"Borodulinskoye": A Farm with a Digital Twin

Creating a digital farm, the management of LLC "Borodulinskoye" aimed to ensure business sustainability. The results achieved during the project not only improved financial outcomes but also prompted the consideration of business scaling and the creation of an experimental training center based on the agro-farm. Alexander Smirnov, IT director of the agro-farm "Borodulinskoye" and nominee for the Data Award, talks about the approaches to implementing the project.

— What does the farm do, and what led to the need for its digitalization?

"Borodulinskoye" is a dairy farm in the Sverdlovsk region. It houses 1,260 animals of the Black-and-White Holstein breed, about 600 of which are dairy cows. As of 2020, the farm’s average milk yield was less than 10.5 tons per day, which is objectively low—it could be much higher.

The farm lacked constant monitoring of production processes, had low manageability, and lacked sustainability. The co-owners of the farm set the task of finding solutions to these problems.

— What exactly did you want to achieve?

Our main goal was to create a stable, efficient, and transparent model for the enterprise, where we would understand all the business processes occurring on the farm. This way, we can anticipate situations and make decisions not based on the facts but by identifying trends and reacting in advance, including through the implementation of new practices. For example, the ability to create a "copy" of the farm in the "digital twin" at any time, where several parameters can be adjusted to get forecast analytics on how the changes would affect the overall result, allows us to reasonably consider requests for business scaling.

— What arguments can be presented to the business to enable the full digitalization of farm operations?

Farm digitalization is only necessary for those who want to see the real picture and manage the enterprise based on data. For the owners who invest in business development, this model is suitable. The main argument, in our opinion, is not just the desire for profit, but for sustainable profit. Only a farm that is resilient to various negative impacts can not only ensure return on investment but also scale the business.

An unexpected argument, which is relevant not only for agriculture, is the effective and complete implementation of software IT solutions. Initially, developers embed extensive functionality into every software product, but the effectiveness of using these tools greatly depends on the current, objective, and quality data entered by employees. Digitalizing processes allows the automation of data collection from various sources and ensures that the implemented solutions receive the necessary information.

— What are the investments in this project? What is the expected payback?

Investments do not exceed 25% of the additional annual profit from milk. They are not yet finished, but the results obtained justify and allow these investments to be secured from the owners. To date, we have invested in the DairyComp305 herd management system, the TMR Tracker feed preparation and distribution system, the Dairy Production Analytics (DPA) and Business Scanner services from "ALAN-IT," climate sensors, and Heatime collars. We are currently finalizing the justification for investments in a system for automating climate control, lighting, and telemetry collection related to milk production volumes.

— What were the stages of the project, and how long did it take?

The initial stage was related to the implementation of modern herd management systems, after which modules for feeding management, reproduction management, sensors, and analytical services were implemented. A significant result appeared by the beginning of 2022 when we increased the milk yield from 18 kg per dairy cow per day to 21 kg. Currently, the milk yield is around 30 kg per day.

The digitalization of the farm would have been incomplete without the digital agriculture module. The core digital component in this part became the "AgroSignal" service from "Infobis." As part of this task, we implemented modules related to transportation and technological operations control, fuel consumption monitoring and accounting, maintenance and repair tracking, speed control, and the accuracy of technological operations. We digitized the boundaries of all fields and refined them during real work. The manager can plan and analyze work for each field, and the system automatically controls the harvesting process, ensuring control from the filling of the combine hopper to transportation and weighing.

Currently, we are working on further data detailing from the fields and their agrochemical and retrospective analysis. Ultimately, we aim to create a "digital twin" of each field, which will be used both for work planning and for the full cycle of precision agriculture technologies. At the moment, we are discussing the integration of both modules into a unified system based on DPA solutions with developers.

— Were there any alternatives to creating a "digital farm"?

We have several business areas, one of which is property management. We are actively implementing the Internet of Things based on LoRaWAN technology from a foreign supplier, which allows us to quickly collect data from various accounting devices in buildings. Therefore, we were also studying the market for similar sensors applicable to agriculture.

