Published on 27.04.2026
NEW on CONNECT: Estimating Soil Properties from Satellite Data with Explainable AI
Image: Rıdvan Salih Kuzu; from HI Project "HYPER-AMPLIFAI" | info
With Hype-Spectral, a top-performing solution in the HYPERVIEW2 ESA Φ-lab Challenges, Aaron Banze from German Aerospace Center (DLR) shows how machine learning and hyperspectral imaging can accurately estimate key soil nutrients, non-invasively and at scale.
The approach integrates explainable AI (XAI), making model predictions more transparent and trustworthy.