AI-based remote sensing of nature restoration and landscape elements

The two-year initiative — funded by the Swedish National Space Agency and Arla Foods and run in collaboration between RISE and Arla — aims to develop AI-based methods that use satellite images to assess the development of natural grasslands over time, as well as to map and identify landscape elements that contribute to biodiversity at farm level. Drone data is also collected to enable annotation (description of image content) and verification at a higher resolution than is possible with satellite data.

The aim is to develop effective, data-driven tools that can strengthen and verify the benefits of farmers' efforts to promote biodiversity and support sustainable agricultural practices.

The project responds to a growing need within the food industry to demonstrate concrete, location-specific results in its work on sustainability and biodiversity. At EU level, regulations on land use, nature restoration and reporting requirements are being tightened, including through the new Nature Restoration Law.

By combining satellite data with AI, the project aims to provide Arla with better insights into landscape elements that are crucial to ecosystems – without having to physically inspect each farm. The tools developed are intended to support advisory services, environmental certification, reporting and incentive schemes linked to conservation efforts.