Reporter | Chen Chen
Editor | Jing Yu
In 2016, Shao Tianlan and a group of returnees from Tsinghua University established Mecamand, focusing on the field of industrial robots. At that time, in the industrial robot industry, the situation of the four major families was already very stable, and few entrepreneurs dared to say that they wanted to take a share in such an industry.
Since the first industrial robot entered the automobile production line in the last century, the process of industrial automation has officially started. But until now, we still see that industrially developed countries like Germany and Japan are transferring their industrial chains to East Asia and South Asia in search of cheaper manpower.
Shao Tianlan told us that from his experience of visiting hundreds of factories, there are still a large number of processes in the manufacturing industry that have not been automated.
“We are Party B among Party B,” Shao Tianlan said. In his opinion, there are many and complicated processes in the industry, and the needs for automation are also different. , Intelligentization has become exclusive to large industries and large enterprises.
As for how to get out of the dilemma of “industrial customization”, Mecamand’s idea is that complex problems are left to software to solve. They make the technology of AI+3D vision the brain and eyes of the robot, combine standardized hardware, and complete complex tasks such as feeding, testing, screwing, cutting, and welding.
Shao Tianlan said that automation is far away from the technology circle and investors, but in the past two years, it is obvious that everyone has begun to pay attention to new innovations in the old industry of industrial automation. In June 2022, Mecamand completed the C+ round of financing.
Sales of industrial robots have increased tenfold over the past decade. Industrial automation is becoming a technological dividend that small industries and small businesses can afford. It is a vast market with potential, and it is also a place waiting for innovation and talent.
The following is the transcript of the speech of Shao Tianlan, the founder of Mecamand, organized by Geek Park:
I came back from Germany 6 years ago and started Mecamand. Mech-Mind is very similar to an autopilot company in the industrial field, except that what we drive is not a car, but a variety of equipment. We are committed to making industrial robots more intelligent through the technology of sensing, perception and planning.
In the direction of AI + industrial robots, we are one of the companies with the most landing cases and the largest financing amount in the world. We have AI, robots, sensing, and perception technologies, but the track we are in is automation, or industrial automation.
Friends here today, is there anyone from the manufacturing industry or industrial automation? very few. Let me first share a number with you, which is also a number that I think is very magical, that is 0.
It’s a great honor to participate in this event. I did my homework before I came, researching so many speakers at the Geek Park Innovation Conference in the past 5 years, to see who’s main business is industrial automation, and the data I got is 0. Today it becomes 1, because I work in industrial automation.
Shao Tianlan, the founder of Mecamand, gave a keynote speech at the Geek Park Innovation Conference IF 2023
In the past few years, there have been many excellent corporate guests, including the Internet, consumer electronics, automobiles, cloud computing, etc., but industrial automation has always been a little far away from the technology circle.
As a practitioner of automation, why should I come here to talk about it? Because it has a huge market, such as Siemens and other top 500 giants with a level of 100 billion (US dollars), this is a huge industry. Of course, there are other huge industries, such as hot pot, but there is not much innovation in these industries. The industrial automation industry is experiencing a lot of innovations, and innovations are rapidly changing this industry that sounds less trendy and less fashionable.
I use a number to represent the rapid change brought about by this innovation, which is 10 times. The sales of industrial robots have increased by 10 times in the past ten years. The Internet may feel that 10 times in 10 years is nothing, but it is great for automation. Today I would like to share with you some of our observations.
The ‘False’ Industrial Automation Boom
The automation circle is far away from the technology circle and investors. In the past 6 years, I have often heard some questions. For example, automobile factories like Volkswagen and Toyota have hundreds of thousands of people at every turn. The same is true for home appliance companies such as Gree and Haier. It can also be seen from the same data that there are still many people working in the manufacturing industry.
Some people also asked, in countries such as Germany and Japan, has such a simple job been automated? If this is the case, their factories will not be transferred to East and Southeast Asia. There is also a common saying that the current factories are very advanced. For this speech, I searched for some materials and searched the factory. Many of the videos in it are actually cut out of parts with a relatively high degree of automation. I have been to hundreds of factories myself, and these places are not automated.
You can usually see a lot of videos like this: like a fully automated potato chip factory, potatoes come in, and then the potato chips are packaged and sent out. These videos will give many people outside the circle the impression that manufacturing is highly automated. However, this kind of video is only a very small part of the manufacturing industry.
What does the real mainstream manufacturing industry look like? Let me give some examples. Rows of such tens of thousands of workers are manually assembling small appliances such as coffee machines, and they need to complete a lot of processes. This is a foreign garment factory, and there are a lot of workers in it, doing sewing and other work by hand. These are typical situations in the manufacturing industry, where there are still a large number of links that are not automated.
What kind of manufacturing industry, or what kind of link, is its degree of automation relatively high?
The most typical ones are stamping, welding, painting and other links in automobile OEMs. These are highly automated, and a large number of robotic equipment are also used. There are also chips, such as fabs or the semiconductor industry, which have a relatively high degree of automation.
