BelSU Scientists Develop Smart Program for Safe Animal Grazing

A growing number of agricultural enterprises are turning to cutting-edge technologies such as the Internet of Things (IoT), movement trackers, unmanned aerial vehicles (UAVs), GPS, GLONASS, and automated systems to reduce operational expenses.

“There are similar developments in countries like Australia and China. Our goal is to advance the scientific and technical aspects of the domestic smart livestock industry. We have proposed a novel model for managing the location of grazing animals through the integrated use of drones and the Internet of Things,” stated Konstantin Polshchikov, a professor in the Department of Information and Robotic Systems at Belgorod National Research University (BelSU).

Professor Polshchikov explained that inexpensive sensors can be installed on cows, sheep, and other livestock to determine their exact coordinates via GPS. These sensors transmit signals to the operator-shepherd through the IoT. The shepherd will have access to a comprehensive interactive map of the pasture displayed on a laptop or tablet. If an animal approaches the pasture boundary, the shepherd can quickly deploy a drone to guide it back to safety using sound cues, mimicking the behaviour of traditional shepherd dogs.

“This approach is significantly more economical than utilizing a drone that is constantly airborne. In our system, the drone remains stationary most of the time and is deployed only when necessary,” emphasized Polshchikov.

The development from the Institute of Engineering and Digital Technologies at Belgorod State University addresses several challenges faced by agricultural enterprises, such as animals wandering far from their pens while grazing, which can lead to injuries, loss, or damage to property. Additionally, this model could substantially reduce the risk of livestock theft.

“While trackers are designed for low power consumption, excessive data transmission can rapidly deplete their battery life, leading to shutdowns. Our system effectively mitigates this issue, optimizing energy use for the trackers,” explained Professor Polshchikov.

The researchers noted that GPS trackers are integrated into a single network through 8-channel and 16-channel LoRa gateways, which have limited performance capabilities. In large herds with numerous connected trackers transmitting data simultaneously, not only does energy consumption increase, but there may also be delays in message delivery to the shepherd due to network congestion. Such delays can hinder timely responses to potential dangers in the pasture.

To address these challenges, Professor Polshchikov and his team developed a mathematical model that considers various factors influencing the efficiency of message transmission from trackers. This model calculates optimal time intervals for data transmission.

According to Professor Polshchikov, the proposed model categorizes animal locations into four classes based on their distance from the pasture boundary, assigning optimal messaging intervals for each tracker. This enables the prompt deployment of a drone to guide animals back to their designated grazing areas. The model also determines the optimal number of trackers and drones required for the safe grazing of specific herds. Computational experiments have demonstrated that the proposed system significantly enhances grazing safety.

The scientists’ model primarily utilizes standard 8-channel and 16-channel LoRa gateways, but they plan to expand the system's capabilities to accommodate a larger number of trackers in the near future.

The developed model has been formalized and registered as a computer program, which can be utilized by agricultural enterprises implementing smart livestock farming practices, as well as by companies providing monitoring solutions for cattle and other farm animals.

“This computational program can be integrated into a broader management system for the entire agricultural process,” explained Polshchikov. Looking ahead, the researchers aim to refine their herd management model further. Their next objective is to conduct field experiments to validate the real-world characteristics of animal grazing.