Unter der Leitung von POLICIES startete Anfang April das neue Forschungsprojekt AISAAC. The aim is to develop an AI-based monitoring system that evaluates the sustainability of agricultural land on a small scale. Damit sollen unproduktive Teilflächen identifiziert, Renaturierungsmaßnahmen unterstützt und der Ressourceneinsatz in der Landwirtschaft optimiert werden.
Thinking about agriculture in a more sustainable and data-based way
New approaches to agricultural land management are needed to meet the growing demand for food while at the same time preserving natural resources. This is precisely where the researchers in the AISAAC - AI-based Sustainability Assessment of Subarea-Specific Managed Cropland project come in. The project is being carried out as part of the Austrian Research Promotion Agency's (FFG) AI for Green funding programme and is financed by the Austrian Future Fund.
‘We want to create a data-driven basis for decision-making in order to harmonise productive and ecological goals,’ says Ulrike Kleb, project manager and senior researcher at the POLICIES Institute.
Precise predictions with Explainable AI
The project centres on the development of Explainable AI (XAI) models. These are intended to visualise the relationships between soil quality, weather data, resource use and crop yields - and use them to forecast the yield potential of individual sections of land. In addition to historical harvest data, high-resolution satellite images (PlanetScope and Sentinel-2) and moisture sensor data are also incorporated into the models. This allows precise maps to be created that support decisions regarding cultivation and renaturalisation.
Focus on sustainability assessment
Another aim is to develop a sustainability index that can be used to assess small-scale agricultural areas in the long term - taking biodiversity, soil fertility and resource utilisation into account. A prototype of the monitoring system is being trialled on a reference field belonging to the project partner Weinland AGRAR.
Broad consortium from research and practice
Das Konsortium vereint Partner aus Wissenschaft, Landwirtschaft und Technologie:
Neben JOANNEUM RESEARCH POLICIES sind auch das Institut DIGITAL, Forschung Burgenland, Maschinenring Agrar Concept, Weinland AGRAR sowie aGRAR-ZT beteiligt.
The project runs until March 2027 and aims to create a basis for more sustainable decision-making processes in agriculture - data-based, explainable and practical.
