Built a program to track logistical intelligence using satellite data [P]

Hey guys, I've been workin on something new to track logistical activity near military bases and other hubs. The core problem is that Google maps isn't updated that frequently even with sub meter res and other map providers such as maxar are costly for osint analysts.

But there's a solution. Drish detects moving vehicles on highways using Sentinel-2 satellite imagery.

The trick is physics. Sentinel-2 captures its red, green, and blue bands about 1 second apart.

Everything stationary looks normal. But a truck doing 80km/h shifts about 22 meters between those captures, which creates this very specific blue-green-red spectral smear across a few pixels. The tool finds those smears automatically, counts them, estimates speed and heading for each one, and builds volume trends over months.

It runs locally as a FastAPl app with a full browser dashboard. All open source. Uses the trained random forest model from the Fisser et al 2022 paper in Remote Sensing of Environment, which is the peer reviewed science behind the detection method

GitHub: https://github.com/sparkyniner/DRISH-X-Satellite-powered-freight-intelligence-