SAR image analysis
Reading Synthetic Aperture Radar imagery of marine surfaces.
02 / Environmental intelligence
FOREARTH
Intelligence for understanding environmental systems at scale.
Software · AI · Research · Hardware
The problem
Scale
MILLIONS OFSQUARE KILOMETRES.
Marine environments cannot be watched by hand. We explore how computer vision on satellite radar can surface the patterns worth a human looking at.
What we build
06 capabilities
Environmental preprocessing
Cleaning and normalising raw radar returns before inference.
Computer vision inference
Models that look for patterns rather than single pixels.
Deep learning classification
Separating oil-like signatures from the noise around them.
Oil / non-oil separation
A binary call that has to survive real ocean conditions.
False-positive analysis
Because a look-alike is not the same as a spill.
Featured build
Enigma Devs / 2026
OIL SPILL INTELLIGENCE
A computer vision pipeline exploring how Synthetic Aperture Radar imagery can help identify and categorize patterns associated with oil spills across marine environments.
Capabilities
- SAR image analysis
- Oil / non-oil classification
- Environmental preprocessing
- Computer vision inference
- Deep learning classification
- False-positive analysis
The pipeline
Image → category
FROM ORBITTO ANSWER.
01
Satellite image
02
Preprocessing
03
AI model
04
Detection
05
Categorization
02
Environmental intelligence
06
Capabilities
05
Stage pipeline
Enigma Devs / 2026
Building technology with a reason.