Row, row robo-boats, gently down the data stream - a new age of autonomous shipping sets sail
The UK is helping a robot launch a future vision of autonomous sea transport. What are the advantages and risks of this new technology?
The UK is providing the navigation data to help an unmanned boat cross the Atlantic this year, a project that the vessel’s backers hope will usher in a new age of autonomous, remote-controlled, or autonomy-assisted shipping.
As with autonomous land and air transport (aka driverless vehicles and drones), the developers hope that robotic vessels could one day offer environmental, sustainability, efficiency, and cost advantages.
These would be on a sliding scale from new AI systems to aid mariners all the way through to fully autonomous, unmanned craft.
The Mayflower Autonomous Ship (MAS) is an unmanned, AI-piloted robotic research vessel, originally conceived to commemorate the 400th anniversary of the Mayflower voyage, as the name Mayflower 400 on the boat itself suggests.
It has six AI cameras, 30 other sensors, and 15 edge devices onboard – but no human crew. The MAS can see 2.5 nautical miles ahead, recognise other vessels, and, via its onboard computer vision system, identify hazards such as partially submerged wrecks or lost shipping containers.
A sound idea?
Later this Spring, the MAS will set out to cross the Atlantic from Plymouth Sound in the UK, where it is moored, to Plymouth in Massachusetts, USA. The inspiring, 3,244 nautical-mile trip will follow the route of the Pilgrims’ 1620 voyage in the original Mayflower to what was then called the New World.
In this way, its developers hope to greet another new world– one of autonomous and AI-assisted shipping, which might one day include container vessels, tankers, and even passenger ships.
This will be the second attempt at the historic crossing, after the MAS was forced to turn back just three days into the first in June 2021, when it developed a minor fault. You can follow the boat’s progress live on its website.
The solar-powered vessel was proposed in 2016 by Brett Phaneuf, founder and CEO of the Submergence Group in the US and of M Subs Ltd in the UK. A new company, Marine Ai, has been spun out from the latter.
The MAS project is now led by ProMare, a US non-profit that promotes marine exploration, and is supported by IBM (provider of the computer vision system) and a global consortium of other technology and academic research partners, including six UK universities. Meanwhile, the UK Hydrographic Office (UKHO) revealed this week that it is the provider of S-100 navigation data for the MAS, and is funding its integration with the boat’s AI and machine learning systems.
S-100 is a new data model and standard developed by the International Hydrographic Organization (IHO).
According to an announcement from the British government, this will be the first time that S-100 data has been tested onboard an unmanned or autonomous marine vessel. Ongoing trials in UK waters are testing the boat’s ability to read, integrate, and use the new data format, and will include the MAS simulating the behaviour of a much larger vessel: a frigate.
Mark Casey, Head of Research, Design, and Innovation at the UKHO, pitches:
This project will be highly relevant for the use case of S-100, as it will provide feedback on the utility of the S-100 data and format.
It will identify gaps in the data and the method of deployment and integration, and provide real-world operational testing that will inform future deployments of S-100 data from the UKHO to other manned and unmanned vessel operators across the world.
The MAS project will also help to advance the state of the art in marine autonomy and will be crucial for the advancement of autonomous shipping.
So, what is the point of the technology?
Many large vessels today, such as tankers, container ships, and ferries, are old, slow, dirty, polluting, and expensive to run, both in fuel terms and because they need extensive human crews.
In addition, this makes it a slow and extremely expensive process to hire ships for a range of other tasks, including maintenance duties on offshore energy platforms.
The backers of autonomous ships believe that developing a new generation of vessels that can, to a large degree, operate themselves will solve many of these problems, while introducing navigation aids and other systems that will benefit human seafarers.
The technology would also have significant military applications: robot warships, autonomous support vessels, unmanned surveillance craft, and submarines. Robot submarines have, of course, been with us for many years. Most are tethered and remotely operated underwater vehicles (ROVs) used for deep-sea exploration, maintenance, and engineering tasks, rather than the new generation of autonomous underwater vehicles (AUVs). Autonomy has real utility beneath the waves, as radio signals propagate poorly in seawater.
…and downside risk
However, as with autonomy on land and in the air, autonomous shipping generally faces massive legal, regulatory, social, employment, and trust hurdles, given the centuries of human expertise and skill that have gone into navigating the oceans, and the local economies that rely on it.
As a result, the market will be freighted with risk, including opposition from the thousands of skilled people that shipping companies currently employ. The recent controversy over P&O firing its UK crews then hiring cheaper agency workers would be a drop in the ocean compared to possible industry-wide opposition to autonomous shipping.
Any death, loss of cargo, collision, or accident involving a robotic ship would also usher in a new world of legal, liability, and insurance complexities – not to mention countless opportunities for crime, such as fraud and theft.
In the years ahead, the real-world progress of driverless vehicles and autonomous or remotely piloted delivery drones will reveal the gaps – if any – between technology promise, social acceptance, and economic payback. Those lessons will inform the decisions of shipping industry disruptors and their investors.
No one doubts the innovation, engineering expertise, academic depth, and ambition behind all forms of autonomous transport, nor the size of the environmental challenge. Such research programmes are already producing technologies that help human beings and may keep them safer: new sensors and driver-assistance systems, for example. These are welcome, and projects like MAS are introducing them to the maritime world.
But common sense is absent in some of the envisaged markets for unmanned transport, as explored in my recent reports on urban delivery drones. Many innovations introduce as least as many problems as they solve, which is why some programmes are being scaled back.
Meanwhile, trust in any crewless sea vessel that carries human beings or essential supplies will remain low for the foreseeable future. Losing an unmanned boat mid Atlantic is one thing, but an autonomous boat full of passengers, food, or raw materials would be quite another.