Artificial intelligence has become a topic du jour across industries, promising, or threatening — depending on one’s own sense of job security — to upend healthcare, finance, manufacturing, security, education and more by outsourcing some parts of decision-making processes to algorithms that can learn.
The maritime industry is increasingly receptive to AI, according to a recent report from Thetius IQ, a research organization for maritime technology. The report found that, in 2024, 420 organizations of some sort were involved in developing, selling, buying or investing in maritime-specific AI technologies. That is up from 276 in 2023. The report counted 125 companies that sell specialty AI systems to shipping companies and 36 shipping companies that have purchased AI technologies or announced plans to. Adoption of AI tripled in a single year.
The report concludes, “The maritime sector, often perceived as traditional and resistant to change, is now embracing AI with remarkable enthusiasm.”
Proponents say AI will streamline nearly every aspect of operating and maintaining a ship; it will direct navigation, manage storage and inventory, help crew comply with safety standards, detect hazards on board and in the water, predict maintenance and repair issues before they become problems, conserve fuel and reduce emissions — all while reducing the number of man hours invested in these concerns.
In an interview with Professional Mariner, Oliver-Andreas Leszczynski, head of artificial intelligence at the INER Institute of Northern-European Economic Research in the Northern Business School in Hamburg, Germany, said one should picture a ship and the places through which it sails as rich sources of data.
Weather conditions and attributes of a destination are sources of data that could decide navigation. Current patterns and wind conditions, plus the ship’s weight, are sources of data that could decide energy usage. Frequency of use, exposure to the elements and physical attributes too small for the human eye are sources of data that could predict the need to replace or repair pieces of equipment. We don’t have a way to synthesize all this data, but we could soon.
“So there are gigabytes, terabytes of data, which we are not able nowadays to really process,” said Leszczynski, “and with artificial intelligence — with the right data model — we will be able to really understand the data that’s produced on the ship … and really get the biggest outcome and the real benefit out of it.”
On a macro scale, he said, AI advancements contribute to more predictable shipping schedules and decreases in accidents.
Some AI technologies are already being tested. In 2022, Wallenius Wilhelmsen, a Swedish/Norwegian shipping company that mainly transports cars and trucks internationally, installed an artificial intelligence tool from Deep Sea Technologies on 65 vessels as a test run. Deep Sea is one of several boutique firms creating AI systems for the maritime industry. The Athens-based company calls its AI tool Pythia, which in Greek mythology is a title for an oracle.
In 2022, Ashlyn Misquitta, a mariner with 30 years’ experience, captained a car carrier vessel outfitted with Pythia, from Australia to Singapore, with several stops at Asian ports. He told Professional Mariner it was helpful for some purposes and not quite ripe for others.
Pythia decreased fuel usage by calculating factors like ocean currents and wind resistance. “The AI model uses a lot of sensors to determine where the combined vector hits you,” Misquitta said.
Reducing fuel consumption was a key goal of the trial — and of the industry in general, as the International Maritime Organization, individual nations and companies institute rigorous emission standards. “We can’t afford to have that variance anymore,” said Misquitta.
But Pythia was not as helpful as a navigator. Misquitta said he often rejected the routes it suggested. “We would plug into the system and we would get a route telling how to go from point A to point B with the speeds. The routes were not up to standard, in terms of safety. It would be too close to land. It would be too close to buoys or outside channels.”
Misquitta is now an operations manager at Wallenius Wilhelmsen and is involved in improving the AI system’s performance. “It’s a lot better now,” he said. “The AI started understanding what the ship can and can’t do when the weather gets bad and adjusting accordingly. There is still scope for improvement, but we are in a much better place today than we were 12 months ago.”
Geir Fagerheim, Wallenius Wilhelmsen’s senior vice president of marine operations, told Professional Mariner the technology is “just at its infancy.”
“I think you can compare it to, say, a very advanced cruise control, which the airline industry has been using for many years already.”
Fagerheim does not think AI will lead to staff reductions anytime soon. “It’s still at a stage where it’s not removing the need for, say, competent people on the bridge or on board vessels. This is purely a solution to help them with a problem that is too big for human beings to calculate.”
But “over time, it will obviously lead to requiring upskilling or reskilling.” Crew members may have to become more technology and information savvy.