In the 19th century, Stevenson invented the world’s first steam locomotive, which overturned the steam locomotive.

Five years ago, Gao Yong founded blue core technology to independently develop robot deep vision, breaking the existing robot cognitive system.

On the path to innovation, there is more doubt than trust, and more booing than applause. Blue core technology believes in “Believe in the power”, believe in the future of the robot, believe that stick to the start-up will see depth vision in the robot industry brilliant day.

Blue chip technology recently announced the completion of a series B financing of over 100 million yuan. Although it has been established for more than five years, the founding team has been accumulating precipitation in the visual field for nearly 20 years, and now has the full-chain product Matrix and full-stack technology capability, has Been working with a number of head clients. Blue Chi venture capital is blue core technology series a financing lead party.

In this issue of the blue chip dialogue, we have invited the founder and CEO of Blue Chip Technology, Dr Gao Yong. We hope that blue chip’s experience in exploring the market can provide some enlightenment to industry practitioners, blue chip technology is expected to “Believe in the power”to bring more power to entrepreneurs.

1、Give the robot the eye

Blue Chi venture capital: There are a lot of start-ups in robot racing, what are the characteristics of blue chip, how does business evolve? Gao Yong: Blue Core Technology’s mission is to let the robot “Understand”the world and serve the world. We hope that through the mobile robot depth vision system, for the robot equipped with “Eyes.”. Today’s qr codes and magnetic strips correspond to “Discontinuous blind paths”and “Continuous blind paths”. 2d Lidar is a semi-blind state, and their perception is very weak. If in front is a suspension chair, if relies on the two-dimensional laser navigation, the robot or judges can sit up.

Blue Core from the beginning to do three-dimensional visual sensors, through 3D vision to solve the robot positioning, perception problems. Along the way, we ran into some technical difficulties. First of all, when we started making sensors, no one paid for them, and customers were afraid to try them. To B’s business, the customer To the security, the stability request is very high, does not attain 99.99% , the customer does not dare To try lightly. Our approach is to go from making sensors to making whole machines -- mobile robots.

After making the whole machine, there are other technical problems. Mobile robots, for example, work well in the lab, but when they’re used in a factory, they face problems that don’t solve customers’problems, because different industries have completely different factory environments. For example, some of the factory’s production management system has a high degree of intelligence or information, our robot into the factory after how with the factory’s original production system docking? So we kept expanding the hardware and software.

From the second half of 2020, we focus on three industries, one is 3C semiconductor industry, represented by Huawei and ZTE, the second is new energy industry, and the other is packaging industry. Blue Chi venture capital: install”Eye”and did not install”eye”what is the difference between the robot?

Gao Yong: If the robot is equipped with “Eyes”, the robot will be more powerful. Deep Vision can make robots more intelligent, and from the customer’s or the market’s point of view, blue core vision robots will be safer and more stable, able to do some of the work that a less perceptive robot can’t do.

At the same time, blue core technology, products, market play, can promote each other, so that more powerful robots. In other words, the more scenes we land, the better our robot’s vision system will become, and the wider and deeper a moat will form.

How does the “Eye”fit with the “Foot”and “Hand”of the robot? There are many technical challenges. We have full stack capability from robot perception algorithm, localization algorithm, obstacle avoidance algorithm, motion planning algorithm to robot scheduling management system.

Taking the Piggyback Transport Robot as an example, we combine the self-developed LX-MRDVS, robot control module and intelligent scheduling system, which can not only make the robot have autonomous navigation ability, but also adapt to complex environment, it can also make robots and other multi-robots coordinate with each other.

Blue Chi venture capital: What factors prompted blue core to focus on mobile robot depth vision system this technical route?

Gao Yong: in 2016, when we are making robot vision, the popular technologies on the market are Qr code, Lidar and so on. At the time, we weren’t 100% sure that the blue core’s path would be 100% successful. Now looking back, there is a kind of empirical intuition, but also based on professional judgment of the industry. We’ve done a lot of industrial work on robot vision in the past.

Most people in the robotics industry come from a background in mechanics and control. I come from a background in CV (computer vision) and perception. I have a natural empathy for robots. At that time I felt that the robot was still living in the dark, more painful, but also more lonely.

When you’re doing it one way and most people in the market are doing it the other way, you keep asking yourself, am I wrong? What if it’s wrong? Some people even told us directly: “I don’t believe that the domestic team can do this thing, or you are a little too naive, too naive...”this is the first difficulty we encountered.

We believe that robots must be smarter. Today’s robots are still in the early stages of application, and future robots should have visual systems like those in science fiction, like humans, and even more powerful than humans, so that they can work better for humans, to serve humanity.

The second difficulty is talent, early to break the professional constraints of the existing team, it is necessary to absorb other professionals. Lei Jun once said, “It takes more than three trips to find someone. It takes 30 trips to find someone. If you have my determination, I don’t believe you can’t make a decent team.”

