Driverless cars, crewless ships, pilotless planes and drones have all quite abruptly thrust themselves into the insurance industry’s collective conscience over the past year. The possibilities have been discussed for some years but quietly filed away in the “Sometime, Never" trays. Now those possibilities are becoming realities and fast rising to the top of the “This Year, Next Year” trays.
As they do, it is not just the obvious underwriters of motor, marine, aviation and transport risks that are having to take notice, but product liability and cyber risk underwriters too. All face a decade or more of unprecedented change. Opinion may vary on the pace of that change but no one doubts that it is coming fast.
“The arrival of driverless cars on our roads probably won’t happen quite as quickly as the manufacturers and producers are projecting but it is definitely going to happen,” says Craig Dickson, Chief Executive Officer of DAC Beachcroft’s Claims Solutions Group. “Google is predicting 2020 but that seems a little fanciful.”
The Google driverless car has already clocked up over 700,000 miles on American roads. But Dickson points out it is not just about the technology in the cars but the road infrastructure and a wide range of cultural considerations.
“A complete shift is unlikely in less than 20 years but we will see a range of early adopters. These will include large manufacturing and petrochemical plants as well as concept cities that will be purposely designed so they don’t have any non-autonomous vehicles. This could easily be somewhere like Milton Keynes, which lends itself perfectly to the technology with clear lines, wide roads and a good cable infrastructure.”
Milton Keynes is one of four locations for the Department for Transport sponsored studies into driverless cars, along with Bristol, Coventry and Greenwich. These involve manufacturers, safety experts, academics and a wide range of transport interests as well as legal and insurance experts.
Early hesitation
Dickson says that the acceptance of autonomous vehicles will take time, but will probably happen faster than people think as people’s understandable early hesitation about letting technology take control of their journeys rapidly recedes. The initial doubts over driverless trains when they were first deployed on the Docklands Light Railway quickly vanished and now people hardly question whether they are getting on a train or tram that doesn’t have a driver.
On the seas the pace of technological development is similar, but the readiness of entrenched interests in the maritime industry to accept change might prove an inhibitor of the deployment of crewless ships.
As with the aerial drones, the American military machine is one of the key driving forces behind the development of drone vessels. It is thought to be close to deploying them for a range of routine patrolling and detection functions, such as sweeping the vast spaces of the Pacific and Atlantic oceans for submarines.
Commercial shipping
The crossover to civilian use has already started with Rolls-Royce leading the way in an EU-sponsored project known as MUNIN – Maritime Unmanned Navigation through Intelligence in Networks. Its development timeline suggests that crewless cable ferries and coastal cargo ships could be at sea by the early 2020s with semi-autonomous ocean-going cargo ships following later in the decade.
“We could see small cargo ships operating in areas like the Baltic Sea where the climate is harsh but the risks very low within a decade,” says Toby Vallance, an Associate specialising in marine, aviation and transport at DAC Beachcroft.
“With commercial shipping it is all about tight margins. The attraction of having a ship with no accommodation block and therefore more space for cargo will be an important factor. Culturally, however, it might seem a long way off as the various classification societies, maritime regulatory bodies and unions are not yet fully aware of the potential.”
To a certain extent it will happen by stealth: “Some vessels are so big now that in order to control them in changing sea conditions you already need powerful computers to run the engines, the stabilisers and so on,” adds Vallance.
Of course, automated submersible technology has been used extensively for some time now, especially in the offshore energy sector, assisting in safety checks, maintenance and repair of subsea equipment and infrastructure.
“Underwriters need to start taking more interest in how quickly these new technologies can be deployed,” says Vallance.
The development of aerial drones is following a similar, if more advanced, course. The military use of unmanned bombers is already extensive, but the transfer to civilian uses is being held back by a mixture of cultural and legal concerns. As with cars and ships there are early adopters. Again in the offshore energy sphere, oil and gas rigs, which need regular safety surveys, now use aerial drones to carry out visual inspections that were previously highly hazardous for human beings to do.
In America trials are being carried out to test the potential of drones to carry out a variety of functions, including realising Amazon’s dream of delivering customer orders by drone, delivering medicines to predetermined collection points and monitoring traffic or antisocial behaviour.
Potential benefits
There are huge potential benefits in taking control of cars, ships and planes out of the hands of humans. About 90% of accidents on roads are caused by driver error according to the Thatcham Research Centre, and similar figures are frequently quoted in the marine and aviation sectors. On top of that there is the potential for black swan incidents such as the Lufthansa/Germanwings crash and Costa Concordia sinking, both entirely due to human decision-making.
On the roads, the deployment of semi-autonomous technologies such as automated front/rear braking, lane drift sensors and automated parallel parking are already having an impact on accident rates and severity.
“All these technologies, combined with telematics data and dashboard cameras, have huge potential to quickly reduce the frequency of low-velocity accidents, which will have a major impact on whiplash claims,” says Dickson. “What the industry needs to do is to get all that into the claims investigation process so that it can defend itself against fraudulent claims but also pay legitimate claims faster.”
