The growth rate exceeds 100%.
In the past year, when most of the capital returned to rationality from passion, the 3D visual track has not diminished, and this craze continues to spread.
On the one hand, financing is frequent, whether it is intelligent robot companies such as Vision Robot, Lingxi Robot, Mecamand, or AI vision companies such as Tuyang Technology, Deep Vision Intelligence, and Zhixiang Optoelectronics, all of which have disclosed large sums of money one after another this year. financing.
According to incomplete statistics, in July this year alone, there were 5 financings for 3D vision in the field of industrial robots, equivalent to the amount in one quarter of last year.
On the other hand, from the perspective of downstream application industries, 3D vision has been implemented in many industries such as automotive, semiconductor, thin film, lithium battery, panel, 3C, photovoltaic, logistics, PCB, food and beverage.
On the frontal battlefield, each family used their housekeeping skills to grab the market and push plans. In the rear battlefield, they dug rivers and built high walls in terms of protection level, stability, accuracy, scanning speed, data transmission, etc.
Behind this joint journey, Leifeng.com has summarized four representative forces: machine vision veterans, AI vision companies, robot tribes, and Internet technology giants.
It is not uncommon for multiple forces to scuffle on popular tracks, but what we care about is, why them? What is the weight of each party in the field of industrial manufacturing? Which side’s foothold is the core position of future competition?
Robot tribe with halo
Robot is one of the most popular application tracks of 3D vision.
If the robot track wants to maintain rapid development, it is inseparable from a pair of smarter 3D vision eyes; if 3D vision technology wants to prove its strength, robots are also an excellent carrier.
Most of the robot tribes are positioned as “3D vision + AI + robot”.
If 3D vision is equivalent to the eyes to obtain external information, then AI is equivalent to the brain, which can guide the eyes to recognize images more accurately, and combined with artificial intelligence, 3D vision can maximize its advantages.
One of the characteristics of robot tribes is that they are born with their own robot genes, participate in multiple links in the industrial chain, and are at the forefront of integrated solutions.
Having a brain and eyes is only the first step, and only when it is mounted on a suitable body can the advantages be truly activated.
In the process of 3D vision moving from the laboratory to the complex industrial scene, it is bound to go through the stage of finding the body. The long technical chain, easy to do and difficult to refine make this process difficult, which has also become a fortress of the current military.
Robot manufacturers choose to take root in robot hardware from the beginning, and have more advantages in the mutual empowerment of AI, 3D and robots.
It is easier for them to reduce R&D costs while ensuring product performance, and it is also more efficient in realizing product iterative upgrades to feed back technology.
In other words, robot manufacturers have established stronger comprehensive strength, and they have the confidence to face the competitive pressure of other manufacturers.
The second feature of the robot tribe is that it has assembled domestic high-end 3D talents to work hard with the strength of the entire enterprise.
The robotics track has passed the stage of technology productization and is in the stage of product commercialization. A large number of technology giants with resources have rushed in.
But the difference is that there is a big gap in the strategic attention that the two sides can give.
For big factories, robots and 3D vision are just one of the industry lines; but for robot tribes, they are the foundation of their existence, and they are bound to do their best.
Furthermore, the field of 3D robotics is a field that needs to be deeply cultivated.
Mecamand has formed a core team of nearly 500 people by setting up multiple subsidiaries and support teams at home and abroad.
Lingxi Robot’s strengths lie in hardware technology and 3D imaging technology. Its team members come from well-known universities such as Peking University, Zhejiang University, Tsinghua University, etc., and have more than ten years of research and development experience in robotics, computer vision and automation equipment.
Seebit Robot is the academic representative of the tribe, and its strength lies in software. The backbone of the core technology comes from universities and research institutions such as Princeton, Columbia, Wuhan University, and the Chinese Academy of Sciences. They have years of research and accumulation in the fields of computer vision, robotics, 3D graphics, cloud computing big data, etc., and the annual R&D investment accounts for more than 50%.
The third feature of the robot tribe is to accelerate the harvest of key scenes.
In the era of Industry 4.0, application scenarios are an important resource for machine vision manufacturers to develop 3D machine vision business.
However, there are many scenarios in which 3D vision + robots can be implemented. Different industries and different scenarios require different technical standards, which greatly tests the technical capabilities of machine vision manufacturers.
