Intelligent Airborne Monitoring of Livestock Using Autonomous Uninhabited Aerial Vehicles

Kuru, Kaya orcid iconORCID: 0000-0002-4279-4166, Ansell, Darren orcid iconORCID: 0000-0003-2818-3315 and Jones, David (2023) Intelligent Airborne Monitoring of Livestock Using Autonomous Uninhabited Aerial Vehicles. In: The 11th European Conference on Precision Livestock Farming, 09-12 September 2024, Bologna, Italy. (Submitted)

[thumbnail of ECPLF-KK1.pdf]
Preview
PDF - Other
Available under License Creative Commons Attribution Non-commercial.

379kB

Official URL: https://www.ecplf2024.it

Abstract

Precision Livestock Farming (PLF) is one of the most promising applications showing the benefits of using drones where a lack of human element in the farming industry is becoming evident. UAV-assisted smart farming within large farms has gained momentum in managing large farms effectively by avoiding high costs and increasing the quality of monitoring. To this end, the high mobility of UAVs combined with a high level of autonomy, sensor-driven technologies and AI decision-making abilities can provide many advantages to farmers in exploiting instant information from every corner of a large farm. The key objective of this research is to develop user-friendly AI-based software that can combine the sensor data sets and accurately detect animals and health anomalies, so the information can be presented in an easy-to-understand on-demand format for livestock farmers to take targeted or preemptive action, and improve the health, welfare, and productivity of their livestock. In this research, an automated drone solution with a cross-discipline approach has been developed to periodically survey livestock in an automated manner using vision-based sensor modalities involving both standard visual band sensing and a thermal imager. The experimental results suggest that the accuracy rates of detecting livestock are very high with very high sensitivity (Se) and specificity (Sp) values. Additionally, the results regarding the animal body heat signatures obtained from the thermal imagery show promising results in detecting disease-related cases. This research is a productivity and sustainability-focused pilot to investigate and demonstrate how drones and artificial intelligence software can provide a better way to regularly inspect animals on a large farm to avoid high costs and to increase the quality of monitoring. The research demonstrates how highly integrated technologies with drones can help the farming industry to overcome the challenging issues in the management of livestock, particularly, health monitoring of livestock in very large farms in an eco-friendly and sustainable way.


Repository Staff Only: item control page