Safety should also be a big motivator for the adoption of technology in the marine sector, says Vallance. “Crowded ports and especially vessel loading and unloading are particularly hazardous. Just having fewer people at sea reduces the risk to marine insurers considerably. There is also an emerging skills shortage. Spending long periods at sea is less attractive than it used to be and as the technology has moved on, the quality of crew training has not kept pace.” He points to the increase in groundings as proof of the extent of this problem, one that would vanish with the removal of crews.
Product liability
The increasing range of highly sophisticated technology being deployed is pushing product liability into the spotlight. The picture here is complex and fraught with potential dangers for anyone who makes glib assumptions about the extent of cover and the ease of recoverability, warns Stephen Turner, a Legal Director at DAC Beachcroft.
"At the top of the tree are the manufacturers, who should be constantly reviewing their exposures as they develop autonomous vehicles,” he says. “Alongside them are the component suppliers, many of whom may not have previously supplied the motor, aviation or marine sectors. Then there are the various delivery supply chains.
“Typical product liability policies have aviation and watercraft exclusions as well as motor-related exclusions because the market has traditionally regarded them as matters for the marine, aviation and motor markets. New entrants to the market will have to review their cover carefully as they should not assume cover is included in their existing policies. Product recall can also be subject to similar exclusions.”
The potential for drone technology to open up new areas of risk extends to the supply chain, especially when considering financial loss. “If any goods being delivered using automated transport systems are delayed, damaged or delivered to the wrong person, the already complicated debate about where liability for that falls will become even more difficult,” warns Turner.
Key developments
International conventions bar the way to driverless cars
The German transport minister, Alexander Dobrindt has raised the prospect of international agreement being needed before fully autonomous cars become a reality. There is a series of international conventions that insist all vehicles have drivers. The key one is the Convention on Road Traffic, commonly known as the Vienna Convention on Road Traffic, a United Nations document ratified by 70 countries in 1968. The most prominent missing signatory is China, although the UK apparently omitted formally to sign the last revision.
Clause 1 of article 8 of the convention says: “Every moving vehicle or combination of vehicles shall have a driver.” This is reinforced by clause 5: “Every driver shall at all times be able to control his vehicle or to guide his animals.”
Crewless ships will need progressive nations to break regulatory barriers
There are strict rules about the ratio of crew required according to the size of ship, which are decided by the International Maritime Organisation and administered by the various national classification societies. Rolls-Royce says that it expects it to take until the later part of the next decade for worldwide agreement on new rules governing drone ships to emerge. However, it predicts that more forward-looking countries, such as the USA, could move faster and allow crewless vessels in their own coastal waters much sooner.
Aviation regulators set regulatory review timetable
Much of the future development in commercial use of drones lies in the hands of aviation regulators and they are working closely to create a roadmap for harmonising the rules, led by the EU and the Joint Authorities for Rulemaking on Unmanned Systems (JARUS). The current roadmap is looking at harmonisation for the under-20 kilogram unmanned aerial vehicles (UAVs) flying in line of sight by 2018, with certain ‘beyond-line-of-sight’ operations allowed in sparsely populated areas. This is already permitted in the UK around Aberporth Airport in Wales and over parts of Salisbury Plain.
This will be followed between 2019 and 2023 with agreement on rules allowing larger UAVs to operate in some types of airspace and limited beyond-line-of-sight operations in populated areas.
In the USA, the Federal Aviation Authority already authorises UAVs for use by public authorities and emergency services. It also has six test sites up and running as it rushes to comply with Congressional demands to produce comprehensive regulations for licensing the commercial use of UAVs by the end of September 2015.
Cyber threat
One of the first new hazards foreseen by all the experts is the cyber risk. “The cyber risk at sea is really an extension of the piracy risk we already have to deal with,” says Vallance. “Piracy has made some parts of the world very dangerous. It is possible to see the attraction of a cyber attack on a crewless vessel to terrorists, criminals and a range of activists. It will be about creating the security of a robust backup. It will be one of the areas that will need most testing during the pilots.”
Security of networks will have to be a top priority with the extension of telematics and the development of autonomous cars. “A significant investment will have to be made by vehicle manufacturers and component suppliers to protect against a range of cyber threats,” says Dickson. “The security of the data networks will be especially important as roads communicate with vehicles and vice versa. I expect vehicles to have a fail-safe mode they switch into if they think their systems have been compromised.”
For cars, the external infrastructure that supports them will also have to be assessed and insured: “So far all the focus seems to have been about the onboard technology but it is pretty clear that the internet of things and the whole connected infrastructure will be important too,” says Dickson. “Underwriters will have to make clear distinctions between the technology in the vehicles and the technology in the infrastructure.”
“The cyber risk really brings this whole area into focus,” says Turner. “There is plenty of scope for failure or vulnerability that will have to be covered. It could be through disruption, extortion or delay. The key question will be where do we draw the line between different covers – cyber and liability. There are parallels and precedents with the control systems for large manufacturing systems so we have some knowledge of how to deal with this. Scale will be the new unknown.”
It will be one of many unknowns as this huge, sweeping, all-embracing revolution in transport gathers pace over the next decade.
