Core Technology
Three sensors. One cognitive picture of the High North.
RD0CT0BR is built around a fusion engine that turns independent, noisy modalities into a single, queryable maritime ground truth.
SARonline
All-weather radar detection
Spaceborne synthetic aperture radar tiles ingested continuously and segmented for vessel signatures — independent of cloud, darkness, or sea state.
- Resolution
- 3 m / pixel
- Revisit
- ≤ 6 h
- False-positive
- < 1.8%
AISonline
Transponder telemetry
Multi-source AIS aggregation with anomaly scoring for dropouts, spoofing, identity collisions, and implausible kinematics.
- Tracks
- 12.4k active
- Latency
- < 8 s
- Spoof recall
- 0.92
OSINTonline
Open-source extraction
Multilingual NER over news, port logs, social and forums — every mention vessel-anchored, timestamped, and source-graded.
- Languages
- 11
- Sources
- 1.9k
- Daily items
- 32k
Fusion Engine
Rule logic + ML correlation
A deterministic core scores cross-modal consistency; a learned model elevates long-tail patterns. Operators see why every alert exists.
| Trigger | Inference | Severity |
|---|---|---|
| SAR contact ∧ ¬AIS | Possible dark vessel | HIGH |
| AIS gap > 20 min ∧ in EEZ | Transponder evasion | HIGH |
| Δcourse > 90° ∧ low SOG | Loiter / rendezvous | MED |
| OSINT match ∧ port mismatch | Identity inconsistency | MED |
| MMSI collision | Spoofing candidate | LOW |
Compute
Edge-deployable inference, GPU-optional
Sovereignty
Air-gap compatible, NATO-aligned classification
Open architecture
Modality plug-ins via documented schemas