Autonomous ships - Fact or fiction?

Autonomous cars have now been on everyone's lips for a few years – but autonomous ships?

Autonomous ships are a new frontier

Driverless car technology is at an advanced stage, with major companies such as Google, Tesla, Ford, General Motors and Nissan paving the way. For quite some time, Volvo has been testing driverless cars on public highways in Gothenburg, in conjunction with the Swedish authorities.

In the field of transportation, autonomous (or unmanned) ships have been and still are being intensely debated. It is virtually on the agenda of almost all transport related seminars. As with all technical developments, there will inevitably be ascribing risks. From an insurer's perspective, it is new territory and these new risks difficult to assess and grasp.

Autonomous ships are a new frontier for the insurance and legal communities. Many questions remain unanswered. For instance, if there is no shipmaster on board, who can be held liable if something goes wrong? Will autonomous ships become a reality or remain fiction?

Fact or fiction?

Let us start off on the right foot. Autonomous vehicles are already state-of-the-art, in many land-based transport modes. Several examples of automated subways, self-driving intralogistics vehicles or automated guided vehicles at modern container terminals already exist.

There are also very wide-ranging approaches to autonomous control concepts in modern aviation. Autonomy is seen as an opportunity for maritime transport to meet today's and tomorrow's competitiveness, safety and sustainability challenges.

One could, therefore, argue that the technology is in place. However, there are still a number of nuts to crack, before the world will see autonomous ships on any larger scale. Some of the challenges involve infrastructure: ports, fairways and cargo handling etc. Other challenges are of a political, geographical and regulatory in nature.

That said, the continued drive for extracting cost efficiencies and increasing competitiveness within transportation and the supply chain continues to be a powerful driver for making this a reality.


Autonomous ships will transform the entire marine business. The operational profile of ships will change, as more tasks and functions become controlled from shore centres. Remote and autonomous solutions will enable entirely new ship designs and concepts for improved economic performance.

Politically, logical reasoning for encouraging autonomous ships ploughing the seas and waterways includes, e.g. that the EU road network is suffering from congestion. Population growth in urban areas will lead to a demand for transportation that exceeds the capacity of existing roads. To alleviate these issues, governments all over the EU are trying to move some of the freight volume from the roads to the waterways.

To this end, the Europe Commission funded the three-year MUNIN (Maritime Unmanned Navigation through Intelligence in Networks) research project, to investigate the possibilities of unmanned ships.

MUNIN, which was completed in August 2015, used ten years of global manned ship data to compare the risks of manned ships to those of unmanned ships. The study projected that an unmanned ship would have one-tenth of the risk of a manned ship of colliding and foundering, where human error often plays a role.

The analysis also predicted a savings of $7 million over a 25-year period, per ship, in fuel use, crew supplies and salaries. With the number of cargo ships operating worldwide, currently estimated at 9,600, the potential savings are enormous.

A lot of time and effort

Some companies that have invested a lot of time and effort into developing concepts for autonomous ships include the Classifications Society DNV GL and Rolls-Royce. DNV GL has announced its concept for an unmanned, zero-emission, shortsea vessel and has built a 1:20 scaled model. Automated Ships Ltd and Kongsberg Maritime have signed a memorandum of understanding to build the world's first unmanned, fully automated vessel for offshore operations. Sea trials will take place and the vessel will ultimately be classed and flagged.

There are many technical challenges. Various sensor systems will be needed to detect problematic situations, such as unexpected objects in the sea, dangerous weather conditions or collision dangers. A major challenge will be to devise sensor systems that reliably detect all relevant dangerous situations and appropriately act on them.

Legal challenges

New strict, international and uniform technical safety standardization will be absolutely essential for the introduction of an unmanned merchant service. This needs to be implemented in the context of the SOLAS Convention and the rules of Classification Societies. The IMO will be scrutinizing the development of such a framework.

New safety standardization will be absolutely essential.

Can existing maritime law still be applied, in principle? What existing rules will require amendment or expansion and what new rules might have to be developed?

Maritime law is a distinct body of both domestic law, governing maritime activities and private international law governing the relationships between companies that operate ships on the oceans of the world and the tidal waters of rivers.

It deals with all matters, including marine commerce, international trade, shipping, marine navigation and the transportation of passengers and goods by sea. Maritime law also covers many commercial activities, which although land based, are maritime in character.

How will current international regulations deal with unmanned ships

How will current international regulations deal with unmanned ships, including the changes required to conventions such as SOLAS, MARPOL, STCW and COLREGS? What is the potential impact of autonomous ships and the carriers' duties to cargo, under the carriage of goods regulations, such as The Hague-Visby or the new Rotterdam Rules? Will autonomous ships have an impact on the carriers' right to limit their liability? Threats, such as cyber piracy, will spur whether further regulations are required.

As for the UN Convention on the Law of the Sea (UNCLOS), it is quite clear that an autonomous ship will have to fly the flag of a state. But is the autonomous ship, with no master or crew on board, still a ship?

