PROTECTED B
IDEaS Challenge 13SPARC Global · DefenseScope
CMMI · ISO 9001 · ISO 27001Sign in

FUSION-SAT · DefenseScope

Multi-modal Geospatial AI for Explainable Arctic & Multi-Domain Situational Awareness

Seven verified ingestion streams lift through cross-modal attention and a dynamic knowledge graph into maritime, land, and air watch floors — honest readiness on every path.

Sign in to consoleDark vessel workspace
FUSION-SAT multi-domain situational awareness overview

500+

Projects delivered

30 yrs

GIS & GeoINT experience

10M ha

Monitored landscapes

Seven core data modalities

§8.1 streams — scroll on narrow viewports; 2×4 grid on large screens.

1OP-POC

Sentinel-1 SAR

Copernicus CDSE

All-weather radar for vessel footprints, ice, and change.

2OP-POC

Sentinel-2 Optical

Copernicus CDSE

High-resolution visual detail in clear daylight.

3R&D

Landsat-8/9 TIRS

USGS / AWS Open Data

Thermal heat blooms from propulsion or structural distress.

4R&D

Tactical Text

SITREPs & patrol logs

Structured ingest of patrol reports and maritime intelligence.

5OP-POC

AIS & RF Telemetry

AISStream · RadioML

Live kinematics plus R&D RF spoofing classifiers.

6R&D

Sonar & Acoustic

NOAA NCEI WCSD

Water-column sonar and multibeam bathymetry for subsurface.

7OP-POC

Arctic Buoys (IABP)

NOAA PolarWatch ERDDAP

In-situ ice drift, pressure, and SST for SAR datum context.

The Fusion Lift

Atlas ingestion lifts seven modality streams through cross-modal attention (CAT) and a Neo4j knowledge graph into operator-ready mission outputs.

Fusion Lift engine diagram — seven modality streams through CAT and Neo4j to mission outputs

Cross-modal Attention Transformer (CAT)

R&D

Combines acoustic/RF bearing lines with SAR detections to pinpoint assets without active communications. Engine scaffold exists in TerraScope-AI — trained weights pending (TRL-5 R&D).

Neo4j Dynamic Knowledge Graph

R&D

Retains vessel identities and bridges temporal gaps when AIS drops out. Schema and adapters are code-complete — blocked on live Neo4j 5 instance.

Implementation mapping matrix

Modality coverage across UC-1–UC-6 and domain extensions (master scoping §8.2)

FUSION-SAT implementation mapping matrix across use cases and modality streams

Interactive use-case explorer

Mission brief · inputs · outputs · visual — auto-advances every 8s; arrow keys supported.

sea · UC-1R&D

Dark Vessel Detection

AOI · Lancaster Sound choke point

Inputs

Modality streams feeding UC-1 (§8.2 matrix)

  • S1 SAR
  • S2 Optical
  • Landsat TIR
  • Tactical Text
  • AIS / RF
  • RadioML
  • Arctic Buoys

Outputs

  • Live AIS tracks + density
  • SAR detections
  • Dark-vessel alerts
  • Contact dossier + triage/export

Watch a choke-point AOI for vessels transiting with AIS off ("dark") by fusing live AIS with Sentinel-1 SAR detections through a space-time gate.

Open Dark Vessel Detection workspace →

Dark Vessel Detection mission visual

Dark Vessel Detection mission infographic

M1 · TRL-5 — semi-integrated components validated in simulated environment.

M4 · TRL-6 target — integrated prototype in simulated operational environment.