Fully automated farming can be a solution to minimize hunger in the world. Keeping track on livestock automatically reduces operational costs and improves productivity at the farm. This project is a feasibility study of system components for automated livestock monitoring using drones and artificial intelligence.
The Federation of Swedish Farmers (Lantbrukarnas riksförbund - LRF) estimates livestock lost to predator attacks yearly in Sweden to a few MSEK. The costs of daily supervision over the livestock are however substantially higher: looking daily for scared och sick animals which got lost due to this or any other reasons is estimated up to 50 MSEK for cows alone. Similar numbers apply to sheep and horses.
Monitoring of the livestock is also required by the EU and national regulation which in particular requests farmers to keep track on places where livestock consumes grass and other nutrition.
This project is a feasibility study and a test of the capabilities of the existing drone and AI solutions applied in the farming context. We need in particular to investigate a possibility to distinguish individual animals with this method. The project will answer with which accuracy drone plus AI system can perform this task.
The tests lay a fundamental foundation for a fully-automated livestock monitoring system.
AI-assisted drone livestock monitoring
Koordinator och AI ansvarig