About the project
A major challenge in Norwegian forage production is the unpredictable variability of protein levels in perennial ryegrass (Lolium perenne). Currently, farmers lack tools to identify the optimal harvest window, leading to reliance on imported concentrates.
ProteinSense addresses this by developing an AI-driven "harvest alert" system. A core element of the project is rigorous testing across different nitrogen fertilization levels to map the specific correlation between nitrogen supply and protein concentration. By combining this agronomic data with drone imaging (multispectral/hyperspectral) and handheld NIRS, the project will create robust models to estimate protein in standing grass. This enables farmers to optimize harvest timing, improving nitrogen use efficiency (NUE) and boosting national self-sufficiency.
Objectives
The primary objective of the ProteinSense project is to establish a proof-of-concept for predicting protein levels in ryegrass. This will enable farmers to harvest grass with higher protein content, allowing the value chain to reduce its dependency on imported protein concentrates.
Participants
NMBU participants
Odd Arne Rognli
Professor Emeritus
WP leader
External participants
- YARA : Anders Rognlien
TINE: Petter Klette