SmartCut
AI-based prediction of yield, quality and optimal mowing time for permanent grassland

SmartCut is an EIP-AGRI innovation project developing an AI-based forecasting model that helps farmers determine the optimal mowing and fertilisation date for permanent grassland. The goal is to increase grassland productivity and forage quality while enabling more sustainable, resource-efficient dairy farming. The project is funded under the European Innovation Partnership for Agricultural productivity and Sustainability (EIP-Agri) via the CAP Strategic Plans 2023–2027 and runs from 2026 to 2028 with a total volume of around €1.09 million.
Why mowing time matters
The dairy industry is the economic backbone of the grassland-rich low mountain regions of Rhineland-Palatinate — Eifel, Hunsrück, Hochwald, West Palatinate and Westerwald — yet it is under pressure: the number of dairy farms is declining sharply while farm sizes keep growing. The quality and yield of grassland play a central role in the efficiency of livestock farms, and the timing of mowing is a decisive factor for both forage quality and yield.
Today, farmers largely determine the right moment by visually inspecting fields and checking weather forecasts. Modern forage harvesters with NIR sensors already record high-resolution yield and quality data during harvest, but small and medium-sized farms in particular derive little benefit from this data so far. Corresponding models for AI-based determination of yield potential in grassland are lacking.
What we do
SmartCut develops a grassland stand development model that allows farmers to determine the optimal mowing time while taking multiple influencing factors into account:
- Field data: NIR sensor data from forage harvesters and fresh-grass samples collected on experimental plots of seven participating dairy farms after every cut, analysed for protein, sugar, starch and crude fibre content.
- Environmental data: local weather stations providing temperature sums and soil moisture, weather forecasts, soil information, elevation profiles and fertilisation practices.
- Earth observation: satellite data including NDVI time series providing up-to-date field information.
- Practical constraints: field accessibility (bearing capacity) and implications for subsequent cuts, biodiversity and plant communities.
The model optimises dynamic yield and quality components such as dry matter yield, energy and protein content under real weather and conservation conditions, and also provides forecasts for subsequent cuts. Predictions are delivered region-specifically through the GeoBox infrastructure and evaluated by the participating farms — as a decision-support tool for more sustainable grassland management and milk production.

Satellite-based monitoring of grassland biomass development across the project’s experimental fields from March until the first cut in May 2025. Basemap: © OpenStreetMap contributors © CARTO.
Consortium
The operational group brings together research, advisory services, industry and practice:
- Deutsches Forschungszentrum für Künstliche Intelligenz (DFKI) — coordinator
- Ferdinand-Steinbeis-Gesellschaft für transferorientierte Forschung gGmbH (FSG)
- Lehr- und Versuchsanstalt für Viehhaltung, Hofgut Neumühle
- Dienstleistungszentrum Ländlicher Raum Eifel (DLR Eifel)
- Dienstleistungszentrum Ländlicher Raum Rheinhessen-Nahe-Hunsrück (DLR RNH)
- Milchviehberatungsring Wittlich-Trier e.V.
- Pessl Instruments GmbH
- Seven dairy farms specialising in milk production, grassland and forage cropping
Associated partners: Förderverein Digital Farming e.V., Fraunhofer Institute for Computer Graphics Research IGD, Hochwald Foods GmbH and John Deere GmbH & Co. KG.
SmartCut in the media
German public broadcaster ZDF reported on SmartCut and how AI can improve milk production — featuring project farmer Tobias Zarth (Wahlerhof), Johannes Steinfort (Hofgut Neumühle), and DFKI director Prof. Andreas Dengel:
Further information
Project team

Marcela Charfuelan
Sr. Researcher

Miro Miranda
Communication & Consulting

Deepak Pathak
Researcher