Data & methods
The conclusion cannot be stronger than the retrieved evidence
MarineAware studies how cooperative AIS, imagery, weather, reporting, and reference data can support bounded maritime analysis. This page distinguishes the wider research catalogue from sources actually retrieved in a specific public build. A configured or commercially available source is not evidence present in an assessment.
Research source catalogue
This table maps source classes that may be relevant to maritime research. It does not claim that MarineAware licenses, retrieves, validates, or uses every listed source in the current public demonstrator. Each output must name only the sources successfully present in that build.
| Source | Type | Access | Detail | Primary use |
|---|---|---|---|---|
| Satellite & terrestrial AIS | Vessel tracking | Commercial | Kpler AIS, Spire, S&P/ORBCOMM; T-AIS near-instant, enhanced S-AIS ~6 min | Live positions, port calls, dark-activity |
| NOAA MarineCadastre AIS | Terrestrial AIS | Open | US coastal historical archives | Free historical traffic analysis |
| Copernicus Sentinel-1 (SAR) | C-band radar | Open | ~5–20 m, all-weather day/night; 1C/1D carry onboard AIS | Non-cooperative / dark-ship detection |
| Copernicus Sentinel-2 (optical) | Multispectral | Open | 10 m, 5-day revisit, cloud-limited | Optical vessel & coastal detection |
| ICEYE / Umbra / Planet | Commercial imagery | Commercial | SAR to ~25 cm; PlanetScope ~3 m daily | Confirm & classify detections |
| Copernicus Marine (CMEMS) | Ocean physics | Open | Currents, SST, SSH, waves; ~8 km global to 500 m regional | Routing & fuel optimisation |
| NOAA GFS / WaveWatch III | Metocean forecast | Open | Wind ~13 km; waves ~50 km, to 5–16 days | Weather routing |
| ECMWF ERA5 / Open Data | Reanalysis / NWP | Freemium | ERA5 ~31 km hourly 1940→; open subset 0.25° | ML training & backtesting |
| GEBCO / EMODnet bathymetry | Bathymetry | Open | GEBCO 15 arc-sec; EMODnet ~115 m (EU) | Under-keel clearance, depth routing |
| Global Fishing Watch | Fishing / events | Open | AIS effort 2012–2024, events & vessels APIs, SAR layers | IUU, encounters, transshipment |
| UN Comtrade | Trade flows | Freemium | Bilateral HS trade, 1962→ | Demand-side flow context |
| EU MRV (THETIS-MRV) | Emissions | Open | Per-ship verified annual CO₂ & fuel, >5000 GT, 2018→ | Emissions ground truth & CII context |
| Equasis / Clarksons / S&P Sea-web | Registries | Freemium | Identity, ownership, PSC, ship particulars | Identity resolution & due diligence |
Latency and resolution figures are typical source or vendor descriptions and vary with coverage, traffic density, tasking, processing, and mode. Verify the official source and licence before use.
Methods under examination
These are research and engineering approaches, not claims that every method is implemented, validated, or production-ready in the public build. A specific result must disclose its implementation, inputs, evaluation, and failure states.
Trajectory modelling & anomaly detection
Learn normal behaviour per ship-type and route from AIS kinematics, then flag deviations — course/speed anomalies, gaps, loitering, impossible kinematics. Sequence models (LSTM/GRU, Transformers), reconstruction and WGAN-GP scoring, over unsupervised route baselines.
Dark-vessel & sanctions-evasion detection
The flagship capability: correlate AIS with SAR/optical detections — a radar hit with no matching AIS is a candidate dark ship — fused with gap, loitering and rendezvous reasoning. Anchored by the 2024 Nature result that ~75% of fishing vessels are untracked.
ETA & voyage / weather routing
ML ETA regressors on historical voyages plus physics-aware routing over gridded metocean (CMEMS, WW3, ECMWF) with bathymetry constraints and vessel fuel curves. Objective trades time vs fuel vs CII penalty.
Emissions estimation (CII / EEXI, MRV-grade)
Bottom-up activity model — AIS speed/time × vessel-specific power/fuel curves → fuel → CO₂ — calibrated against verified EU MRV totals so AIS-based estimates are credible to lenders and verifiers.
Demand & flow forecasting
Fuse port-call time series (AIS geofencing) with trade signals (UN Comtrade) and seasonality; classical and ML time-series, increasingly time-series foundation models for zero/few-shot forecasting across many ports.
Foundation models for geospatial & time-series
Fine-tune geospatial FMs (Prithvi-EO, TerraMind) for SAR/optical vessel detection, and apply time-series FMs (TimesFM, Chronos) to ETA and demand. Maritime-specific fine-tunes are R&D, not turnkey — we treat them accordingly.
LLM / agentic maritime-intelligence pipelines
Retrieval over vessel, registry and regulatory corpora; agents orchestrating AIS, SAR and weather APIs; natural-language alerting with cited evidence — explainable copilots in the spirit of research systems like AIS-LLM.
Identity resolution & network analysis
Resolve MMSI ↔ IMO ↔ beneficial owner across Equasis, S&P and Clarksons; graph analytics over ownership, management and co-occurrence to expose the shell structures behind sanctioned or IUU fleets.
Provenance by build
Publish build time, collection window, source successes and failures, formula version, and a stable evidence record. Do not infer retrieval from configuration.
Missing is not zero
No captured AIS position, reporting item, thermal value, or image does not establish real-world absence. Preserve missing, zero, stale, and failed as different states.
Abstention is a capability
When the available sources cannot answer the stated question, report insufficient evidence. A plausible narrative is not a substitute for observation.
Evidence states
Say what the build can support
Context only: weather or coarse imagery is present without relevant activity evidence.
Partial: at least one relevant activity source is present, with material gaps disclosed.
Corroborated: two suitably independent source types support the same bounded observation.
Insufficient: the available evidence cannot answer the stated question.
AI disclosure
Narration is not sensing
A language or vision model may turn already-computed evidence into a draft narrative. It does not become a sensor, set the deterministic score, or count and identify vessels from imagery unsuitable for that task.
Machine-generated language should be visibly labelled and reviewed before external use. A model statement is not an independent corroborating source.
Freshness
Current means refreshed
A daily or live page should show the build time, last attempted refresh, last successful refresh by source, and a prominent stale warning once its promised cadence is exceeded. An old successful value must not silently stand in for a current observation.
Corrections
Preserve the change
A material correction should record the original conclusion, revised conclusion, date, reason, and supporting evidence. Send a challenge with the exact page, wording, and source to contact@marineaware.com.
Guides & reference
Go deeper on the open-data foundations
The guides describe possible open-data foundations and their limits. A demonstrator page must still disclose which sources were successfully retrieved for that build.
Open datasets for MDA
A practical catalog of the free AIS, SAR, optical, metocean and registry data.
Maritime OSINT guide
The collection-to-assessment workflow and the signals that matter.
MDA glossary
Plain-English definitions of AIS, SAR, dark vessels, spoofing and more.
Open demonstrator ↗
Inspect public outputs with their build-specific data gaps and limitations.
Challenge a method or propose a bounded research case.
MarineAware welcomes research critique, methods discussions, seminar invitations, and bounded collaboration proposals.