MOSCOW, September 7. A unique system for smart farms, which has no analogues in the world, is being created by scientists from Tyumen State University. According to them, the development will be able to “revolutionize” the urban farming industry by combining all the necessary knowledge and skills in one neural network assistant. The results were published in the journal Informatics and Automation.
Today, one of the main directions of development of agribusiness is the development of smart farming systems capable of automating plant care, experts reported.
Researchers at Tyumen State University are developing a unique system of this type, which, according to them, can act as an assistant and consultant for users with the most superficial knowledge of crop production and farming.
The goal of the system is to make urban and other modern agricultural industries radically simpler and more efficient, the creators emphasized. According to them, there are no analogues to this product in the world today.
As experts explained, the development is implemented as a “superstructure” for existing smart farming systems. Using elements of artificial intelligence (AI), it analyzes the parameters of plants and the environment based on sensor data, and allows not only to optimize routine processes, but also to find solutions to non-standard problems.
“
“With the help of AI neural network elements, our system will make smart farms even more intelligent. Now such a farm will be able not only to monitor and control the parameters of the production process, but also to give advice on how to act in case of equipment breakdowns, plant diseases and other emergency situations,” said Alexey Prokhoshin, head of the Tyumen State University’s scientific agrobiotechnical complex.
< /span>
As the creators emphasized, unlike many existing automated agro-industrial complexes, the new system will have the competencies of several experts at once.
«Standard algorithms for smart farms do not take into account a number of regularly occurring problems. Equipment breakdowns, diseases or plant pests — not every specialist is able to quickly and competently cope with such situations. Our development will help to promptly recognize problems and find the necessary solutions for them, even with a lack of experience and knowledge among farm workers,” explained the scientific director of the project, head of the Department of Information Systems at Tyumen State University Igor Glukhikh.
The system takes into account many parameters — from microclimate and lighting to signs of various diseases. Based on this analysis, the system will formulate specific recommendations for the user aimed at increasing productivity and product quality, the university reported.
< /span>
«Imagine an agronomist with vast experience and intuition, who tirelessly monitors every plant and all the parameters of his farm, knows about any needs and instantly responds to any changes. Relying on the power of AI, our system is able to implement such supervision,” noted Prokhoshin.
A large interdisciplinary group of Tyumen State University scientists is working on the creation of a new system, which includes biologists, mathematicians, systems engineers and other specialists.
«Today, neural networks are being actively created that can recognize plant diseases from video or photos We are taking the next step: our system will not only recognize, but also find real solutions to eliminate the pathogen, taking into account many factors, including the availability of certain drugs,” Glukhikh emphasized.
Currently, a team of scientists is completing the preparation of the development for commercial implementation, and is also working on a mechanism for continuous learning of the system through user feedback.
Tyumen State University is a participant in the state program for strategic academic leadership «Priority-2030». The research is carried out within the framework of the strategic project «Biosafe World».