For several years, Alexander Khanin was the head of the image processing department at a large Russian research institute, and at the same time, together with friends in the dorm room, he created his own object recognition system. Khanin believes that an effective tracking robot should be seen as a person, that is, understand what he is watching. His company, VisionLabs, creates programs that are embedded in video surveillance systems and devices that minimize the likelihood of error.
With the help of Hanin’s technology, you can track the actions of criminals, control the movement of vehicles, observe the work in another part of the world. Initially, he was going to offer his product to ordinary users, but soon realized that it was better to work in the B2B sector. This decision allowed to sell the company's products in the amount of $ 20,000 over the past two months.
Field of activity
Research activities in the field of computer vision, data analysis and robotics
Creator of the VisionLabs Project
We are engaged in video analytics: we make software and equipment that can see the same thing as a person. We recognize the numbers of cars, people's faces, their actions, track movements.
The idea to create such a project visited me when I was still a third-year student at MSTU. Bauman. I studied at the Department of Robotics and wrote a term paper. I built a robot such as "Valley", which could perform various actions: for example, to reach the stairs, go to the door and more. I wrote an algorithm, loaded it into a robot, and simply drove these commands into the program. And how he will fulfill it, which path he will choose, “Valley” has already decided for himself. Over time, there were like-minded people with whom we eventually formed an excellent team and named it VisionLabs.
At first, the project was just a hobby for us. At that time I lived in a dormitory, in the evenings there we organized our first round tables and brainstorming sessions. Then I worked as the head of the department in one large research institute, was engaged in image processing and went to many exhibitions, seminars, conferences. Learned the wishes of potential customers. We knew the security market, we knew the robotics market, saw what was missing, saw new markets and new opportunities.
In order to continue work and reach a new level, we organized the Center for Technical Vision in the trade union committee of the university, but it did not receive much development, and we decided to write an application in Skolkovo. To do this, we found guys who were well versed in marketing and finance, walked around the market, set priorities. After a couple of months, we received approval. And in the spring of 2012, we became residents of the innovation center and began working as a startup.
In March 2012, I quit my main job at a research institute and took up only the VisionLab project. I think this can be considered the birthday of our company.
It is clear that the sooner we create a robot that can see and perceive the world as a person, the sooner we will come to a brighter future. This requires artificial intelligence. Without it, the development of both robotics and technology as a whole is impossible. It is the lack of development of intelligence today is the main obstacle to the development of technology. It is important for the car not only to see, but also to understand what is happening. Otherwise, it will remain a piece of iron, and the maximum we can invent is a robot vacuum cleaner, suitable only for washing floors.
The very first we created an algorithm for reading labels. Thanks to him, it became possible to determine the number of cars, read passports with minimal error. It began to be created even before the formation of the company. With the advent of VisionLabs, we started writing an algorithm for recognizing faces that is able to detect them, distinguish them from each other, track the same face, and, accordingly, a person. The robot sees the face, remembers its main characteristics - the distance between the eyes, the size of the mouth, and so on - and subsequently recognizes it on other records. The last one was the motion recognition algorithm. Each of the programs was written for about six months. All of these products are either already sold or are in the pilot testing phase.
As we work in the IT sector, our only equipment is computers. We spent about 100,000 rubles on them. First, we rented a small office in the Skolkovo technology park. Now we have expanded and moved to Strogino. We passed all the examinations of the Ministry of Economic Development and the Government of Moscow and got an office on favorable terms: 5,000 rubles a month for 40 m2 with furniture. At Skolkovo, we paid about 15,000 rubles for Raskolnikov’s closet.
We invested our own funds in the project. We have several shareholders: I (the main one), my colleague at the university and two partners from Skolkovo. From the launch to the present, we have spent about $ 100,000. Each of us has invested and continues to invest our personal funds.
