Our project, carried out by PRIOT Digital Systems under the auspices of the European Space Agency (ESA), explores how deep learning and high-resolution Earth Observation (EO) data can detect European spruce bark beetle (Ips typographus) outbreaks at the earliest 'green attack' stage.

Satellite Bark Beetle Monitoring Interface
AI-driven spectral analysis identifying early tree stress before visual canopy discoloration.

The Green Attack Challenge

Traditional forestry inspection relies on observing crown discoloration (red and grey attack phases). Unfortunately, by the time foliage turns brown, adult beetles have already reproduced and spread to adjacent healthy stands.

By leveraging multispectral satellite bands (Sentinel-2 and commercial sub-meter constellations), our deep learning architectures detect subtle changes in chlorophyll absorption and water canopy indices weeks before human visual perception.

Architectural Stack & Pipeline

Python / PyTorch Sentinel-2 Multispectral U-Net Segmentation FastAPI Geo-Service Mapbox GL JS

Foresters and emergency response units can now monitor thousands of hectares with weekly risk maps, prioritizing ground scouting and sanitary felling with pinpoint accuracy.