Driverless boats: How automation is coming to the ocean
Wireless Technologies

Driverless boats: How automation is coming to the ocean

The world is becoming automated. But it’s not just cars and factories. The mining industry, for example, has long been a proponent of taking people out of the equation with remotely-controlled or entirely automated vehicles now doing most of the actual digging and mining. Now the ocean is set to follow suit.

Containerisation and shipping are the lifeblood of today’s modern world. Some 90% of the world’s trade is carried by sea. But the shipping industry is in a state of flux. Last year a major Korean shipping company went bankrupt, and in March this year a $60 million container ship just seven-years-old (the average lifespan of such a vessel should be around 25-30 years) was scrapped for a mere $5.5 million.

“Cargo shipping nowadays is not a good business. Almost everybody is losing money at this,” says Dr. Jaakko Tavitie, Innovation Scout at DIMECC, a Finnish organisation dedicated to exploring and enabling digital transformation.

“There is more shipping capacity available on the market than there is need to ship goods. There's an over-capacity situation, where prices come down and the profitability is pretty weak.”

These hard times, however, means the industry is ripe for innovation and disruption.


Empty - but not rudderless - vessels

Where once upon a time a boat at sea without a crew was the stuff of nightmares and mystery, today it is a realistic proposition. We could see the first remotely-operated vessel operating in local waters by 2020, according to Rolls-Royce, with open sea operations by 2025.

“Five years after that we expect unmanned ocean going vessels to be a common sight on the ocean,” says Oskar Levander, Rolls-Royce’s Vice President of Innovation-Marine. “Tugs, along with road ferries, are likely to be one of the first places we will see the commercial use of remotely operated and autonomous vessels.”

Ships have had autopilot features known as Iron Mike for nearly 100 years. Today there are increasingly sophisticated autopilot features that rely on GPS. But the vision for the likes of Rolls-Royce and others is to remove the need for any people on board at all and automate the entire journey from port to port.

“We believe autonomous shipping is the future of the maritime industry. As disruptive as the smart phone, the smart ship will revolutionise the landscape of ship design and operations,” says Levander.

“They have the potential to redefine the maritime industry and the roles of the players in it with implications for shipping companies, shipbuilders and maritime systems providers, as well as technology companies from other sectors, especially automotive.”

But, just like the roads are full of all different kinds of vehicles, the seas have a plethora of different vessels doing different jobs. Which means a single autonomous solution applicable to all vessel types is unlikely.

“Some could be completely un-crewed and look radically different from current vessels,” says Levander. “Others will be a blend of autonomous and remote control; sailing autonomously in open water, independently picking the best route and speed, and remotely controlled where more advanced manoeuvres are required – navigating in congested waters and entering and leaving port. Some, such as cruise ships, are always likely to need crew if only in a customer service, safety and reassurance capacity.”


AI boats will be sailing soon

Much like driverless cars, the technology isn’t a far off sci-fi idea, but tantalisingly close.

“The bits and pieces that will be needed already exist,” says Dr. Talvitie. “They may not be at the proper maturity level yet, but they exist. We do not need to create any new technologies to do this; a vessel which goes through a certain route without a captain on board, that could be done almost today.”

Levander explains: “The sensor technology needed is sound and commercially available, and the algorithms needed for robust decision support systems – the vessel’s ‘virtual captain’ – are not far away. The challenge is to find the optimum way to combine them cost effectively in a marine environment. 

“Vital to the development of remote controlled and autonomous operations will be a ship’s ability to establish and communicate what is around it – in effect develop its electronic senses in order to navigate safely, avoid collisions and perform complex manoeuvres such as docking.”

Tests of sensor arrays in different operating and climatic conditions is currently being carried out in Finland. Rolls-Royce is currently working on situational awareness systems which integrate data from multiple sensors into an analysis of the ship’s immediate environment.  This data would then be transmitted back to a remote operations centre or used by the ship’s on-board computers to generate the appropriate next action.


