must serve these enterprises well.
The third threshold is the complex industrial chain.
From technology to business, the more new and advanced technology is, the longer the industrial chain will be.
Upstream includes core components such as reducer, servo system and controller; midstream body manufacturing involves key components, logic design, control algorithm, environmental configuration, etc.; downstream integrators need to carry out secondary development, automation supporting equipment integration, and provide solution.
The cooperation of existing related technologies, the matching of market demand, and the industry experience accumulated by paying tuition fees and stepping on pits are enough to eliminate some enterprises, and the companies that survive also need to rely on massive financing to survive.
The entire accumulation process often takes several years. The advantage of early entry will only become increasingly prominent in the field of AI industrial robots. Han Fengtao said frankly, “If a new company wants to enter the AI industrial robot market, it will be very difficult, because this industry has grown to a certain scale.”
Epilogue
The more chaotic the market, the more “quantity” is needed, and the more clear the market is, the more “quality” is needed.
It is true that the AI industrial robot market is still an immature market, and it is difficult to answer when its growth will pay off the huge investment.
On the bright side, the industry is saying goodbye to the rough expansion and explosion of the past, and is slowly returning to rationality and value creation.
Under this consensus, the growth of the field of AI industrial robots needs to endure loneliness. In the long run, companies with core technologies, identifying core values, and continuously expanding downstream market applications are expected to stand out.
As long as the hot and cold interweaving of the industry is controlled within rational limits, the overall advantages outweigh the disadvantages.
Leifeng.com has been paying attention to the field of intelligent manufacturing for a long time. If you have good projects, stories, opinions and corporate insiders, welcome to add the author WeChat Yukuaikuaier or MOON_ERS for breaking news and discussions.
This article is reprinted from: https://www.leiphone.com/category/robot/A6iZNAK733w01UZd.html
This site is for inclusion only, and the copyright belongs to the original author.