For the first year of operation, our revenue was generally ridiculous, we earned only a few thousand rubles. Now, over the past two months, we have sold our products in the amount of approximately $ 20,000. The number of our partners is constantly growing. I think that by next year the revenue indicator will grow significantly.
Search for grants and partners
We wanted to get to Skolkovo in order for the project to undergo a detailed examination. In the process of filling out a large application, an understanding of the project was structured in our head. In addition, Skolkovo is a large community of people with burning eyes, where you can find like-minded people and partners. Resident status gives us the right to be in the ecosystem of the fund, have access to experts, it is easier to promote the project in the media, to look for customers and partners. From material wealth we get preferential taxation and the opportunity to apply for grant financing for the project. By the way, investors and Skolkovo are not mutually exclusive concepts. In most cases, to get a grant you need to find a co-investor.
We wrote the application for about four months, because then the submission process has not yet been established. The form for filling is constantly changing. This is a fairly lengthy document, and not everyone has the patience to fill it out. The application is written in two languages, Russian and English, 30-40 pages in each language. Plus, a bunch of different links are affixed and a lot of additional materials are indicated. The application is reviewed by 10 experts, as a rule, half of whom are Russian and half are foreign. They study everything, write comments, evaluate the feasibility of the project, the state of the business idea, financial indicators according to forecasts, and so on. In our case, everything turned out well: the experts felt that we met all the requirements and gave us the green light.
Whatever the cool, breakthrough, world-changing technology, without the right implementation, it does not cost a dime. We are constantly looking for the right partners for our technology.
Market and customers
When we first launched the project, we thought that we would sell our products to end customers. After working with them for about three months, we realized that it would be very difficult for us to educate them, so we decided to switch to a B2B model. Today, our customers are large development companies in which the sales structure has already been debugged. These are companies developing security systems, video analysis systems, and vision systems. We do what we can do well: write algorithms. We are like Intel Inside: no matter which laptop you choose: Sony, Toshiba, Apple - Intel processors everywhere. And without them, nothing will work. Manufacturers compete among themselves in price, marketing, and service. But it doesn’t matter which of them wins, they will still take intelligence from us, because we make the core.
We are not the only ones in Russia, but we have one of the fastest algorithms in the world, and we enable developers to integrate our algorithms into their cameras and video management systems. This is a unique opportunity, and only a few companies abroad have it in the world. There is a fairly tight binding to the platforms and requirements for implementation, but our requirements are very simple and loyal.
The sphere in which we work is not very clear for a simple user. And people are very cautious about what they poorly understand. Therefore, our main difficulty was to choose the scope of our technology, that is, the audience for it. Whatever the cool, breakthrough, world-changing technology, without the right implementation, it does not cost a dime. Therefore, we are constantly looking for the right partners for our technologies. Our plus is that we have more than one product, but a whole portfolio, respectively, more likely to find a client suitable for us.
Our second main concern was the pursuit of excellence. We tried to create an ideal, but then we realized that it was not always needed. It was hard to find a moment, say to myself: “Stop, stop redoing everything endlessly.” It was difficult, as it seemed to us, to produce raw developments on the market, where they should have already been tested. It’s important to understand: you can spend a lot of time creating some cool technology and hiding it from the world, and at that time someone else will do something similar, but not so elaborate, will launch it on the market and win.
Now, for example, we have a new face recognition technology, we need a year and a half to complete it, but in three months we will begin to launch its special cases on the market. The sooner we do this, the sooner we will get feedback, we will correct what the client did not like, we will make the product better.
Another problem was remoteness from the market. We thought that we knew the market extremely well, talked about things that seemed absolutely logical to us, but in fact everything turned out to be not quite as we expected. We did not take into account factors such as mentality, nationality, and business peculiarities in Russia and in other countries. For example, we thought that the technology of face recognition would be very necessary and important for banks, we were going to create a network of cameras for them, but over time we realized that not everyone was ready to meet us. Many people prefer the old well-known security cameras, rather than the new, not yet known technology.
Text: Yana Borodyuk