Features

We build region-specific AI models trained on real AIS data to learn the behaviour of different vessel types. This enables highly realistic forecasting of vessel movements between flexible locations, delivering unmatched precision and reliability.

Request a Free Trial

Usage

Generate single predictions through Pathfinder, our no-code, browser-based interface, or integrate directly with our low-code APIs:

• Trajectory - predict the most likely route a vessel will take to a destination
• ETA - estimate arrival times with data-driven confidence bounds
• Plan - generate fuel-efficient passage plans aligned with a target transit time

To find out more about our APIs, visit our Documentation.

Use Cases

Our region-specific AI models support any application that requires accurate and reliable vessel movement predictions within a defined area.

Port Operators

Optimise berth allocation and reduce disruption with continuously updated ETAs during final approach to a chosen location, such as a pilot boarding station or specific berth.

Vessel Tracking

Enhance situational awareness by integrating future trajectory predictions into tracking and monitoring systems.

Autonomous Vessels

Validate and benchmark navigation systems, or embed our models to support reliable routing in complex and constrained environments.

Models

Explore our deployed region-specific AI models, as well as those under development.

Request a Region if it is not listed below.

About

Different vessel types behave differently. Tankers, container ships, ferries, and tugs operate at different speeds and under different constraints. That’s why established routing systems separate models by vessel type and size - using a single model for all vessels leads to inaccurate results in practice.

The same principle applies to geography for the same reason. Every waterway has its own navigational patterns and practical constraints that shape how vessels move, including narrow channels, port approaches, traffic separation schemes, and local operating practices. Just as vessel type matters, so does location.

In reality, achieving global coverage requires models to simplify or generalise these regional details due to data, computational, and operational constraints. This reflects a fundamental modelling trade-off: prioritising coverage means sacrificing detail. Attempts to add complexity to mitigate this often produce models that are difficult to deploy, challenging to interpret, and time-consuming to update.

This trade-off becomes critical as vessels approach their destination or operate in constrained environments, where small errors have outsized operational consequences such as costly delays or safety risks. In these situations, regional factors dominate vessel decision-making - dictating timing, limiting routing options, and forcing real-world adjustments.

Global models struggle to capture this nuance. We don’t.

We build region-specific AI models trained on real vessel behaviour derived from AIS data, capturing local navigational patterns and constraints in great detail.

Our modular regional design delivers practical advantages that global architectures cannot match. We can retrain and deploy updates to individual regions in hours rather than weeks, responding rapidly to changing conditions, new regulations, or emerging patterns. Our models are computationally efficient, enabling deployment on edge devices when needed. Most importantly, our models remain transparent and interpretable.

We specialise our models not only by vessel type and operational context, but by geography, ensuring predictions are finely tuned to the reality of how vessels actually move.