Challenge
Vineyard managers across the Iberian Peninsula needed real-time, data-driven monitoring of vine maturity, disease, and yield anomalies without relying on continuous cloud connectivity.
Solution
Our embedded systems and AI team designed and implemented an on-device inference pipeline, from model optimization for CubeSat-class hardware to multispectral data preprocessing and object-detection workflows, as part of the EU Horizon Europe AGRARIAN programme.
Results
- Selected as an Open Call winner under EU Horizon Europe's AGRARIAN programme
- Near-real-time vine maturity estimation, disease detection, and yield-anomaly identification
- Runs on constrained, satellite-grade edge hardware without continuous cloud connectivity
Technology
Python · Edge AI · Multispectral Imaging · NDVI · Kubernetes · CubeSat Emulation
Powered by DAC Digital