Maxim Zverkov is an accomplished entrepreneur and business development expert with two decades of experience. He owns and leads 7 independent companies, employing hundreds of staff, and together they achieve an annual turnover of $50 million.
With a strong emphasis on strategy and business development, Zverkov has amassed extensive C-level management expertise.
His journey with the Abagy startup began when he encountered challenges integrating robots into his equipment and metal structures production business.
Recognizing the need for new technology, he set out to enable the use of robots not only in mass production but also in high-mix, low-volume scenarios. In the interview below, Maxim gives insights into this success story.
A few years ago, I installed robots at my company and tried to use them in painting structures for exhibition stands and equipment. We had a very high production volume, and the idea was to automate technological processes in order to reduce production costs and manufacturing time. However, the experiment failed: the robots were not used to the extent intended.
I began to analyze what the problem was, and I realized one simple thing. With our impressive sales, only a tiny part of our production could be considered high-volume. Most of our production was small-batch or single-unit. In such conditions, "classic" robot technology does not justify itself.
In terms of a single operation, it is much more profitable to use a robot manipulator invented forty years ago than human labor. But this is true for a large number of single-type operations, that is, for mass and high-volume production. The total robotization of the automotive industry is a good example of this fact.
Why did it take place? Because automobile production means a lot of precision piece parts, multiple and accurate repetition of operations. Here, robots perfectly replace humans and bring clear benefits to the automotive company.
However, in forty years, only 2 million such robots have been installed worldwide, which is less than 1% of work positions. This is the International Federation of Robotics' estimate for the end of 2023. There is a strange asymmetry: on one hand, robots bring benefits, on the other, their implementation is slow. Why is this the case?
Robots are good for everything except one thing: they cannot "think". Industrial robots, which are mostly used in modern production facilities, are not adaptive.
They work according to a predetermined program and cannot react independently to changes in external conditions. To perform a new task, such a robot must be reprogrammed. This sometimes takes up to a week, which is costly for small-scale production.
And since the trajectory of the manipulator is based on reference points, it is necessary to accurately position the workpiece. This is not always possible. And it turns out that a human worker, thanks to their ability to quickly adapt and perceive new tasks, wins over the robot in the classical automation paradigm.
The idea behind Abagy was to eliminate the bottleneck of classical robotics - the need to program the robot. What was the way to do it? The 3D model of a workpiece that a designer creates in CAD is essentially a set of numbers. The trajectory of the manipulator is also a set of numbers. Our task was to fit them together.
After going through many hypotheses, we found the engineering solution in combining machine vision and artificial intelligence systems. First, the robot scans the real workpiece and then matches it to the 3D model. On the finished 3D image, the technician sets the trajectory of the tool, and then the robot is ready for operation. The whole process, depending on the complexity of the workpiece, takes 5-20 minutes – just what is needed for small-scale and single-unit production. This is how the Abagy project came to life.
It’s hard to formulate a concept that could be replicated universally. But there are things that have been fundamental to me from the very beginning.
I always count on people and pay attention not only to the necessary qualifications. Our core team has been intact from the very beginning, and there have been team members who have made a great contribution to the development of the project and have moved on.
New people have taken their place, and it is important for me that they are interested as much as possible in what we do at Abagy.
As far as business management is concerned, I always stay up-to-date. But at the same time, each manager chooses the ways and mechanisms of developing their own line of business.
Any startup is a small and generally flexible structure at the beginning of its operation. It is often this flexibility that is its advantage over large corporations. With minimal resources, it can implement ambitious projects that giants cannot do because of their sluggishness and bureaucratization of processes. Taking advantage of our youth, flexibility and mobility, we were always testing hypotheses, improving our product, and entering new markets.
For example, in the beginning, in addition to technology, we also offered the market a pay-per-use model. Here's how it worked. We installed robots at the client's facility at our own expense, and the client paid for the work done by the robots using our software. For this market where equipment is expensive and difficult to reuse, such a model was new.
We gradually moved away from pay-per-use, although we still work that way with one client. Our main focus now is on the technology itself, because it can bring business value to industry.
We did not choose software development, but came to it as a result of analysis. Why have robots been effectively assembling automobiles and microelectronics for many years, but are being used to a little extent in other industries? Back then, the robotization density was 1 robot per 100 workers, now the figure is slightly higher.
The answer was simple – robots are economically inefficient for most businesses because, unlike workers, they can't adapt to changes in environment and tasks.
Therefore, we started making "brains" for robots. Our software enables the robot to adapt to external conditions. Thanks to this, we have lowered the threshold for implementing robotics tenfold. Installing a classical robot is profitable if the cost of the operation for its entire time starts from around $1 million USD. For our robots, this threshold is about $100,000 USD.