A tidal wave of oceanic automation

As well as startups such as Sea Machines and established organisations such as Yara International, the number of initiatives and collaborative projects currently investigating the role of autonomous systems in shipping is growing rapidly. MAXCMAS , the Norwegian Forum for Autonomous Ships, the UN-backed Maritime Unmanned Navigation through Intelligence in Networks (MUNIN) project, DIMECC’s One Sea Autonomous Maritime Ecosystem and the Autonomous Waterborne Applications Initiative (AAWA) from the Finnish innovation fund TEKES are  just a few examples. A conglomerate of Japanese companies are also exploring this space.

DIMECC’s One Sea is looking at how to create a business environment which enables autonomous shipping to become a viable commercial business. The AAWA initiative’s mission is to explore the economic, social, legal, regulatory and technological factors which need to be addressed to make autonomous ships a reality.

Many of these projects include major players such as Rolls Royce, Meyer Turku, Wärtsilä, and ABB.

I have seen a lot of enthusiasm,” says Tekes Program Manager Piia Moilanen. “New technologies can solve so many challenges, like safety, and give new competitive edges to the industry.”

This enthusiasm is despite the AAWA report saying the marine industry needs to overcome its “conservative nature” in order to benefit from the technology.

“I have also faced some scepticism over the pace, as many see that international legislation is slow in responding into the developments,” adds Moilanen.


The future of boats: new designs, new services

Why do we need to automate seafaring? In short, says Levander, because it’s expected to be safer, more efficient and cheaper both to build and to run.

“Remote-controlled and autonomous vessels can be designed with a larger cargo capacity, better hydrodynamics and less wind resistance.”

With no crew to accommodate, accommodation and elements of the ventilation, heating and sewage systems can be removed, making ships lighter. This leads to a reduction in energy and fuel consumption, lower operating and construction costs, and eventually entirely new designs. Levander suggests transport costs of a fully autonomous vessel could be around 20% compared to today.

As with their four-wheeled counterparts on the roads, safety is being touted as a major potential benefit of seafaring automation. According to a report from Allianz [PDF], there were on average 11 deaths per 100,000 seafarer-years between 1996-2005 for UK vessels; a figure some 12 times higher than the general workforce. In Hong Kong that figure was 56 per 100,000 seafarer-years, 84 per 100,000 in Poland and 90 per 100,000 in Denmark. Around one ship in every 670 was lost at sea in 2010. According to that report, up to 96% of accidents are a result of human error.

“This is often as a result of fatigue,” says Levander. “Remote controlled and autonomous ships don’t get tired and will reduce the risk of injury and even death amongst ship’s crews with all the potential loss or damage of valuable assets.”

But the benefits of automation wouldn’t stop once a boat pulls into port.

“Constant real-time remote monitoring of vessels worldwide will see ships become more closely integrated into logistics or supply chains, enabling global companies to focus on using a whole fleet to best effect, generating cost savings and improving revenue generation.”

Quite how this massive wave of disruption will impact businesses is unknown. Most of the people we talked to for this piece agree that there will be new players in the game while some old ones will disappear or at the very least shrink considerably.

This has the potential to create new shipping services, such as online cargo service marketplaces, more efficient pooling and leasing of assets, and new alliances.

Some of these services will support existing players in the market and others will be more disruptive – allowing new players to enter and potentially capture a significant share of business in the same way as Uber, Spotify, and Airbnb have done in other industries

While it is out of the remit of DIMECC’s One Sea project, Dr. Talvitie predicts that initially there will be a spate of retrofitting autonomous systems – replacing a whole fleet of boats that have decades left to run overnight simply wouldn’t be practical – before boat designers look at redesigning vessels from the bottom up with autonomy in mind.

“They will have to be completely different from the current ships, because there will be less or probably no crew, all the ship systems that are in there have to be much more reliable than they are today. Many of the systems are built to be serviced en route, and if there are no people to do this service, they will have to be redesigned and rebuilt so that they will last the routes without servicing.”


The rise of robot captains

It’s not yet clear how this automation of the oceans could affect the skills of sailors. The AAWA report warns that in the short term at least, there could be a degradation of skills as automation takes over, or that fewer crew members will need a wider range of skills, especially in the early days were one person is remote controlling several ships at once. 

