19. sep. 2025
Deliverable D4.1 Report on test scenarios and benchmark trajectories is an important output in the GASS project’s work to make shipping greener and smarter.
The GASS project develops AI-driven solutions to cut fuel use and emissions in global shipping. To succeed, we need to validate the concepts and ensure that our technology works under real operational conditions. That’s where task 4.1 comes in.
"Task 4.1 – Deployment and Validation Setup is about preparation. It defines how we select vessels, design test scenarios, and build benchmark trajectories. Together, these steps form the backbone for validating the GASS solution in practice, making sure that later trials and demonstrations rest on a solid and realistic foundation," says Marianne Bjelland, Sustainable Energy.
The result of this work is documented in Deliverable 4.1, which was delivered in June 2025 as planned and is now ready to guide the next phase of the project.
Why benchmarks matter
"Deliverable 4.1 is more than a report. It is the foundation for all validation activities in Work Package 4. Without a clear setup of test scenarios, and benchmark trajectories, it would not be possible to measure the true impact of AI-enhanced voyage optimisation," says Khanh Quang Bui, NAVTOR.
In Task 4.1, benchmark trajectories are defined to represent real industry practices. These serve as reference points against which AI-optimised routes can be compared. The GASS project integrates data from multiple sources, a Digital Twin (DT), and a Weather Routing Optimisation Service (WROS) powered by a genetic algorithm. This enables vessel-specific, predictive voyage planning under realistic operational conditions, providing a scientific basis for evaluating improvements in fuel consumption, emissions, waiting times, and total voyage costs.
What the testing framework looks like
In Deliverable 4.1 we defined the framework for how the GASS solution will be validated. At the core of this framework are three reference points:
Baseline routes – the shortest navigable (great-circle routes) without optimisation.
Benchmark trajectories – routes that reflect common industry practice, often provided by existing weather routing services.
AI-optimised routes – generated by the DT coupled with the WROS and compared against both the baseline and benchmark trajectories.
Testing begins in a simulation environment using digital twins of vessels and hindcast data. This setup enables the evaluation of performance under realistic conditions while controlling key variables, such as wind, waves, swell, currents, and safety constraints.
So far, one hindcast simulation has been carried out on a trans-Pacific voyage from Vancouver to Yokohama, applying a 4-metre wave height limit. This first test provided valuable insight into how the methodology works in practice and marked Milestone 3 of the project.
This framework now serves as the basis for the upcoming trials at sea, where the same approach will be applied to evaluate the system on board real vessels.
Next steps
With Deliverable 4.1 completed and Milestone 3 achieved, the GASS project is ready to move from simulations to real-world testing. The next phase, Task 4.2, focuses on demonstrators and impact assessment.
Seven vessels with diverse operational profiles will be equipped with the GASS solution and tested across different trading routes and seasonal conditions. Each voyage will be evaluated against baseline routes and benchmark trajectories, providing a clear picture of how much can be gained in terms of fuel efficiency, emissions reduction and voyage costs.
The validation will be guided by defined KPIs, targeting:
20% reduction in fuel consumption
20% reduction in GHG emissions
30% reduction in waiting times for tramp operations
10% reduction in voyage costs on average
Deliverable 4.1 has set the stage for validation in the GASS project. By defining vessels, scenarios and benchmark trajectories, the project now has a clear framework for testing AI-driven voyage optimisation. With this groundwork in place, the upcoming demonstrators will provide real evidence of how GASS can make shipping both greener and more competitive.
The GASS-project is made possible by The Green Platform Initiative (“Grønn Plattform”) funding scheme by Norges forskningsråd, Innovation Norway and Siva SF under the Grant Agreement No. 346603