Robots can also be used effectively by low-volume enterprises – factories that manufacture a variety of products in small batches or even in single pieces.
Abagy's solution is aimed at them: we connect robot solutions, i.e. hardware, to our own software in the cloud. And the manipulators start to "see" and "think": even if the spare parts meet just the minimum accuracy criteria, the machine will still understand how to deal with them.
Let's start with the competitors. There is a common issue: programming a robot is time-consuming and requires it to be taken out of the production process. There are several startups in the world developing adaptive robotics solutions: Path Robotics, Covariant, Omni Robotics, AGT Robotics, Zeman, Kranendonk, Inrotech, and others. Each company has its own approach and direction: some solutions are tailored to the production of specific products, while others are aimed at optimizing the programming process.
For example, there are niche technologies from AGT Robotics and Zeman. They are mainly designed for I-beam production. There are offline programming systems Delfoi and Octopuz which use a digital twin of the robot. Firstly, it is easier to program a twin than a "live" robot. Secondly, it is not necessary to stop the robot for the programming period.
We make a different product. Abagy doesn't make programming more convenient, in fact, it removes it. And this allows our software to be used in a wide variety of production processes.
Now about the secrets, so to speak. It is very important to go into production as early as possible. You need to check via live operations how viable your software is.
We had been making a product for about two years before the first implementation in production. It wasn't an MVP, it wasn't a pilot, it was a production unit. And after the implementation, our technical manager spent about half a year at that production site, around the clock. As a result, the software in its current form is 90% rewritten compared to what it was before installation.
Only real production reveals the true requirements for a technical solution. You can interview 100 managers of different levels and find out how the process is organized, what they expect from robots, what tasks they want to perform, and much more.
But in a factory, things are quite different. And here the developer starts to adapt the product to the tasks of specific production managers, chief welders and other people responsible for the production process. The earlier we start doing this, the better.
70% of Abagy's employees are programmers and engineers, they are highly qualified staff. That's what makes it special. But it probably has nothing to do with the industry.
For sure, we hold on to our professionals. In the labor market, it is difficult for a startup to compete with large companies willing to pay programmers large salaries. That is why we are working on motivation. I think Abagy employs people who believe that our technology will overturn the established technological order.
Our technology is very difficult to grasp. Only professionals who have already faced the problems and limitations that exist in the industrial automation market can understand its ability to solve real production problems.
That is why we consciously do not use marketing tools designed for a wide audience: mailings, articles, organizing events. We focus on courting experts who solve specific production tasks.
We regularly make videos about our technology. We show cases that are of concern to industrial owners: welding specific products with robots or comparing the time factor of the same operation between a robot and a human.
We keep a video blog and regularly update our website, from which we receive requests every day. These are live leads, people who are interested in solving their production tasks using robots.
We also participate in Fabtech, a specialized exhibition for welding, forming and metalworking. This is a very impactful event for our company. For the second year in a row, we gain a lot of leads and receive new contacts.
Abagy has several investors: they are both business angels and funds. At the moment, we are focused on sales and on new implementations. Still, our development strategy includes global projects from which we plan to attract investment.
Abagy works on equipment from all major manufacturers. We are interested in broad cooperation and find it unacceptable to focus on a specific brand of robots, which would limit the development of the technology overall.
As the number of Abagy installations grows, we will consider the prospects of a partnership sales business model.
Here is one example of how acute the labor shortage is. According to the American Welding Society, the United States is projected to face a shortage of 400,000 welders by 2024. Right now, 80% of welders are over 35 years old, and young people seem unwilling to take on this dangerous and difficult line of work.
The same goes for other workers. At the same time, the process of robotization is progressing very slowly. The International Federation of Robotics estimates robot density in the United States at just 274 units per 10,000 employees in 2021, or 2.74%.
For us, 2022 was the year of launch in the US market. And we are experiencing a lot of interest from US manufacturing companies. We are optimistic about next year and expect an increase in the number of installations in the US.
We have two product formats. In the first format, we offer a ready-to-use solution that includes our software and robotized equipment. The second format involves upgrading an existing robotic cell. For example, Abagy upgraded a robotic cell for Schenck Process by installing machine vision and other equipment compatible with proprietary software.
India is certainly a promising market for us, and our solution could be in good demand there. There are quite a lot of production plants in India, but there is also a shortage of labor in some regions. Therefore, Abagy can be of great value in industries such as shipbuilding and construction materials production.
We have gained a lot of experience along the way in developing the Abagy project. We have analyzed industry needs, developed and implemented innovative technical solutions – and we continue to develop them. We have grown into a close-knit, motivated team. We learned how to set ambitious goals and achieve them. Certainly, this experience will be useful for new startups as well.