“We see that the maritime industry will in the coming years need remarkably more ICT-experts than today,” says Tekes’ Moilanen. Perhaps, however, Finland’s history with Nokia and the startup scene it helped produce means the country could be well-placed.

“In Finland, we luckily have an excess of highly skilled ICT-engineers. However, there is a need to educate more people into this branch.”

Levander argues automation will actually help stem the skills shortage that already exists within the maritime industry:

“Ships are becoming increasingly complex with more systems, needing more skilled, operatives. At the same time changes in lifestyle and expectations are reducing the attractiveness of seafaring as a career with fewer people wanting to spend weeks at a time away from home and family.”

“Remote and autonomous operations could see the transfer of seafaring jobs, requiring high levels of education and skills, from sea to remote operations centres on land and make them more attractive to young people entering the industry.”


A new horizon mean new legislation

Much like on the roads with autonomous cars and in the skies with drones, current legal structures will need to be changed to accommodate self-piloting ships on the seas. The laws of the sea are governed by the International Maritime Organization. Robert Veal, Research Fellow, Tutor In Law at the University of Southampton, recently claimed that the regulation of unmanned ships will be “a bigger challenge than the technology”, partly down to the fact technology often moves faster than legislation.

“It typically takes anything up from five years to get a new regulation through the IMO,” says DIMECC’s Dr. Tavitie “I don't specifically know how much the IMO rules would have to be changed, but it's clear that they will have to be changed, and that will have to happen with collaboration between countries otherwise it won’t happen.”

Under current Maritime Law, “Every vessel shall at all times maintain a proper look-out by sight and hearing as well as by all available means appropriate”. This is similar in design to the 1968 Vienna Convention on Road Traffic, which stated at that “every driver shall at all times be able to control his vehicle or to guide his animals...[and] every moving vehicle or combination of vehicles shall have a driver”, until the UN last year updated the legislation to feature provisions for autonomous vehicles.

“Remote control will be important for the first generation of intelligent ships from a legal point of view,” explains Rolls-Royce’s Levander. “There will have to be a human being in charge although they may be in charge of a large number of ships at any one time switching between them as the need arises.

“Different types of ship, or ships at different stages of their voyage, will require different levels of remote control. A cargo ship far out at sea will require limited human supervision whilst ships operating in more congested shipping lanes, close to shore or on entering port will require much more.”

A Research Professor at the University of Ghent claimed in a paper [PDF] that Maritime law ‘appears to be relatively well armed’ for the arrival of such systems; current requirements around paper documents on board can be updated to digital, while the rules regarding the powers of a boat’s commander quickly become redundant if there is no captain or crew. Even requirements of a lookout could be updated given most large vessels have an array of monitoring equipment.

“Legislation can be changed if there is the political will. Questions of liability for autonomous ships are subject to national variations, but generally it seems that there is a less urgent need for regulatory change in this field. What also needs to be explored is to what extent other liability rules, such as product liability, would affect traditional rules of maritime liability and insurance.”


The sinking feeling of security

As is the case with any industry, with increasing use of technology comes an increased cyber-security risks. At a boating conference earlier in the year, BlackBerry researcher Campbell Murray demonstrated a live hack of a superyacht, and told the audience he had seen examples of financial information being stolen from on-board IT systems, and even entire navigation systems being locked out and held to ransom. In July it was revealed that many VSAT systems – which provide internet connection via satellite to ships – are easily accessible and often only use default log-in credentials.

“Headline news ten years ago would report attacks of pirates to capture cargo or passengers for ransom,” says Alex Manea, Chief Security Officer at BlackBerry. “But today, holding a ship’s internal system to ransom using malware is far more common – as is the compromise of shipping companies’ systems to identify and track individual items of valuable cargo that may be of interest to modern attackers, with some attackers even going as far as changing the delivery destination of containers so they can be easily intercepted.”

Manea explains that the reason behind this insecurity is two-fold. The first is that many technology components on board are often the same as in homes and offices – which come with the same vulnerabilities – which will only become more wide-spread as internet connectivity costs lower and coverage improves. The other is the old classic of IT: a large and complex attack vector create by brining together a number of disparate systems.