Lingxi lays out warehousing logistics, industrial manufacturing and other fields; Mecamand lays out automobiles, construction machinery and other industries; Seebit aims at three major industries of construction machinery, logistics and automobiles.
The intelligent robot startups represented by the above three companies have chosen to accelerate their landing in key areas.
On Mecamand’s official website, its 3D vision and the cases of large auto OEMs and leading energy companies have become benchmarks.
For intelligent manufacturing and intelligent logistics, which are two typical deep-water scenarios, Seebit has created a visual brain of “deep integration of 3D visual perception and robot flexible control”, and has overcome the variety of manufacturing scenarios with standardized and low-cost solutions. Difficulties brought by the small batch production mode.
AI visual rookie coming from the waves
If the head manufacturers of machine vision in the 2D era are compared to calm and sophisticated middle-aged people, then AI vision companies are young people in their prime.
Armed with emerging 3D vision technology in one hand, they are always ready to perform a terrifying drama.
The 2D vision technology that has been developed for many years can only get the plan view of the object. 3D vision is like the human eye, which can completely reproduce various three-dimensional scenes. In the process of industrial production, whether it is the length, width, height of the regular parts or irregular The surface, radian, and depth recognition of parts and components can give real-time, high-precision 3D data measurement results with just a “flickering glance”.
This means that from 2D to 3D, it is not a simple technical upgrade, but behind it is a “qualitative” leap in the quality and quantity of information that can be obtained. The latter has high underlying optical technology and algorithmic imaging technology barriers .
Therefore, if you want to engage in research and development in the field of 3D vision, technical strength is a hard indicator.
The development of the 3D vision market requires the organic combination of 3D visual perception technology and AI artificial intelligence technology.
On this basis, digital reconstruction of objects can make the virtual world more realistic, bringing large-scale implementation to fields such as AR/VR, virtual shopping, autonomous vehicles and advanced driver assistance systems.
Technology is the strength of such enterprises, adhering to the idea of ”small but refined”, to stand firm in the foothold of differentiation.
A strong opponent often has excellent comprehensive qualities, and an opponent who can take root in one place and specialize in a single point is equally awe-inspiring.
If a powerful enemy is coming, if you want to protect the city, you must know your own strengths and weaknesses.
Some of them focus on deeper and more detailed upstream core components, and some rely on self-developed software algorithms or outsourced cameras to build their own barriers.
Aobi Zhongguang is a leading company in the field of 3D visual perception. It conducts research and development of underlying technologies through the model of “chip + module + algorithm” to build a moat. At the same time, it covers the needs of small and medium-sized customers with standard products, identifies potential market segments, finds leading customers in segmented industries, and provides them with customized development services.
Guangjian Technology told Leifeng.com that its strengths lie in the development capabilities of deep learning algorithms for 3D cameras and overall software and hardware solutions for optical deep coupling. With the application of nano-optical technology in the field of 3D vision, Optics Technology has developed the world’s first mass-produced consumer-grade nanophotonic chip, forming a unique technical barrier in the field of 3D vision.
It is worth mentioning that among many 3D devices, the sales of 3D cameras/3D acquisition devices have the fastest growth rate. In 2021, the domestic machine vision market size will be about 16 billion, of which 3D cameras account for nearly 5%. It is estimated that by 2025 The proportion is expected to reach nearly 10%, so this has become the key layout direction of most manufacturers.
Willing to endure hardships, pick the hard bones that machine vision is hard to crack.
Lingming Photonics focuses on the research and development of 3D sensor chips, and applies cutting-edge single-photon detector technology to mobile phone 3D modules, LiDAR and other high-performance depth sensing systems.
At the beginning, Ewa Technology took the dual R&D route of “chip + algorithm”. Nearly 80% of the company are technicians. As a start-up company, the current R&D investment has exceeded 100 million.
Founded in 2020, VisionCore also entered the market as a 3D vision AI chip developer, delivering highly integrated chip solutions in the fields of machine vision and artificial intelligence to the market.
Zhang Dingjun, vice president of R&D of Aobi Zhongguang, also said that Aobi Zhongguang chose the “difficult” mode from the beginning, starting from the chip.
