SoilDirect

Core digital services: Spatialise develops AI-powered satellite soil analysis technology that converts Sentinel-2 multispectral satellite imagery into high-resolution soil health maps. Our proprietary deep learning model, trained on over 350,000 ground-truth soil samples in collaboration with Wageningen University & Research (WUR) and ISRIC — World Soil Information, predicts soil organic carbon (SOC), nitrogen (N), phosphorus (P), potassium (K), and pH at 10x10m spatial resolution with >85% accuracy. We deliver these insights through a cloud-based platform with API integration capabilities.

Application to the agrifood sector: Our technology enables farmers and agricultural advisors to monitor soil health continuously and affordably — without physical soil sampling. This supports precision fertilisation (reducing input costs by 15-25%), compliance with upcoming EU soil monitoring regulations, and evidence-based regenerative farming practices. Our model was rated best-in-class in a Rabobank Global RfP for soil carbon prediction, and has been commercially validated through contracts with Province of Drenthe and Rabo Carbon Bank. We have successfully completed two MIT feasibility projects funded by Province of Zuid-Holland (SOCMO 2023, NPK Monitor 2025).

Type of agrifood partner we are looking for: We seek an agrifood SME with hands-on agronomic expertise and established farmer networks who can translate our satellite-derived soil data into actionable farm-level recommendations. Ideally, a partner experienced in soil management advisory, crop optimisation, and farmer training — bridging the gap between digital soil intelligence and practical agricultural decision-making.