“A ship’s system is rarely provided by a single point of technology provision, meaning that multiple vendors’ solutions are integrated by the master platform on board a ship. This can include engine management, navigation, and communications, so the master platform poses a complex threat surface making it harder for cybercriminals to attack.”

“The marine industry also has a number of pseudo standards that component manufacturers must adhere to, allowing the data integration to the control centres, but the number of vendors feeding into the main system can present a vulnerability. There is also a perception that as a ship is isolated, and most of the communications are internal, that the risk of hacking is less. More sophisticated attackers, along with more complex technology, presents a pervasive threat to a ship’s data security.”


The ‘Digital Ocean’ as a network

However, while the technology is largely ready and legislation slowly but surely being looked at, one challenge still remains: connectivity. Where on land, autonomous systems are being complemented with increasingly smart infrastructure, there’s no such comparison on the world’s waters.

“If you think about how sensor-networked we are on land,” says Gary Gysin, President and CEO of Liquid Robotics, a subsidy of Boeing which designs ocean-going drones. It’s pretty good, and getting better and more autonomous. The issue with the ocean is that sensors exist, but they're typically on very, very expensive platforms, they're not networked, they're not integrated.”

Liquid is leading the charge to create what it calls ‘The Digital Ocean’; a mesh of sensors and connected devices which come together and form a pervasive network, and provide ubiquitous connectivity to the oceans.

“Maybe 5% of the ocean is actually known - we know more about space than we do the ocean. And yet if you think about all the things that are critical; our climate, our food supply, shipping commerce, security and defence purposes, there's just a whole lot that goes on in the ocean, and it's not sensor-networked. The whole idea of the digital ocean is to have a persistent network of sensors devices from seafloor to base.”

While having such sensor information fed into such autonomous systems would no doubt be useful – think real-time weather and traffic reports but for the entire Pacific or Atlantic – there’s another challenge around getting that data out of the boat and back to land. AI systems create a massive amount of data – Intel CEO Brian Krzanich last year predicted the average driverless car will generate around 4,000 GB of data per day once they are commercially available – and at least some of that data will be sent to a data centre for processing.

Yet large expanses of water have little to no connectivity options, and even in areas that do, the speeds are slow and the costs can be exorbitant. Ships mostly rely on radar and transponders to avoid each other, then satellite comms or perhaps traditional telecoms coverage if close enough to land to communicate with the outside world.

“Data will flow differently depending on where the ship is – satellite on the high seas but may take advantage of cellular communications technology close to shore,” says Rolls-Royce’s Levander.  “Of course, when the ship involved is autonomous or remotely operated, getting data to shore in a timely manner is vital; such ships will require constant real-time communications links.”

Quite what the answer is to these questions, no one knows. Nano-satellites could be one option, as could internet-providing drones such as Facebook’s Project Aquila or Google’s balloon-based Project Loon, or perhaps some sort of 5G or IoT-centric mesh network.

“The connectivity needs are biggest at the beginning; when we have remote control going on,” suggests DIMECC’s Dr. Tavitie. “But when we get rid of remote control and we get to full autonomy, the connectivity requirements go down to almost zero. Meanwhile, it will be an issue; I don't see any kind of feasible developments going on in the connectivity field that would solve the whole thing just like that.”


Also read:
Whether by pod or autobahn, driverless cars are coming
Facebook drones: 'There are no rules for this.'


«Behind the scenes at the Tour de France


Banks ripe for the next cyberattack»
Dan Swinhoe

Dan is Senior Staff Writer at IDG Connect. Writes about all manner of tech from driverless cars, AI, and Green IT to Cloudy stuff, security, and IoT. Dislikes autoplay ads/videos and garbage written about 'milliennials'.  

  • twt
  • twt
  • Mail

Our Case Studies

IDG Connect delivers full creative solutions to meet all your demand generatlon needs. These cover the full scope of options, from customized content and lead delivery through to fully integrated campaigns.


Our Marketing Research

Our in-house analyst and editorial team create a range of insights for the global marketing community. These look at IT buying preferences, the latest soclal media trends and other zeitgeist topics.



Should the government regulate Artificial Intelligence?