Skip to content

< 2 days · > 1,000 km² · AI · Remote sensing

Satellite data pilot

From brief to prioritised list in under two days – AI screening, blind review and chain of evidence.

Abb. 01Tile grid: AI screening an area tile by tile (synthetic visualisation)
from brief to finished, prioritised list

< 2 days

from brief to finished, prioritised list

of satellite data analysed

> 1,000 km²

of satellite data analysed

image tiles screened in 12 minutes

14,600

image tiles screened in 12 minutes

01Starting point

The client’s question: can a large study area be sensibly pre-screened from satellite data – and how quickly is a reliable result available?

02Approach

  1. 01Split the area into uniform tiles and enriched them with context data (terrain, land cover).
  2. 02AI screening of every tile with fixed criteria and a reasoned rating per candidate.
  3. 03Blind review: the reviewer knew neither location nor expectation; rejections were reasoned and cross-checked.
  4. 04Handed over a prioritised list with a chain of evidence for each entry.

03Outcome

Within two days there was a short, reasoned list. By hand – at 20 seconds per tile – the screening alone would have taken around ten working days.

Client and region are deliberately anonymised, figures rounded; no images, locations or hit rates are shown.

  • Python
  • Multi-Agenten-Workflow
  • Satellitendaten
  • Blindprüfung

A similar task on your desk?

In the intro call we tell you how we would approach it – and how quickly you would see results.