Most commentators accept that, for the purposes of the law of the sea, unmanned vessels must be regarded as ships. Other national laws simply define the master as any person to whom the authority of the ship is transferred or as the person who effectively exercises that authority.

It can, however, be concluded, with a considerable degree of certainty, that unmanned ships would be covered by the great majority of the existing regulatory definitions, conventions and national laws.

The duty of the carrier

The duty of the carrier to apply due diligence in making the ship seaworthy, which is a fundamental element of The Hague-Visby Rules and the Rotterdam Rules, will, in the light of the available case law about the general state of the ship, imply that the guiding IT mechanism and the shore-based vessel controller must function satisfactorily.

Collision law also appears to be able to stand up well against the arrival of autonomous ships. The 1910 Collision Convention governs the liability for collisions on the basis of the errors of ships and not the errors of the master and the other crew members (even though these are, of course, the cause of the collision). Should the collision be caused by the error of a shore-based vessel controller nothing will change regarding liability.

As for contracts of affreightment (usually a charter party), which determines how the ship is operated, there is nothing to stop the various types of charter parties continuing to play a role. The fact that ships will, in the future, perform their tasks without anybody on board does not appear to have an essential impact on these contracts. Some amendments need to be done, especially regarding competences, recruitment and the functioning of the master.

As for the rules of the special conventions relating to pollution damage, they appear to remain entirely unaffected.

To the extent that the unmanned ship is operated or controlled from the shore, the question arises whether or not the shorebased vessel controller can, under the current state of maritime law, be regarded as the master?

The shore-based vessel controller, sitting at his control desk, will have the responsibility of handling transports carrying valuable cargo. Just like the master, the shore-based vessel controller must have certain qualities, such as good judgement, the ability to communicate well, a cool head in emergencies. The shore-based vessel controller will have the duty to maintain a proper look-out and to proceed at a safe speed.

Relatively little will have to change?

All things considered, it appears that, with regard to the liability of the ship owner, relatively little will have to change. But, the introduction of fundamentally new liability rules may become the subject of political discussions. Would it be preferable to have liability based on strict liability for ships controlled by artificial intelligence?

Maritime law, with its long history, appears, on first examination, to be relatively well armed for this technological innovation and the necessary and undoubtedly extensive adaptations of existing public and private maritime law will be unlikely to bring about a revolution.

Maritime law will enter a new phase of development; but, it will certainly not die out.

Insurance – risky business?

At first sight, autonomous transport would eliminate the human factor, which, according to estimates, accounts for roughly 80 per cent of all maritime casualties.

However, to the extent that unmanned ships are operated or guided by human operators, the human factor will continue to play a role. The shore-based vessel controller will be unable to react with the same intuitive feel for the situation (or at the very least a good deal less).

Safe autonomous operation will ultimately depend on the satisfactory operation of the on-board equipment, the required connections with shore stations and the stability of the computer programs. There will be communications and cybersecurity considerations.

What risks will follow an autonomous ship being isolated on the ocean? How will the duties of the crew, to ensure the cargo is inspected, transfer to an autonomous ship? Who will fight a fire that breaks out?

A need for a backup plan

There would need to be a credible backup plan, if something goes wrong at sea. The risk is paramount, as there would be millions of dollars of cargo on board, on larger merchant autonomous ships. Who would be there, if there was a breakdown of the computer systems?

Will an unmanned autonomous vessel at sea be at higher risk of piracy or will human presence on board, with active piracy measures in place, be a more effective deterrent to a pirate boarding? It would be naive to expect that pirates and terrorists will disappear from the high seas.

Could an autonomous ship, whose controlling software has been hacked from the shore or another ship, be regarded as a pirate ship? Whether or not the hacker of an IT system controlling an unmanned vessel would be considered a pirate in the meaning of the UNCLOS, is not so clear.

Will an unmanned autonomous vessel at sea be at higher risk of piracy?

Interaction between autonomous and manned ships

What will happen if a fully autonomous vessel is unable to avoid a collision with a traditional vessel and is rendered holed, grounded or otherwise unresponsive and uncommunicative?

The interaction between autonomous ships and large manned ships can probably be solved by route exchange and coordination. However, a large unknown remains in how to deal with the interaction between smaller vessels, such as leisure craft, coastal fishers and, for example, kayaks in inshore areas.

Must the shore-based vessel controller be protected against unlimited personal liability? Would there be reason to make software and hardware developers, system engineers and manufacturers subject to a special product liability?

If there is no longer any crew required on vessels, what will happen to the pool of available experience, traditionally obtained from individuals with prior seagoing experience that is essential for shore-based legal and insurance companies and brokers?

The list of potential policy and risk questions is almost endless. At present, it is still too early to try to answer all of them.

If we can put a rover on Mars and have it autonomously conduct research, why can we not sail unmanned vessels across the Oceans? The dawn of the autonomous ship is upon us.

Article by

Tony Schröder

Manager of Major, Complex and International Property Claims Team, If

Article published in Risk Consulting Magazine 1/2017