Jan Sjöström has captained vessels for Stena Line for 40 years, and he knows more than anyone how to navigate waters in the most efficient way. For the past several months, Sjöström has had an intriguing new student: an artificial intelligence. The Swedish ferry firm has teamed up with Hitachi to build an AI that can help pilot ships, with Sjöström helping train the system.
The aim isn't to use AI to replace captains such as Sjöström, but to capture in a model as much of what they've learned about navigating in quickly changing conditions to help other captains make the most efficient decisions in terms of fuel efficiency.
Stena Line have a goal of reducing fuel consumption by 2.5 per cent yearly, and predicts the model could be an important contribution to this, in turn cutting emissions, and helping the shipping company reach sustainability goals – highlighting how social innovation can help the environment as well as reduce costs. "This is an important step in our aim to become a cognitive ferry company but also a leader in sustainable shipping,” says Lars Carlsson, Stena Line's head of AI.”We believe there is a reasonable potential in fuel saving, which will have a positive impact on the environment, and also on the cost efficiency of running the vessel."
Vessels already have autopilot, which computes the shortest distance from one point to another, notes Per Kristian Egseth, sales lead for EMEA AI and Machine Learning solutions at Hitachi. "But it's not really taking into account the fact that the environment they are operating in is dynamic," he says, noting that direction and strength of winds and subsea currents can quickly change. "The captain has to manoeuver this vessel in a highly dynamic environment where the adverse impact for not being able to read the environment can be significant in terms of extra fuel consumption."
The model doesn't only seek to avoid windy weather or difficult currents. A phenomenon called "squat effect", in which ships running through shallow areas push some of the water away, meaning the vessel is even closer to the seabed, which increases fuel consumption. "It allows the Captain to optimise the route and the speed to best tackle the conditions," Egseth says.
The system is currently being trialled onboard the ferry Stena Scandinavica, the ship captained by senior master Sjöström, ferrying passengers between Gothenburg, Sweden and Kiel, Germany. "It gives you a suggestion on how to handle the vessel," Sjöström says. However, the system is still being developed, and Sjöström is helping to teach the model. "It's under development at the moment, so Jan and his colleagues provide feedback to improve its overall performance," says Carlsson.
Has the AI model taught Sjöström anything about captaining a vessel yet? "No, not really," he says. "But I've seen great improvements since we started in July, so we're really going in the right direction."
Though fully automated vessels do exist, that isn't the goal with Stena Line. Egseth says the aim is to use AI to help humans extend their knowledge beyond themselves. "Stena Line and other companies operating complex assets can capture the tacit knowledge of those super-users and embed that into an AI," Egseth says. The algorithm and model become an "extension" of the captain that others can use and learn from, making Sjöström and his skills available to help more of his colleagues – even after he retires.
“Planning a trip and handling a vessel in a safe and, at the same time, fuel-efficient way is craftsmanship. Practice makes perfect, but when assisted by AI, a new captain or officer could learn how to fuel optimise more quickly,” Sjöström says.
"The captain we're working with is the most senior one — he's been sailing for 40 years," adds Egseth. "He is really a human algorithm. He knows when you should change speed and direction of the vessel, not because of any data coming in, but because he's been out there for a long time," Egseth says. "We are building a mathematical model that sort of proves his instincts and allows all captains subsequently to use that gut feeling." The aim is to have AI-assistance on all of Stena Lines's 38 ships across 21 routes by 2021.
"How can we use AI to extend the best capabilities of the best of employees to the rest of the field to drive better decision making — this is at the heart of what we're trying to achieve," adds Egseth.
The work on-board Stena Scandinavica is only the beginning, with Stena Line hoping to extend AI tools across its entire business by 2021, for social innovation such as sustainability as well as improving operations. "I see applying AI to improve fuel efficiency as a starting point for analytics and machine learning," says Egseth. "They are gearing now to become the leading cognitive ferry company in the world."
--
Modern life is saturated with data and technologies are emerging nearly every day – but how can we use these innovations to make a real difference to the world?
Hitachi believes that social innovation should underpin everything it does, so it can find ways to tackle the biggest issues we face today.
Visit Social-Innovation.Hitachi to learn how social innovation is helping Hitachi drive change across the globe.
This article was originally published by WIRED UK