In 2020, we came across a case study on the implementation of LoRaWAN sensors on one of the websites, which described the approach to digitalization on a farm in the Yaroslavl region. We contacted the solution developer, "ALAN-IT," which had already implemented a Russian platform solution for a digital farm. It turned out that alternative solutions were either tied to specialized imported equipment or were universal solutions like SCADA systems. These systems, without major modifications, could not upload data from herd management or animal feeding systems.

Another important factor at that time was the Russian origin of the software products. By 2022, most farms were using foreign herd management solutions distributed via SaaS models and faced problems. At each implementation stage, we try to adhere to "digital independence," where the lower-level sensors and communication channels can operate independently of the next-level software products and even send data to several systems if necessary.

— How have business processes and decision-making changed? What led to the growth of performance indicators?

At the beginning of the process, not much effort was needed. Using the DPA solution, we analyzed the current situation on the farm and provided recommendations for improving data entry processes into the herd management system. For example, it turned out that when animals left, the specific reason was often not indicated or was listed as "Other." The same situation occurred with diseases. The first modification of business processes was quality data input, which allowed for a deeper analysis of animal care processes, insemination, feeding, and milking. The farm's management also made it possible to not only solve specific problems but also see if they were being solved and within what timeframes. The staff's understanding that their actions are visible and managed improved the quality of routine procedures, and the timeliness of actions had the desired effect.

Another example: we never thought about how much of a negative effect long-lactating cows had on the farm. They made up a high percentage of the herd, meaning we were at the end of the lactation curve. When we started the work of culling such cows, inseminating, and gradually replacing them with heifers, it allowed us to achieve much higher productivity in the herd.

— Were there any insights from data analysis that were completely unexpected?

We have installed climate sensors that measure temperature and humidity on the farm and show the animals' heat stress index in the summer. In July-August of last year, we had a reduction of about two tons of milk per day. The presence of sensors allowed us to understand the reasons for the decrease in milk yield and the negative effects of heat stress on disease rates and insemination success during this period. This, in turn, allowed us to justify the implementation of a climate management system from an economic perspective.

— What business results did you expect when creating the solution? How do they compare with reality?

We expected a 20% increase in gross milk yield in one year, and the actual results exceeded our expectations. Previously, we were milking around 10-11 tons per day on the same herd, but now milk yields are 17-18 tons, resulting in a profit increase of over 40 million rubles from raw milk compared to 2021. The rest of the farm's economy looks about the same.

— What role can the project play in the development of the market?

We want to implement the idea of creating an experimental training center based on our agro-farm. The Ural region needs specialists who can apply such technologies, and we would like to share our experience and exchange knowledge with other farms through the training center. We are also ready to collaborate with universities and specialized colleges to train young specialists. Students who have just graduated or completed advanced training courses lack experience working with modern equipment and software, which they need to use. The experimental training farm can provide this opportunity in collaboration with solution developers.

Moreover, we see the future training center playing a role as an IT accelerator in agricultural projects. We encountered the fact that many small IT startups are entering the market, but they don’t always have the opportunity to test their solutions in practice. The target audience (agronomists, animal husbandry specialists) has conservative views on innovations and is not ready to engage with such companies. Startups need expert evaluation of the solutions they plan to develop and test them on real objects. Agricultural enterprises, in turn, need IT specialists who must gain experience in real conditions. In the future, specialists from the training center could help scale successful IT solutions to other agricultural enterprises.

A third important component is the opportunity to be among the first to test and implement innovative solutions. For example, in the summer of 2022, our farm participated in the pilot project of testing boluses – capsules with sensors placed in cows' stomachs to monitor their health. The prepared communication networks allowed us to quickly implement and test any modern solutions. Data from the boluses was formatted in Excel, but that didn’t stop us from quickly transferring it to DPA for comparison with data from other systems on the farm.

— How does the farm see its future?

We want to renovate the farm and increase the number of dairy cows. We will continue to solve other tasks, digitize the remaining business processes, and scale up a new business direction – the production of premium dairy products.