GASS Advances Integrated Weather Routing Technology

As part of the Green AI for Sustainable Shipping (GASS) project, Deliverable 3.2 System Integration, NAVTOR and Simula have reached an important milestone in the development of integrated Weather Routing and Optimisation Services (WROS).

The work under Work Package 3 has successfully brought together key technologies required for end-to-end voyage optimisation within NAVTOR's advisory framework. The achievement builds on earlier prototype work and demonstrates how vessel performance models, weather information, optimisation algorithms, and voyage planning tools can be combined into a unified technology framework.

Header image: Screenshot from NavStation

From Research to Integrated Technology

The first end-to-end prototype, connecting vessel data and performance modelling with route optimisation, was completed in 2025. Since then, the focus has been on integrating the various components into NAVTOR's architecture, creating a foundation for future operational services.

Today, NAVTOR has solutions available for all major subsystems involved in the workflow, from vessel data handling and performance modelling to route optimisation and voyage planning. While further work remains before a fully integrated customer-facing solution is available, the project has established a strong technological foundation for future deployment.

Fast and Practical Optimisation

At the core of the solution is a genetic algorithm that evaluates and improves route alternatives.

Unlike traditional optimisation methods that can require extensive computation time, the approach can generate viable route suggestions quickly and continue refining them over time. Recent testing indicates that high-quality weather-optimised routes can typically be produced in around two minutes, making the technology increasingly relevant for operational decision support.

Rather than searching endlessly for a mathematically perfect solution, the focus is on delivering practical and highly effective route recommendations within a timeframe suitable for real-world maritime operations.

Combining Maritime Expertise and AI

The current WROS implementation uses a vessel performance model developed by NAVTOR and primarily trained using noon report data. In parallel, the GASS project has explored how high-frequency sensor data can further improve vessel performance predictions.

A key contribution from Simula has been research into machine learning methods for vessel performance modelling using sensor data. The research has also investigated transfer learning techniques, which may enable future models to adapt more efficiently across vessel types and operating conditions. While transfer learning is not yet part of the operational solution, the work has provided valuable insights for future development.

Supporting More Sustainable Shipping

By combining optimisation algorithms, vessel performance modelling, weather information, and AI-driven research, GASS is helping pave the way for more efficient voyage planning. Better routing decisions can contribute to reduced fuel consumption, lower emissions, and improved operational efficiency.

The completion of this integration milestone represents another important step towards transforming advanced research results into practical tools that can support the maritime industry's transition towards greener and smarter operations.