Pick your jaw up off the floor, as NAVTOR’s Bjørn Åge Hjøllo explains how the Green AI for Sustainable Shipping (GASS) initiative promises to transform the ability of shipowners and operators to slash fuel consumption, emissions and OPEX, ushering in a smarter, greener, more connected maritime future.
Sorry, what did you say?
Bjørn Åge Hjøllo, Chief Sustainability Officer at NAVTOR, smiles in recognition of the obvious disbelief.
He tries again: “I said that this project has the potential to cut 1% of all global emissions.”
1% of all shipping emissions?
“No, 1% of total global emissions… for everything.”
Data driven decarbonisation
This may be the first, but it won’t be the last, time Hjøllo is met by someone that needs to reboot their brain while struggling to comes to terms with the ambition of the Norwegian government backed GASS research project.
Led by NAVTOR, the initiative is a partnership with Grieg Star, Maritime CleanTech, Scandinavian Reach Technologies (ScanReach), Simula Research Laboratory, SinOceanic Shipping, and Sustainable Energy/SIVA, with support from the Norwegian Research Council, Innovation Norway, and SIVA.
Over the course of the next three years, it aims to champion what Hjøllo calls a “data driven approach to decarbonization” enabling shipping companies to identify, analyse and address inefficient energy use on any vessel, in any location, in any weather conditions, in real-time.
Powered by machine learning algorithms, digital twin technology, and a constant stream of high-quality data, the end result will be, says Hjøllo, “a simple, powerful decision-making tool that allows users to maintain competitiveness, achieve regulatory compliance and, in short, unlock more sustainable shipping.”
Although he makes it sound ‘easy’ there’s a lot of hard work that needs to be done first.
And this is where the partnership model comes in.
As in any ambitious team, each GASS player has clearly defined roles.
NAVTOR is the world’s leading supplier of maritime technology for e-Navigation and performance solutions, with products and daily services on over 18,000 vessels in the world fleet. With expertise derived from building and continually developing an integrated digital ‘ecosystem’ connecting vessels, fleets and entire organisations it is well placed to design, integrate and eventually bring this type of innovation to market. Grieg Star, SinOceanic Shipping and Sustainable Energy/SIVA provide both invaluable domain expertise and the crucial test vessels required to build, run and constantly refine the solution for real-world operations.
ScanReach, meanwhile, offers a unique wireless IoT platform that connects sensors, equipment and systems across complex steel vessel environments to harvest the data needed for GASS’ high power processing engines. Simula’s role sees the team leveraging their renowned developer experience to create the machine learning and digital twin back-end to empower the front-end benefits. Finally, Maritime CleanTech is on hand to disseminate information and encourage interaction with another potential 150 partners through its future-focused industry cluster.
So, that’s the big picture, but we have to zoom in to see how this will work in reality.
Hjøllo is happy to ‘go granular’.
“At present, there are no systematic data‐driven solutions for improving energy efficiency onboard, and GASS aims to address that,” he says.
“We want to capture granular information from a very wide range of high frequency data that basically allows us to predict what a vessel’s fuel consumption should be, regardless of vessel type, location, weather and so on. That means integrating precision data from vessel operations and exact operating environments - combining MetOcean condition and forecast data, AIS data, and a whole range of reporting and performance data gathered in real-time, all the time, from sensors. That can span anything ranging from propellor information, to engine RPMs, navigational data, speed and so on.”
Once they have this building material, Hjøllo explains, it can be used to craft a digital twin of any vessel that, regardless of operational parameters, can be used to demonstrate optimal fuel consumption. If the ‘real world’ ship is failing to live up to its virtual sibling’s performance, then the data can be instantly analysed to find out why.
“Maybe there’s an issue with the trim of the vessel, or a problem with fouling, or perhaps the auxiliary engine has been used in a congested area and it’s still running when there’s no need,” he says.
“With much richer, real-time data than ever before, we can unlock up to the minute awareness and performance analytics that enable dynamic voyage optimization – as opposed to today’s ‘static’ standard – and allow onshore teams and onboard crews to address issues/deviations from plans as they actually happen.
That’s an incredibly powerful advantage to have, especially in the new regulatory reality of CII ratings and EU ETS, for example, and the associated costs and investments associated with compliance.
“This could be huge.”
Once extensive testing and validation has been completed, the AI module will be integrated into NAVTOR’s existing, joined-op portfolio. Although it’s early days, Hjøllo suggests that this would be on both NavStation (the company’s onboard digital chart table/planning system) and NavFleet (a shore-based management, monitoring and optimization solution) allowing both vessel and office teams to make the most of up-to-the-minute insights and enhance decision making.
“We’ve always been focused on developing innovations that simplify life at sea for our customers,” he states. “This is the epitome of that; gathering and utilizing vast amounts of complex data to deliver straightforward, actionable and powerful recommendations. It’s an advisory functionality that we’re constantly building, helping deliver added value for anyone that’s looking to gain advantage.”
As his job title suggests, the key advantages Hjøllo’s setting his sights on revolve around sustainability, which, as he rightly points out, goes hand in hand with commercial benefits.
“If we can dynamically optimize voyages and energy consumption, we can also, by extension, dynamically optimize costs,” he says, “helping owners around the world cut down on their greatest OPEX outgoing.
“But, as you might expect, it’s the impact on emissions that should create the greatest excitement.”
Bringing us back to that 1% figure.
How is that possible?
Hjøllo admits that the number is both a best and worst case forecasted scenario.
Here’s how the GASS partners arrived at it:
Today, worldwide shipping accounts for nearly 3% of greenhouse gas emissions. However, widely reported research suggests that it could reach as much as 17% by 2050 as global trade expands and other industries cut emissions faster than maritime. GASS expects that a machine learning application that dynamically optimises vessel energy use should be able to reduce consumption, and therefore GHG emissions, by 20%.
So, when the 30%+ of vessels in the world fleet that use NAVTOR products have their breakthrough AI technology powered up, that translates to over a 5% cut in all of shipping’s emissions and (assuming the 17% figure) a 1% cut in total global emissions.
Hjøllo can’t supress a smile, but is cautious enough not to get carried away.
“I think the important thing is there’s nothing to suggest we can’t do this – we, and our partners, have the track record, technology, and domain expertise to succeed here. The 1% figure obviously depends on a lot of variables that are beyond our control, but the 20% reduction per vessel doesn’t. That is a very realistic target.”
He concludes: “If we can play a part in reducing energy use and emissions by one fifth on every vessel we serve – regardless of type, age, location, weather, whatever – imagine how powerful that could be. Imagine the difference environmentally, from a regulatory perspective and commercially. In an industry, and a world, where every marginal gain is a major win, this is…well…”
He shrugs his shoulders, temporarily lost for words.
Even Hjøllo, it appears, needs to reboot sometimes.
Partners in the GASS project: