AI and machine learning into maritime tasks

Real-time vessel monitoring

Last November, the digital team at the CSL Group, a founding Green Marine participating member, was recognized with the 2024 International Bulk Journal (IBJ) IT Solutions Award for their project leveraging AI to optimize vessel scheduling, safety, and fuel efficiency.

“The project is still quite new – in the developmental phase – but our digital team is out there working to lead the industry, so I’m glad our people got recognition for their efforts,” says Allister Paterson, the CSL Group’s executive vice-president and chief strategic partnerships officer.

CSL’s digital team is continuing to work on the company’s proprietary Operational Optimizer (O2) technology that provides a real-time vessel monitoring system that continuously oversees the status of vessel systems and indicates data trends for analysis by CSL’s ship and shore teams.

“All of this began with us retrofitting our fleet over the past seven years with sensors on most everything to generate data to help us operate more efficiently and to predict any problems before they happen,” Paterson adds.

It is also being used to optimize vessel scheduling from port to port. “We know that if we go 10% faster, it takes 35% additional fuel, but if we travel slower to save fuel, we also need a greater number of vessels to carry the same amount of cargo,” Paterson explains. “Using AI to process our daily real-time data will help us to determine the ideal equilibrium at both the macro and micro level.”

______________________

AI to serve biodiversity

Green Marine partner Whale Seeker has gained international recognition for developing scalable, industry-ready AI tools that help maritime stakeholders increase efficiency, reduce risk, and meet regulatory obligations related to biodiversity. The Montreal-based company has been cited by the UNESCO’s Center for Research on Artificial Intelligence as having one of the world’s top 10 AI projects.

Leveraging AI, Whale Seeker is creating tools to make it easier to monitor whales and other marine life, as well as birds. Whale Seeker’s tools are already being used by clients in government, the maritime industry, the energy sector, in addition to environmental research and conservation organizations. They help to reduce operational delays, support compliance reporting, and make for smarter planning and decision-making.

“By embedding ecological context directly into the AI development process, we’ve created tools that are not only faster and more cost-effective than manual annotation, but scientifically robust and standardized for regulatory and research applications.”

The company’s team of marine biologists works closely with AI specialists to ensure that every model built reflects ecological reality. “This interdisciplinary approach leads to detection tools that dramatically accelerate image processing while maintaining the scientific integrity required for conservation, impact assessment and policy compliance,” Charry Tessier says.

______________________

Simplifying scientific data management

At the St. Lawrence Global Observatory (SLGO), a Green Marine supporter, AI is being tested internally to more quickly process the various data that the centre receives from academic, governmental and NGO research partners, so that it can provide comprehensive standardized information to its stakeholders along the St. Lawrence River region.

“We’re in the process of integrating AI to assist us in what historically took us so much time to do manually,” shares Anne-Sophie Ste-Marie, SLGO’s director of Partnerships and Communication.

SLGO’s goal is to feature AI-assistance in a number of its internal data-processing tools.

SLGO is exploring the use of AI to feed its online catalogue that already hosts well over 300 open scientific research datasets to respond to searches for various scientific information of potential relevance to the St. Lawrence River’s ecosystem and marine life. “The amount of information is so vast and dispersed in different open data portals that it’s often difficult to track down what’s available,” Ste-Marie relates. “We’re testing AI technology to connect external data sources to SLGO’s catalogue to help end users to quickly find the relevant up-to-date information that might not turn up through regular Internet search engines or with a look at individual data portals.”

______________________

Give oceans a voice

Ocean Sonics, a Green Marine partner, is working on how to effectively use AI machine learning (ML) to consistently identify specific underwater sounds and measure their level in real time in subsea areas that could be teeming with sound from various sources.

Cofounders Mark Wood and Desiree Stockermans founded Ocean Sonics in 2012 after working seven years in developing the icListen, a smart digital hydrophone. The icListen records underwater sound at the source allowing users to manage the amount of data they capture so that it can be sent to the surface in real time as visual data.

“Most underwater sound is beyond our limited hearing range,” Wood says. “We’re giving the oceans a voice by providing the instruments to see and visualize the sound occurring under water.”

The icListen technology has already been used in various marine industry applications that include measuring the underwater radiated noise by vessels, as well as detecting the presence of endangered whale species within a harbour. It can also detect the ‘hissing’ from a leaky pipe that might visually go undetected.

Ocean Sonics is now working on how to use ML to better manage the complexity of sounds within marine environments. “ML can process complex information a lot more quickly than humans without getting bored or tired, but it needs the expertise that people have acquired,” Wood notes.

An example is how some biologists can identify specific whales by their calls.

“The thing is: it’s not easy, with one of the biggest challenges being what information to filter out.”

Ocean Sonics is collaborating with maritime enterprises to determine what is important to them in terms of outcomes. For example, if there are plans to build a bridge, an AI program could integrate all required noise regulations and standards for work done near or nearby the relevant body of water and quickly issue a report with graphs to indicate acceptable sound levels for different construction activities.

Examples of other possible applications include assisting third-party regulators when it comes to verifying if a ship owner has carried out improvements to decrease underwater noise from mechanical or operational vibrations, or to continually monitor a vessel’s operations to immediately identify and remedy an unusual source or level of noise.

“We’ve not yet commercialized these services because we’re still working on how to use AI machine learning to optimally manage complexity, which is complex in and of itself because we have to ‘train’ the AI algorithms to recognize a vast array of information and how to respond to each of it,” Wood explains. “That’s why we’re instead undertaking a few pilot projects.”

One of them involves Ocean Sonics monitoring the soundscape at two locations within the Port of Halifax – everything from the underwater radiated noise from the vessels to construction activity to the marine life. “We’re working to establish what’s happening within this busy harbour from a sound perspective,” says Jennifer Aftanas, the marketing and public relations lead at Ocean Sonics, adding that some of the findings will be presented at GreenTech 2025 in New Orleans.

Another pilot project involves using AI modeling to determine how the bow thrusters from cruise ships arriving at the port should be used to avoid damaging the port’s dock walls. “We’re using sound as a way to monitor how ships park,” Aftanas shares.

______________________

Accelerating the blue economy with AI

Canada’s Ocean Supercluster, a Green Marine supporter, is assisting new and existing Canadian companies, as well as enterprises keen to invest in Canada, to further build the country’s ocean economy.

“There are just so many ways now that AI can help companies to be more efficient, safe, sustainable, and competitive and expand their business,” says Jennifer LaPlante, who oversees Chief Growth and Investment. “Canada is a global leader in AI development – particularly with its stronghold in Montreal – but that’s not always translating into commercialized opportunities, especially in terms of the blue economy.”

One example is Canada’s Ocean Supercluster investing $5 million in a $12-million project involving Green Marine partner Global Spatial Technology Solutions (GSTS) Inc. to develop and implement the first AI-powered collaborative predictive berth scheduler. The OCIANA system integrates the information provided by ship owners, pilotage services and port authorities to determine the optimal arrival time for a vessel.

GSTS is leading the project in collaboration with Canada Steamship Lines, the Montreal Port Authority, and the Laurentian Pilotage Authority (LPA) – all Green Marine participants – as well as Green Marine supporter Clear Seas. The proprietary AI technology is being leveraged to support the greater efficiency and sustainability of green shipping corridors through advanced digitalization.

Canada’s Ocean Supercluster has also co-invested in MarineLabs, another Green Marine supporter. Marine Labs is using AI to process the varied and detailed weather and wave information garnered from coastal buoys to provide insights towards advancing marine safety and climate-resistant coastlines.

The country’s major ports, as well as other enterprises, are in discussions with Canada’s Ocean Supercluster to discuss possible AI applications. “We hold round tables to understand where the gaps are in the ocean economy’s adoption of AI solutions,” LaPlante explains. “And we’ve recently launched a program to educate the maritime sector about AI possibilities, which includes safety, privacy and ethical considerations.”

The courses are open to Canada’s Ocean Supercluster members, but it’s free to join. Canada’s Ocean Supercluster can also help innovators to connect with a maritime enterprise if they require specific data for ML testing and/or training purposes, or to just to discuss how a maritime challenge might be addressed using AI.

One of the biggest issues is comprehensive and consistent data. “If you have AI tracking engine performance for preventive maintenance, for example, you need a consistent vocabulary for relating issues and abnormal sounds,” LaPlante says. “You can’t have one person saying it’s a banging noise and another calling it clunky.”

Yet another large concern is the return on investment. “The upfront cost might be significant but the efficiencies in the long run might make it really worthwhile,” LaPlante says. “In any case, it’s good to find out, which can start with a simple email to us.”

______________________

Optimizing human resources through predictive AI

New Green Marine partner Airudi is filling a particular niche by using AI to optimize workplace operations and human resources management to improved productivity, health and safety, and work-life balance.

“With our workforce prediction and allocation solution, we’re seeing a 15% to 30% reduction in errors in terms of optimal labour deployment,” says Mathieu Charbonneau, Airudi’s vice-president and its Transport Division’s general manager. “And employees appreciate knowing earlier when they can expect time off so they can better plan their personal lives.”

The company was founded by two specialists in human resources seven years ago. “Pape Wade and Amanda Arciero saw this need and collaborated with AI experts to create platforms that better predict labour demand as well as provide more visibility to the workforce that can be deployed for operations,” Charbonneau relates. “All this enables us to be more responsive in schedule planning.”

Airudi’s services help companies to sort through résumés to hire the best available people, simplify the verification of their qualifications and eligibility, and then optimize their work schedule.

In the maritime industry, Airudi is specifically focused on using its AI allocation predictor to get the right types and number of employees to the correct vessel/berth at the required time.  “We’ve been working with the Maritime Employers Association at the Port of Montreal to use AI to propose an optimal scenario for the deployment of labour that respects collective agreements, the availability of port workers, and the required skills,” Charbonneau explains. “AI facilitates setting regular schedules farther ahead of time, while also suggesting immediate changes in response to a ship arriving 48 hours early or 24 hours late or some other unanticipated development.”

Airudi is also partnering with SOGET, France’s leader in technological logistics, to integrate the AI expertise in human resources management into SOGET’s logistics platform.

Charbonneau anticipates that Airudi’s expertise will extend in due course to coordinating labour for connecting rail, aviation and trucking services at ports. “The transportation sector has been collecting data for years,” Charbonneau notes.

The company partnered with Green Marine to underline the greater sustainability achieved by AI-informed labour management. “Any time we reduce the wait time for vessels at a port we’re also reducing the required fuel use while anchored or docked and the related greenhouse gas emissions,” Charbonneau notes.

______________________

Optimize fuel consumption

Green Marine Europe partner Opsealog has been using AI in its Software as a Service (SaaS) platform to help charterers in the energy sector to optimize their overall fleet operations through the better use of their support vessels in offshore operations, as well as optimizing the performance of each vessel.

“Our core value proposition is to generate 10% to 15% in fuel savings for our customers, depending on the operational and geographical context,” says Martin Smetek, Opsealog’s director of Products and Innovation.

As soon as Opsealog receives data, whether from reports or sensors, its SaaS platform establishes vessel-specific consumption baselines to assess initial operational behaviour. At the same time, the platform also models activity-based consumption profiles for each vessel, drawing on a continuously enriched knowledge base powered by new data points.

“Then aligned with our client’s priorities, we establish optimization targets which form the basis for the automatic generation of recommendations, which are transmitted directly to the operational units at sea.” Smetek explains.

“This automation helps to make our services more efficient by quickly identifying the levers for action to optimize a client’s optimal performance,” he adds.

Opsealog is working with a consortium of companies on a project funded by ADEME, the French agency for ecological transition, which has been a supporter of Green Marine Europe since its creation. The CASSIOPEE project is an R&D initiative supported by the Conseil d'Orientation pour la Recherche et l'Innovation des Industriels de la Mer (CORIMER) that ambitiously aims to use AI to process the information from a connected fleet of approximately 10 ships to generate several hundred million data points daily.

Challenges remain, including a consistency in data supply. To date, there are no maritime regulations that require digital logbooks. The wide discrepancy in reported information and sensors, along with various levels of connectivity among vessels, makes it challenging to establish consistent data. (See Opselog’s white paper: Creating Value for Data Standardisation.)

“No data is free from the risk of error,” Smetek adds. “Data entered or validated by a human operator always carries a risk of error, just as data from a sensor may be subject to bias or to limitations inherent in its technology. It is precisely this complexity that we address with our methods and tools, by securing the quality of the data to guarantee reliable and usable analyses.”

“It must integrate harmoniously with human expertise, which will evolve both at sea and on land,” he explains “Maritime professions are set to be transformed: seafarers will benefit from tools to simplify decision-making, while shore teams will develop skills focused on developing algorithms, analysing data and supervising intelligent systems.”

