Vessel movement predictions you can act on
AI that learns how vessels really navigate each waterway, replacing assumptions with observed behaviour
Problem
Vessel movement is hard to predict accurately
Vessels don’t move in straight lines. Routes bend around traffic, tides, and local practice,
and these patterns differ by vessel type and by waterway. Yet planning today often relies on
subjective judgement: rules of thumb, assumed routes, and the experience of individual operators.
That makes it inconsistent, hard to scale, and difficult to justify to regulators, stakeholders,
or consenting bodies.
Port scheduling, traffic management, and offshore planning all depend
on knowing where vessels will actually go.
Solution
Predictions grounded in observed behaviour
We train AI models on AIS data to learn how vessels move - each model
is specialised by waterway, vessel type, and operating context, so predictions reflect real
behaviour rather than generic or subjective routes.
Our models predict:
-
Trajectory - the most likely route the vessel will take to the destination.
-
Positions - synthetic AIS positions for the trajectory with time, speed, course, and heading attributes.
-
ETA - the estimated arrival time with confidence bounds.
Required inputs:
-
Vessel characteristics - vessel type, dimensions, and latest speed over ground.
-
Transit information - live or historical vessel position and a destination, supplied as simple coordinates.
AIS-Driven Objectivity
Real vessel behaviour, across vessel types and operating contexts, learned directly from carefully processed AIS data.
Regional Intelligence
Traffic patterns, navigational constraints, and local operating practices captured and reproduced in great detail.
Regularly Updated
Rapidly retrained and redeployed to stay aligned with changing behaviour and constraints.
This AIS-driven approach allows us to produce predictions that are evidence-based, consistent, and specific to each waterway.
Services
Live predictions
Predict the movement and arrival time of vessels currently in transit using their latest
information, obtained from live AIS feeds such as
MarineTraffic.
Use cases:
-
Port operations - continuous predictions during final approach provide an accurate view of arriving vessels to support the confident planning of tugs, pilots, berths, and other valuable resources.
-
Vessel traffic services - continuous predictions for all vessels moving within an area provide enhanced situational awareness, allowing proactive routing and abnormal behaviour detection when vessels deviate from their expected route.
Access routes:
-
No-code - use the Viewer to generate and visualise individual predictions in your web browser, powered by our API.
-
Low-code - read the API Docs to learn how to easily integrate our API into existing systems.
Replay predictions
Replay historical vessel traffic in an altered navigational environment to see how vessels
would have behaved if the change already existed. Development areas are traditionally assessed
individually, but replay predictions make it possible to rapidly and objectively quantify
the cumulative impact of multiple projects on vessel traffic. This approach can also be
used to augment or test subjective vessel traffic impact assessments using a purely
data-driven methodology.
Use cases:
-
Navigational risk - feed synthetic AIS data into your existing tools, just like real data, to identify hazards such as route deviations, increased encounter rates, and congestion.
-
Marine spatial planning - test and refine layouts by quantifying how candidate development areas would change vessel traffic, before committing to a design.
-
Cumulative impact assessment - assess multiple projects together, including as they are built out in phases, rather than one development at a time.
Access routes:
-
Custom - we work with you on your project, taking your proposed development areas and returning a synthetic AIS dataset with an impact analysis, ready for your existing GIS and risk tools.
Models
Validated against real transits
Every model is tested against thousands of real vessel transits before deployment, with
results aggregated across each region for reliable evaluation. The spatial extent and
performance metrics for each live model are published, so you can see exactly where a model
applies and how well it performs before you rely on it.
Each model is benchmarked on two tasks:
-
Trajectory reconstruction - how accurately the model recreates vessel transits over significant distances, using only the start and end coordinates of the voyage.
-
ETA prediction - how accurately the model estimates arrival time, benchmarked against a historical baseline for the same transit.
Europe
Celtic Sea
LIVEEnglish Channel
LIVEIrish Sea
LIVENorth Sea (NW)
LIVENorth Sea (S)
LIVE
About
Rigorous by training, impactful by experience
Nautalyx brings together advanced machine learning and specialist expertise in maritime AIS data processing, backed by extensive experience in developing novel scientific research and building production-ready systems that are robust and impactful.