Guangjian Technology has successively cooperated with ZTE mobile phones and OPPO mobile phones, and launched a new under-screen 3D structured light technology based on the original patent of self-developed nanophotonic technology, breaking Apple’s patent and industrial chain monopoly.
From hardware to solutions, there are no shortcomings in the subdivision field.
Businesses around 3D vision can be roughly divided into two categories: 3D vision cameras, sensors, chips and other hardware products, or 3D vision system solutions.
The problem with the former is that it takes hardware products as its main business, which requires integration and cannot directly face customers.
The latter also has limitations. If you choose to do the overall software solution or integration, the core 3D vision technology accounts for a small proportion, and it is easy to be “stuck”.
Therefore, from hardware products to system solutions, they are all involved, focusing on solving a problem in each sub-scenario.
Through 3D technology, AI vision companies have entered different links in the industrial chain such as 3D cameras, 3D sensor chips and 3D vision platforms.
Several machine vision companies with relatively early start and relatively superior technology have become small giants in the field of artificial intelligence.
Deep-rooted machine vision veteran
History has used its screening process of survival of the fittest countless times to show us the cruelty and rewards of the business world.
Those companies that can cross the historical cycle law and stand proudly not only know how to enhance their strengths and avoid their weaknesses, but also know how to make haste in their areas of expertise. They also know that haste is not enough. Make up for shortcomings.
There is nothing more than a simple truth behind it: big waves wash away sand, and those who sink are gold.
KEYENCE and Cognex, which were born in the 1980s, are well-known veteran companies in the machine vision track.
Compared with foreign countries, the domestic machine vision industry started late. After 20 years of catching up, it has gone from a rutted road to a full-scale development, and finally competed with international giants and shared its reputation.
As a leading enterprise in the era of 2D vision, their sliding movements towards 3D vision are smooth and natural.
The key is to use the unique threshold of time to draw a word “early” between oneself and another.
Behind the word “zao” is the profound industry background that they show everywhere.
First, technological breakthroughs require the accumulation of time, money and talents.
In the past few decades, they have accumulated vision technology early, polished their core competitiveness in product performance and craftsmanship, and gained wide recognition for their mature 2D vision technology.
It is true that 2D and 3D belong to the same visual system, but the technical threshold has been raised a lot. It involves not only interdisciplinary design problems such as optics, structure, and heat dissipation, but also complex system design such as chip and algorithm composition.
They are strong in 2D vision, but they have also humbly invested in 3D vision for many years. Based on past experience, they are testing the gaps that are difficult for start-ups to easily overcome.
If you don’t have enough funds, ammunition and a solid technical team, you will venture into the blue ocean of 3D vision, and a big wave may cause a run. They are well aware of technology, capital and talents, which are the foundation for them to stand in front of the big wave of 3D vision.
Secondly, it has the advantages of a smooth supply chain and a stable customer base.
In the 2D era, they had a good run-in with the upstream and downstream of the industry chain, and they have stepped out of a road for themselves.
Plant trees in the front and enjoy the shade in the back. This time-drained “ally” relationship involves a wide range of factors. Behind the interweaving of strength, interests, trust and habits is the high cost of substitution. Therefore, once formed, it is difficult to shake easily, making it difficult for newcomers to catch up.
Now in the 3D era, they can better borrow the previous channel.
Furthermore, the brand halo and word-of-mouth advantage are its hard armor.
In the early days of the industry, they attracted a group of “hardcore fans” for their product versatility and high quality, as well as excellent customer service.
After customers form brand stickiness, they can occupy market share through stable shipments, and customer recognition is also conducive to the inclination of resources such as capital, talent, and supply chain, thus forming a virtuous circle.
Internet technology giants coming across borders
The era of milk and honey flowing in the Internet era is over, the opportunities for innovation go to the deep water area of the market, and the placement of digitalization goes deeper into the industry.
It is not uncommon for a large-scale Internet technology giant to attack left and right in the field of industrial subdivisions and meet narrowly.
3D vision is one of the tracks that has been “chased and blocked” by tech giants.
Under the wave of intelligent manufacturing, Internet giants want to carry out technological innovation, in addition to having the foresight and determination to enter the game, they also need the wisdom and endurance of deep water.
Frankly speaking, hard technology requires long-term research and accumulation, with high technical thresholds and clear application scenarios.