Standardized foods, such as potato chips, cola, various cans, etc., will also have a high degree of automation. In addition, since the industrial revolution, the degree of automation in the textile industry is also quite high. There are also daily chemicals such as shampoo and cosmetics with relatively large volumes, including various medicines.
In addition to those mentioned above, the degree of automation in almost other industries and processes is relatively much lower.
In industries with relatively high automation, if we look for some rules, we will find:
First, large enterprises dominate. Such as pharmaceutical factories. The output value of pharmaceutical factories is very exaggerated, and the same is true for automobile factories, which can easily reach hundreds of billions.
Second, the production scale is large. Whether it is producing Coke, potato chips, or automobiles, its production scale is very large, and it is produced continuously throughout the year.
Third, the product iteration cycle is long. For example, Coke, except that the outer packaging is changed every year and holidays, most of the things inside change very slowly. There are also cars. The current car changes may be faster, but the production time of the previous car was relatively long.
Fourth, the process is stable. For example, beer, the craftsmanship of beer may not have changed much for decades.
Fifth, raw materials and production environment are controllable.
Make a standardized solution and let the complicated things be transferred to the software
Industrial automation is a very difficult industry, and we consider ourselves to be the second party of the second party. The demands for industrial automation are numerous and complex. There are many processes in thousands of industries, and the automation required for each process is different.
Previous automation combined with hardware such as sensors and motors has the characteristics of high degree of non-standardization, large environmental modification, and low flexibility. Low flexibility means that if the customer produces this product today, your automation solution is OK, but if his product demand changes tomorrow, the automation solution will no longer be suitable.
We hope that no matter what the needs are, the solution can be highly standardized and highly intelligent. It is best that this intelligence is the same as that of humans. It does not need to be designed for the environment. What humans did in the past, robots will still do in the past. At the same time, it is easy to use and easy to deploy. It is best if the customer needs automation, send a robot over, and let it do it directly in the afternoon.
Shao Tianlan, the founder of Mecamand, gave a keynote speech at the Geek Park Innovation Conference IF 2023
This matter has been talked about for a long time, but it can’t be done because it is a great technical challenge.
Humans are much stronger than robots in several points, on the one hand is the actuator. I am not a very strong person, but it is easy for me to lift twenty or thirty kilograms, but this kind of energy density is unimaginable for the current robot motors, which is an order of magnitude worse. For example, my hand is not very strong, but it has more than 20 joints and a lot of muscles. This level of integration and flexibility is beyond the reach of today’s robots. Moreover, our brain is still a very good control system. Such complex control capabilities are also difficult for robots to achieve today. Even if you want to break through, the cost will be very high.
But we want automation to be better, so what can we do? As much complexity as possible can be moved from hardware to software. Although we cannot replace humans like humanoid robots, we can still make the industry more standard, smarter, more agile, and more flexible.
Industrial automation can also be “inclusive”
Let me give some examples of our product deployment in the past few years, such as the most common loading link in the manufacturing industry.
There are many machines in the manufacturing industry, such as machine tools, thermal processing, and various testing equipment. Each of these machines has a demand for feeding materials, especially in discrete manufacturing factories. The situation in many factories is that workers stand in front of the machine, and in front of a box are semi-finished products from the previous link, and they need to put these semi-finished products in a specific place. But we can also solve this task by using AI+3D vision technology and using standardized hardware.
The robotic arm is this kind of standardized hardware, which will be shipped in batches in the factory. We did not make a specific mechanical design for this workpiece, but let the robot know how to grab an object and put it in the frame through vision + AI.
In our laboratory, we randomly placed a lot of items that can be seen in supermarkets in a box, with a variety of sizes and materials. The task of the robot is to place these objects one by one like a supermarket cashier. take it out.
In addition to feeding, there is also gluing. Thousands of different types of windows and panels need to be glued. We use standardized robots and standardized vision systems to allow robots to identify and position different windows and panels, and then apply glue. The robot is a completely standardized product, and so is the vision. Its conveyor belt does not have any special mechanical design for the workpiece, so its complexity, that is, the trajectory of the robot’s movement, is completely resolved on the software. Another example is tires. Now put a standardized robot and install a standardized system to track the movement of the car.
Today’s robots can solve highly complex and non-standard automation needs through advanced vision, artificial intelligence technology, and very standardized hardware, that is, robotic arms.
Here is a video introducing our products. Through optical design, we can allow the robot to perform very fine 3D scanning, and then use highly standardized hardware and software to complete many links, such as feeding, testing, screwing, and cutting. , Welding etc.
In the past, industrial automation was more exclusive to a small number of large industries and large companies such as automobiles and chips. However, through intelligent technology, we can enable more industries and many small and medium-sized enterprises to enjoy automation. Technology.
The above is my speech today, thank you!
This article is transferred from: https://www.geekpark.net/news/312743
This site is only for collection, and the copyright belongs to the original author.