______________________

Smarter maritime corridors

Donald Roussel is a founder and board member of the Intelligent Maritime Corridors International Council, a Green Marine supporter. A former mariner and past director-general of marine safety and security for Transport Canada, he launched the not-for-profit in early 2021 after seeing how far ahead Finland, Norway and Sweden are in comparison to North America in terms of developing AI for autonomous vessels.

Roussel is concerned about North America lagging when it comes to AI use within the maritime industry. He notes the European Union’s more significant support for maritime AI research and development. In 2018, for example, the European Commission began offering up to €20 million to build and demonstrate a full autonomous vessel for inland and coastal waterways.

Companies had already started R&D. In Norway, Yara International announced its partnership with the Kongsberg Group in 2017 to build the world’s first autonomous zero-emission container vessel. The 262-foot (almost 80-metre) Yara Birkeland is powered by a 6.8-MWh battery bank and can transport 120 TEUs of cargo. By transporting mineral fertilizer between Norwegian ports, it’s replacing 40,000 diesel-powered trips by truck, which eliminates 1,000 metric tons of CO2 yearly. The vessel completed its two-year testing in 2024.

The Central Commission for the Navigation of the Rhine has 48 autonomous zero-emission projects listed at last count stretching into 2029. Efforts are under way to use existing permits for the Lower Rhine trials and to extend this solution to other waterways. For example, the Port of Hamburg is in the application phase for sections of the canal in north-west Germany, as well as the Mittellandkanal and other parts of the Rhine.

In Asia, the Nippon Foundation is administering the Joint Technological Development Program for the Demonstration of Fully Autonomous Ships under the MEGURI 2040 Fully Autonomous Ship Project. Launched in 2020, the project involves 51 Japanese companies and is aiming for full-scale commercialization of fully autonomous ship technology by this year.

Roussel says Canada is well-positioned with its inland waterways for testing and adapting emerging technologies. “Some studies indicate 11-to-17% of vessels globally will be remotely assisted in the near future,” Roussel says.

Roussel notes that Canada’s Shipping Act has already been modified to allow for trips for testing purposes. “So we need to promote this and get it tested, because right now, the customers for this are not there,” he says, adding that Canada risks losing its competitiveness if it doesn’t move faster.

He also says the regulatory framework must keep up with upcoming fast-paced changes. “Maritime UK has already issued Version 7 of the Maritime Autonomous Ship Systems (MASS) UK Industry Conduct Principles and Code of Practice,” he notes.

“We need industry leaders, unions, government and the people who do the prospective analysis altogether discussing this now,” he emphasizes. Roussel points to the partnership between the International Pilots’ Association (IMPA) with the Canadian National Centre of Expertise on Maritime Pilotage and the Canadian Coast Guard to explore the feasibility and impacts of remote pilotage as a step in the right direction.

______________________

Accelerating processes through categorization and documentation

InnovMarine, a Green Marine partner for 10 years, is also stepping into the world of artificial intelligence applications for the maritime sector. The enterprise founded in Lévis in 2014 is already making a name for itself in the development and marketing of advanced digital solutions for shipbuilding and vessel operations and maintenance.

Its current suite of AI tools includes ShipConstructor for naval design and engineering, QuickBrain which is applicable to intelligent maintenance management with 3D visualization, CMPIC for wiring management, and IMSurvey which is used for digital ship inspections.

QuickBrain is being developed in partnership with the French company Ennovia, which collaborates with IBM on AI solutions applied to industrial maintenance. InnovMarine has also just signed a marketing contract with IBM for a portfolio of AI and automation solutions to produce specific tools for the shipping industry.

IBM has already tested the technology to improve its internal processes, particularly in human resources management and accounting. "So, we're working with engineering consulting firms, shipyards, and ship owners to identify possible applications, based on priorities," Drapeau adds.

The administrative applications tested by IBM are already relevant and transferable to shipping companies, but AI’s role can go even further.

The company is already active in supporting the Royal Canadian Navy's digital transformation, enabling advanced management of maintenance operations and better intervention planning. The partnership with Ennovia has notably enabled InnovMarine to expand into North American and European markets, particularly thanks to the QuickBrain solution which has already been adopted in France by the navy and several shipyards.  This collaboration with Ennovia is part of a dynamic fleet modernization effort in Canada, with prospects for major contracts for the construction and renovation of military and civilian vessels.

Drapeau adds that these AI advances are transferable to other types of vessels. "Our contracts with the navy are not related to defence per se,” he explains. “They cover functions that are the same as those in the marine industry, such as ship operations, ship inspections, and administrative management."