How We Detect Launches
We monitor weather satellites that capture thermal imagery of the entire planet every 10 minutes. When a rocket launches, it produces a heat signature we can see from space. Our system automatically scans every known launch site worldwide and uses machine learning to identify real launches from false alarms like wildfires or industrial activity.
Tracking The Flight Path
Once we detect a launch, we simulate where the rocket is going using physics models. We calculate when rocket stages separate, predict the final orbit, and visualize the ground track. This works for any type of mission - whether it's heading to low Earth orbit, a sun-synchronous path, or a high geostationary transfer.
Getting The Data
All detection data streams in real-time through our API. We output in standard formats that work with mapping software and command-and-control systems. You can integrate directly into existing workflows or use our dashboard.
Machine Learning
Our models analyze satellite imagery to distinguish launch events from other heat sources. We extract dozens of characteristics from each thermal signature - brightness, shape, movement patterns - and classify them instantly. The system runs in the cloud for speed.
Verification
After detecting a launch, we confirm it by checking subsequent satellite images along the predicted flight path. We look for heat signatures at expected stage separation points. Each detection gets a verdict: confirmed, partially confirmed, trajectory deviation detected, or insufficient data.