Field-health maps from drone and satellite
A precision-agriculture cloud platform: NDVI crop-health maps from drone and satellite, geospatial data in PostGIS. The system is in production and the customer works in it.
Fields judged by eye.
Crop condition was hard to assess without walking the field. Drone and satellite data both existed, but sat apart and nobody combined them into a usable picture.
Fuse drone and satellite into one map.
We built a platform that combines aerial drone imagery with Copernicus satellite data and computes NDVI maps of crop health. The index computation used to be implemented twice and the numbers disagreed, so we split it out into a dedicated Java analytics library as the single source of truth; a GDAL service classifies the raster, which is uploaded straight to object storage through a signed URL. PostGIS holds the geospatial data, and drone licensing pairs over a QR code, with an audit trail of every connection attempt carrying correlation IDs.