Project aims to prove AI can cut fuel use and emissions
The initiative set out by four maritime players seeks hard evidence that artificial‑intelligence (AI) route optimisation can deliver measurable reductions in bunker consumption and greenhouse‑gas (GHG) output when a ship is operating under everyday conditions. The consortium describes the effort as “an objective method for measuring fuel and emissions savings from AI‑based voyage optimisation, taking into account factors such as weather, sea conditions and vessel operations.” This wording appears in both the Ship & Bunker release and the Hellenic Shipping News brief.
Consortium composition and each partner’s remit
At the centre of the trial is an Evergreen Marine Corporation liner that will be fitted with Samsung Heavy Industries’ autonomous navigation suite, known as the Samsung Autonomous Ship (SAS). Samsung Heavy Industries will run the speed‑optimisation algorithm and provide real‑time vessel control through SAS, while Weathernews Inc. supplies both baseline routes and the meteorological‑oceanographic data required for accurate forecasting.
ClassNK, the Japanese classification society, has agreed to review the verification methodology, carry out an independent evaluation of the results and eventually publish a Statement of Fact that records the findings. The Ship & Bunker story notes that ClassNK will “issue a Statement of Fact based on the findings.” Hellenic Shipping News adds that the society’s contribution is intended to “enhance the objectivity and credibility of the findings through technical review of the verification methodology.”
The four entities formalised their collaboration by signing a memorandum of understanding (MoU) at Evergreen’s headquarters. While the exact signing date is not listed, both sources place the agreement in August 2026, aligning with the publication dates of the announcements.
Verification methodology and data flow
Operational data – including speed, engine output and fuel consumption – will be captured directly from the Evergreen vessel during normal voyages. Weathernews will overlay this information with real‑time weather and ocean conditions, as well as historical datasets that help define a “baseline” route against which AI‑generated plans can be compared.
Samsung Heavy Industries’ SAS system will execute the AI‑derived speed profile, continuously adjusting the vessel’s propulsion to match the optimal plan. The Ship & Bunker article stresses that the project will “analyse operational and fuel consumption data from the vessel to assess the system’s performance in real operating conditions.” The analysis will be carried out by a joint technical team, after which ClassNK will independently verify the calculations and confirm whether the AI tool achieved the projected savings.
Upon completion of the data‑review cycle, ClassNK intends to produce an objective Statement of Fact that details the verified fuel reduction percentage, associated GHG abatement and any caveats arising from weather variability or operational constraints. Both primary sources underline that this independent certification is a core deliverable of the partnership.
Regulatory backdrop and related class society work
ClassNK’s involvement in the AI verification project coincides with the recent release of its updated Guidelines for Ships Using Alternative Fuels, Edition 3.1. Maritime Gateway reports that the new edition expands safety requirements for vessels running on hydrogen, methanol and ethanol, signalling the society’s broader push toward greener ship operations.
The timing suggests a strategic alignment: while the AI‑optimisation trial focuses on immediate fuel savings through smarter routing, the alternative‑fuel guidelines address longer‑term decarbonisation pathways. By participating in both initiatives, ClassNK positions itself as a facilitator of incremental efficiency gains and as a regulator for emerging low‑carbon fuels.
Industry observers note that the combined effort of a classification society, a shipowner, an equipment supplier and a weather data specialist creates a rare “end‑to‑end” validation chain. Should the trial confirm significant savings, the methodology could become a benchmark for future class‑society verification programmes and may influence regulatory bodies that are drafting performance‑based standards for AI‑driven navigation tools.
What this means for operators
For ship operators, the project's outcome could translate into a clear, third‑party certified route‑optimisation product that can be integrated into daily planning without fear of non‑compliance. A verified fuel‑saving percentage would allow owners to justify higher upfront costs for AI software, negotiate better charter rates based on lower operating expenses, and potentially claim carbon credits or meet emissions caps under existing regulatory regimes. Moreover, the collaboration demonstrates that robust data collection – from engine performance to weather inputs – is essential for any credible efficiency programme, prompting operators to invest in integrated monitoring systems that feed into such verification frameworks.