I did run about 10 times to get a current core member on board, convincing him over the course of a year or two to see the blue core move forward. When you first came into contact with him, he didn’t believe it was a matter of how long you could do it. Not everyone will understand if you tell them once, and many will still wonder.

With a team, the process isn’t all plain sailing. In 2017, after our first sensors were built, we initially judged that the technology was OK and the team was OK. When the sensor team finally did the data transfer, I remember four or five of them were in the company lab, and they didn’t come home for three days and three nights. I was touched, too, but the results were devastating.

I thought after everything we’ve done, the numbers would be good. We go back to square one and reflect on whether we are wrong, or if there is something wrong with this direction. I know very well that if you go in the wrong direction, the more you invest, the further away you are from the future.

When I have a little vacillation, my own years in the visual field industrialization experience and intuition to give me confidence, support my choice of direction. There is no hope before there is persistence, but there is persistence before there is hope. In my experience, I think this direction is right, and it turns out that this direction is also right.

In 2017, I found the right team, and if there’s a problem, I’m going to see it through. The debugging process continued until the beginning of 2018, achieving one of our expected results. One client had been using a lidar system from another supplier, and there were about two to three robot crashes a month.

Most of the factories are manned, and it’s more complicated when there are people. Since the blue core’s visual obstacle avoidance system, the number of robot crashes has dropped to almost zero, which means that the customer’s production line can run smoothly and efficiently, and the customer’s input output can naturally meet expectations.

2、Jump out of a parachute first

Blue Chi venture capital: blue core products have been in Huawei, ZTE, Midea, Toshiba, Foxconn, COMAC, as well as photovoltaic, lithium batteries and other industries, the first customer batch application, in the process of market development, what are the challenges encountered?

Gao Yong: The market challenge is mainly divided into two stages. The first stage is how to create a benchmark customer. The second stage is how to evolve from a benchmark customer into a stable expectation and systematic sales growth, instead of relying on one or two big clients.

We didn’t take the easy route, the Qr Code, the LIDAR. With a background of our team, we can make a product in three months and sell it. But we down hard-hearted, must first make the visual technology, do not rush to do first product. Later, a large domestic 3C enterprises to cooperate with us, is because the market already has the technology can not meet their requirements, so there is the opportunity to blue core. Sometimes, choose a slightly difficult path, although a little lonely, but there will be a good opportunity to give you in the end.

Customer’s suspicion has many kinds, for example: If a small company finally did not have, after-sales maintenance to look for who? One of our early clients tested us for most of the year. The goods on the shelves of a customer’s factory may be worth millions of dollars. If the robot goes wrong and falls into the wrong position, we will have to pay the full amount, and I will have to personally guarantee it with the company.

I don’t think there’s a chance of that happening if we don’t open the door. If I don’t have confidence in my product, my customers won’t use it. The parachute that produces oneself oneself jumps first, parachute you descend oneself first, all quality problem can discover and solve.

In the first half of the year, when I got a call from this client, I was very nervous, afraid of problems, and now we’re working very well together, said Any Time, Any Place, a 24 year old business executive who works for the company. For startups, the strict requirements of a benchmarking client are a “Steep learning curve.”. Looking back, many of the industry’s leading customers on our strict requirements, so that we have accumulated a lot of experience in robot vision, for our future development to lay a solid foundation.

It was a difficult start, but it was a much smoother one. Many customers believe that the blue core products are trustworthy, they will take the initiative about the blue core every year together with the planning of automation system construction.

Blue Chi venture capital: What is Blue Core next key plan of development?

Gao Yong: at the technical level, we hope to enhance the robot’s semantic perception and self-learning ability. For example, a robot can walk through a factory by itself and recognize the environment of the workshop. It can identify production equipment, workers, walls, posts, boxes, shelves and so on. This is a hard-core challenge for us to overcome in the future.

In terms of market development, we will continue to cultivate 3C electronics, focus on developing new energy, packaging, railway and chemical industries, accelerate the commercialization of the core technology LX-MRDVS, and accelerate the construction of the enterprise’s upstream and downstream partner ecosystem, to complement each other’s advantages, win-win cooperation, and jointly explore the depth of vision technology in new industries, new areas of application.

Blue Chi Venture: Blue Chi venture choose not to be defined, dare to be the first of the entrepreneurs, how do you see “Not defined, dare to be the first”?

Gao Yong: In the 19th century, when Stevenson invented the world’s first steam locomotive machine and was first tested on the road, instead of flowers and applause, he was ridiculed and smashed by crowds, and no one believed that his “Monster”could be faster or better than a carriage.

If Stevenson had balked and given up, there would have been no story, and trains and high speed rail would have been delayed all over the world. Rather than admire Stevenson’s skill at the time, it was his superhuman vision and inner strength.

Most people believe what they see, and some people don’t believe what they see. At bluechip, we believe in the power of belief; we believe in seeing because we believe; we believe in our judgment about the future of the robotics industry; I believe that stick to the start-up will see the depth of vision in the robot industry brilliant day.