How One of the Most Digitized Livestock Farms in Russia Implemented Innovations

The path from restructuring to digitalization is interesting and useful, but not easy.

LLC "Agromilk" is a producer of raw milk from the Vologda region. Today, the farm has 1,300 Black-and-White breed cattle, with an average milk yield of about 21 liters per cow per day. In addition to livestock farming, the farm processes 2,800 hectares of arable land, primarily growing feed for cattle, and 82 hectares are dedicated to vegetable growing.

Sergey Blyuma, CEO of "Agromilk" — on the stages from restructuring to full digitalization.

There was a problem, but the experienced team took everything into their own hands

The farm restarted its activities two years ago after bankruptcy. Since November 2021, I became the CEO of "Agromilk." At the time of restructuring, the farm had difficulties with accounting and production monitoring.

We decided that it was important to control processes by striving for completeness and accuracy of data, and to implement unified standards. This helps to identify and correct mistakes on time. Ultimately, it makes production more efficient.

The digital update of "Agromilk" was not something new or unexpected for me or my team. In parallel, I head the Continental Group, where we work on telemetry, microclimate automation, collecting data from manure removal, water supply, feeding control, digitalizing milk loading, and lighting. The experience gained at other enterprises was confidently brought to "Agromilk."

Farm transformation, from equipment tracking to computer vision

In a year and a half of work, we created one of the most digitized farms in Russia. The farm has been transformed across many parameters, including staff accounting, equipment tracking, crop and livestock management, and planning. We control the microclimate, water supply, and milk shipment. We launched a pilot project on computer vision and artificial intelligence.

A massive amount of information started to appear, which can now be analyzed and interpreted. AgroSignal is one of the new elements of the farm's digitalization. It’s our newest project, and we continue to work on its implementation. In the first season after the restructuring, we immediately understood that it was important to know where the equipment is, how it moves, where it goes, and why. To do this, we equipped the equipment fleet with sensors. Through tracking and AgroSignal, we have already optimized the number of necessary pieces of agricultural machinery, fuel costs, and personnel. This solved some problems, but the main one remained — feed costs. So now, we are focusing more on crop farming to grow high-protein grasses, without making them "golden."

The lack of a reliable knowledge source seriously hindered the farm’s operations

While I have 27 years of experience in livestock farming, the same cannot be said for crop farming. Now, we are gradually getting a handle on this area. Yes, we equipped the machinery with sensors and launched tracking, but the question arose: what else can we do to improve crop farming performance?

The complexity with staffing also pushed us toward digitalization. Two years ago, long-term employees left the company, and young staff took their place. The knowledge accumulated by the previous generation was not recorded anywhere. As a result, the transfer of processes was difficult for the younger employees. Having been in this situation, we were once again convinced that we needed standards and a knowledge repository. The ability to scale and share knowledge. For this, digitalization is also necessary.

What showed us that crop farming needed to be quickly placed on solid digital rails was the history with feed. Two years ago, the farm simply could not provide its animals with enough feed. Part of the herd had to be sent to the meat processing plant. In the first year, we were only able to harvest the required amount of silage, but we could not meet the quality standards.

Of course, we could blame the weather, but I see the problem in organization and understanding of technology — poorly optimized processes, disjointed interaction with people, lack of exchange of experience and problems, with each having their own ideas. As a result, there was quantity but no quality. We decided that digital products and automation would help us solve most of these problems.

Searching for a digital solution for crop farming

We didn’t wait for a guru to approach us, promising to open the way to enlightenment. We started approaching different companies with a request. The first product we reviewed was Cropio. Its capabilities interested us, but at that time we already understood that foreign software could encounter support issues, as happened with livestock farming software. The program worked, but it stopped being updated and modernized. So, knowing the potential risks, and based on the experience of our partners, we came to the understanding that we needed to look for a Russian product. One that would be regularly updated and function smoothly.

By regularly attending industry exhibitions, we saw several attractive solutions for crop farming. However, when communicating with company representatives, we encountered implementation and support issues.

Everyone said, “Buy the product, our employees will come and set everything up” — but we’re implementers ourselves, and we know from experience that you can’t always travel everywhere. There is a lot of software on the market, but not every product has detailed instructions and descriptions of key functions. As a result, when you hand the program to a specialist, they try to figure it out, but end up saying, “Sorry, Sergey Nikolaevich, the software is useless.” The data on production processes is not filled in, and we understand why. The specialists lack the knowledge of how to work with the program properly. It takes a lot of time and effort to figure it out. Our employees aren’t just studying manuals, they are actively working in the fields.

Therefore, when searching for a digital product, we started negotiating with companies that could not only implement it but also provide ongoing technical support. We didn’t expect to be convinced that their software would bring benefits to the business. The main thing for us was not just implementation support, but that employees would have continuous real-time support. In case of questions, they could ask for help and find out how to use the program.

Meetings with other teams gave us fewer hopes than AgroSignal. If you don’t have technical support, the product already feels flawed. Since the end of last year, from the first days of January, we started the project to deploy AgroSignal. Given that we had already completed the financially intensive part of equipping the equipment with sensors, the installation of the program and the transfer of key field parameters were done promptly, in about a few weeks.

Excluded processing of foreign fields and obtained measurable crop farming processes

Now we receive valuable data that includes information about who and when conducted the planting, which specific plots were used, what type of grass was sown, humidity levels, and temperature during the work. We also know who worked the fields, what the movement patterns were, and, of course, the actual productivity.

Surprisingly, the productivity parameters of agricultural machinery are a huge issue in our industry. I never thought about it before. In livestock farming, we don’t discuss the productivity of machinery. Of course, there are parameters like speed and other characteristics, but when you start talking about such simple things as “how many hectares per hour can this mower process,” the range of answers is huge. And it boils down to such abstract words as “it depends on the field, weather, soil.” This seriously undermines labor organization efficiency because many, including machinery suppliers and agronomists, give vague numbers like "half to the ceiling."

Today, thanks to AgroSignal, we have a deep understanding of production processes. We realize that equipment, for example, a mower, can process 60 hectares of field, while another only 40 hectares. Based on these parameters, to avoid agronomic mistakes, it needs to be worked differently and logistics need to be planned.

Clearly, the result entirely depends on the efforts of the manager; data alone does not generate profit. They only provide an understanding of what efforts need to be applied for the farm to work productively. AgroSignal has certainly helped us understand the effectiveness of the work. It takes time to turn this into tangible monetary Image by senivpetro on Freepik