AI and drones target moose damage in Finnish forests

Researchers are testing drones and artificial intelligence to protect tree seedlings from hungry moose, Yle reports. The HirviDrone project aims to identify which seedlings should be treated with a repellent and apply it only to those trees.

Drones first photograph a plantation, after which artificial intelligence identifies the trees and assesses their condition. A spraying drone could then distribute a bio-based feeding deterrent to selected seedlings.

Jere Kaivosoja, a senior researcher at the Natural Resources Institute Finland, said the method could eventually be much faster than current backpack spraying. Cameras and artificial intelligence could identify tree crowns and possible damage, allowing a spraying drone to determine where the substance should and should not be applied.

The repellent contains sheep fat, which moose find unpleasant to taste.

Moose were the most significant animal group causing forest damage in Finland in 2025, according to the Natural Resources Institute Finland. Damage was recorded across about 337,000 hectares. In northern Finland, damage affected 175,100 hectares, while southern Finland recorded 162,000 hectares.

The aim is not to drive moose completely out of plantations. Juho Matala, also a senior researcher at the Natural Resources Institute Finland, said the trees intended to grow into large, undamaged trees would be treated, while other seedlings and lower branches would remain available as food.

Drone application is currently permitted only for research purposes. It requires permits from the Finnish Safety and Chemicals Agency and the Finnish Transport and Communications Agency.

Kaivosoja said spraying drone technology was developing quickly, but the recognition algorithm needed for plantations was still under development. Matala said current equipment was largely designed for agricultural use and was not well suited to treating individual seedlings.

Regulations governing aerial application and equipment approval requirements are also slowing the technology’s introduction.

In Kuusamo, the local jointly owned forest has about 1,500 hectares of plantations where moose cause damage, according to its managing director, Jarmo Korhonen. Damage is severe across about 1,000 hectares, and some areas have had to be replanted. Korhonen said compensation for moose damage covered only a small share of the losses.

The repellent was tested by drone at the jointly owned forest’s Jokilampi tract, which covers about 10,000 hectares. Hunting is currently the most important way of limiting damage there, but Korhonen said drones could eventually provide additional help.

HirviDrone is funded by the European Agricultural Fund for Rural Development and involves the Finnish Forest Centre, the Natural Resources Institute Finland, the National Land Survey of Finland’s Finnish Geospatial Research Institute and the Finnish Wildlife Agency, along with several forestry organisations. The project continues until the end of 2026.