As a 3D vision head company, Obi Zhongguang has taken 9 years from its establishment to its listing.
The Internet technology giants often intrude intensively at the time point when the commercialization of products is approaching, and the technology landing and industry understanding are short boards.
Therefore, it has become a common choice for Internet companies to make good use of a large number of resources in hand and communicate with each other.
First, cooperate directly with upstream companies.
Alipay has cooperated with Obi Zhongguang, an innovative company that provides 3D face recognition modules.
WeChat Pay and Guangjian Technology have carried out in-depth cooperation, and Guangjian has also won the complete market share of WeChat face payment.
Xiaomi also chose to cooperate with Lu Shenshi on face recognition to customize the 3D face recognition module “Chongming”.
In addition to mobile phones and face-swiping payments, Internet technology giants have also gradually extended the application of 3D vision technology to many fields such as smart manufacturing, smart security, and smart hardware.
Second, borrow chickens to lay eggs and expand their 3D visual territory through foreign investment.
In 2020, Xiaomi invested in the 3D machine vision start-up Deep Shallow Vision. The latter’s main product is a micron-level intelligent depth camera, which has the ability of micron-level measurement accuracy, intelligent analysis, and restoration of the three-dimensional appearance of objects.
In 2020, Xiaomi invested in the machine vision start-up Sunshine Youshi, a 3D industrial camera developer with optical precision measurement as its core technology.
Fostering strengths and circumventing weaknesses and building an ecosystem with influence are the unique advantages of Internet companies.
The manufacturing industry chain is long, the construction cycle is slow, the production process is extremely complex, and the investment in technology research and development is high, which is a tough nut to crack.
In addition, technological transformation and production line upgrades are happening all the time, and behind every upgrade is a change in the supply chain, supply ecology, or even a reshuffle.
The strong brand influence of Internet technology giants helps to aggregate the ecology and attract the joint investment of upstream and downstream industries. In platform-level wars, it can easily kill opponents.
3D vision is still in its infancy, and the opening of the industrial chain and the large-scale implementation of products will take time to accumulate.
From the perspective of competition, these invisible rivals have successively assembled in the field of 3D vision and exchanged their capital and influence for the market.
From a higher perspective, the industry needs disruptors, and the ecological leadership of the giants will also help drive the development of the entire industry chain.
There is an endless stream of players, who can lead in the field of 3D vision?
As 2D vision is not enough to meet the ever-increasing precision requirements of precision manufacturing industries such as consumer electronics, smart cars, and semiconductors, the depth information of the physical space has been increased, and 3D vision that can record the physical world more comprehensively and realistically has appeared.
For the discretely manufactured auto parts industry, flexible manufacturing is an inevitable trend, and 3D vision just solves the key flexible positioning problem in the flexible production line of auto parts.
In 2021, the growth rate of my country’s machine vision market will exceed 45%, and the 3D vision market, one of the branches, will grow rapidly, exceeding 100%. It has become a hot favorite in the CV industry this year.
And this is also an excellent opportunity for Chinese companies to use technology and capital to overtake on curves.
As an emerging technology, the starting time of industrial-grade 3D vision technology at home and abroad is basically around 2014, and the starting line at home and abroad is not far behind.
The country has issued many favorable policies, and at the same time, the market demand for intelligent upgrading of industrial manufacturing is strong. Under the background of high capital attention and the heating up of the track, the fierce confrontation between new and old players has begun.
According to Zhu Li, CEO of Optical Vision Technology, there are two signs that 3D vision is truly mature: whether the hardware systems of sensors and 3D cameras are standardized in the scene, and secondly, whether 3D vision can cover 10% of the market demand.
The former means that the entire upstream component supply chain has entered maturity and standardization; the latter means that 3D vision has completed high-end scene coverage, and the market has entered a stage of rapid penetration and growth.
With the advent of the era of the Internet of Things, there are about billions of smart devices in the world with 3D visual perception requirements. This makes the first-generation products equipped with 3D vision technology have entered the market, and allowing machines to see the three-dimensional world is just the beginning.
This article is reprinted from: https://www.leiphone.com/category/aijuejinzhi/TLumiPEXBdn40Frj.html
This site is for inclusion only, and the copyright belongs to the original author.