4.1.4 — Maritime Intelligence — maturity: live
Vessel Behaviour Analytics
Detecting anomalous, illicit or commercially significant vessel behaviour from AIS time-series and supporting data — meeting points, loitering, port avoidance, identity manipulation.
When AIS positions alone no longer tell the full story, pattern-of-life analytics applied to satellite-derived vessel tracks expose the intent behind the movement — at national-fleet scale, around the clock.
Once you can see every vessel, the next question is what each one is doing. Vessel behaviour analytics applies pattern recognition to AIS time-series and contextual data to flag activity that an operator should look at: a tanker meeting another tanker for hours in the middle of nowhere (likely a sanctions-evading ship-to-ship transfer); a fishing vessel loitering inside a marine protected area; a cargo ship turning off its AIS as it crosses into territorial waters; a vessel changing its declared identity mid-voyage. The category includes EMSA's STAR Automated Behaviour Monitoring service, commercial offerings from Windward, Pole Star, Lloyd's List Intelligence, Kpler and others, and increasingly large-language-model-based reasoning over multi-source vessel records.
The economic users are insurance underwriters (sanctions risk, route risk, claims fraud), commodity traders (oil-on-water inventory, freight rates), banks doing trade finance compliance, and government agencies (sanctions enforcement, customs, fisheries). For middle-power states the application is doubly valuable — it surfaces the suspicious activity their own ISR cannot cover, and it produces evidence packages that hold up in international forums.
Frequently asked
What is vessel behaviour analytics and how is it different from simple AIS tracking?
AIS tracking tells you where a ship is and how fast it is moving. Vessel behaviour analytics ingests that positional stream alongside SAR detections, RF emissions, optical imagery and historical patterns to classify what the ship is doing — loitering, transhipment, boundary-crossing, identity-switching — and flag deviations from expected behaviour. The output is an intent signal, not just a location.
Why should a coastal state own this capability rather than subscribe to a commercial maritime-intelligence platform?
Commercial platforms such as Windward or MarineTraffic offer excellent global products, but they are built for global customers, not for the specific EEZ boundaries, fishing agreements, port-state priorities and sanctioned-entity lists of any single nation. A sovereign system can be tuned to domestic law, integrated with coast-guard command systems without data-sharing obligations, and kept operational when a vendor withdraws service or raises prices. Critically, the underlying satellite data — the raw AIS and imagery — remains under national control.
How many satellites does a nation actually need to achieve useful coverage of its EEZ?
For a mid-sized EEZ (500,000–2,000,000 km²) a constellation of 6–12 VHF-AIS nanosatellites in a sun-synchronous LEO orbit at 500–550 km provides mean revisit of under 90 minutes per point. Augmenting with two or three SAR microsatellites brings sub-hourly cueing for specific high-priority zones. Nations can begin with as few as three satellites and a ground-station partnership, scaling incrementally.
Can vessel behaviour analytics detect ships that have switched off their AIS transponders?
Not through AIS alone — that is precisely the dark-vessel problem. A behaviour analytics platform cross-cues AIS absence against SAR-detected radar returns and RF-signal clusters from providers such as HawkEye 360 to infer that a vessel is deliberately dark. The behaviour history of that vessel before it went dark (speed, heading, time of day, proximity to known transhipment zones) then sharpens the risk classification significantly.
What data latency is acceptable for fisheries enforcement versus sanctions monitoring?
Fisheries patrol tasking can tolerate 2–6 hour latency for initial cueing if aircraft or surface assets are hours away anyway. Sanctions evasion monitoring — where a vessel may complete a ship-to-ship transfer of oil in under four hours — ideally requires sub-60-minute alert latency, which demands either a large LEO constellation or priority tasking agreements with commercial SAR operators such as ICEYE or Capella.
What international legal framework governs action taken on the basis of satellite behaviour analytics?
UNCLOS Articles 56, 73 and 110 define the coastal-state right to board, inspect and seize vessels in the EEZ for fisheries violations and to exercise hot pursuit. Satellite-derived evidence is increasingly accepted in flag-state proceedings and regional fisheries management organisations, but chain-of-custody documentation and sensor metadata conforming to ISO 19115 standards are usually required for evidence to be admissible in port-state or flag-state courts.
How does a country prevent misuse of vessel behaviour data — for example, commercial intelligence gathering on competitor fleets?
Governance frameworks — aligned with IMO MSC.428(98) on cyber risk and national data-classification law — should distinguish between maritime safety data (open), enforcement intelligence (restricted) and strategic economic data (classified). Many nations model their approach on the EU's CISE (Common Information Sharing Environment) framework, which separates operational layers by access tier. A sovereign system makes this tiering technically enforceable rather than contractually dependent on a vendor's policy.
Is this technology mature enough to deploy today, or is it still experimental?
The core capability — satellite AIS combined with ML-based anomaly detection — is fully operational. Global Fishing Watch has run production systems since 2016; Windward, Pole Star and Anagog-derived platforms are used by navies and coast guards across Europe, the Gulf and South-East Asia. The frontier research sits in multi-modal fusion (combining AIS, SAR, optical and RF in a single latency-optimised pipeline) and in explainable AI for legal evidentiary standards — both areas where a sovereign R&D programme can make meaningful contributions.