Driverless cars Driving redefined | Risk info | Large | Business | Zurich Insurance

Driverless cars: Driving redefined

Cars

Driverless cars are not a distant theoretical concept, they are on our roads today. In the UK, the Meridian Shuttle in Greenwich hit the streets at the beginning of the year and trials have also been approved elsewhere around the country in Bristol, Milton Keynes and Coventry.

Google launches driverless car

2012

Google driverless vehicles clock up over 1 million miles

2015

BMW, Nissan and Volvo announce they intend to bring driverless cars to the market

by 2020

Frost and Sullivan predict there will be 1 million driverless cars in the US

by 2020

Volvo predicts no-one should be killed or seriously injured in a new Volvo

by 2020

IHS predicts there will be 11.8 million driverless cars sold globally

by 2020

IHS also predicts there will be 54 million driverless cars on the world's roads

by 2035

Apple are rumoured to be further along than expected with their self driving car 'Project Titan'

Future developments

In the US, Google is going for the so-called ‘Big Bang’, with its driverless vehicles having clocked up more than 1 million miles since their launch in 2012. It is also rumoured that Apple have booked time at a secret test facility in California and will have a game changing product to market very soon.

Car manufacturers are alert to the potential upside of this new technology and are spending millions not just on research and development, but on establishing themselves as credible producers of driverless cars in order to gain profitable market share over the next few years. BMW, Nissan and Volvo have all announced that they intend to sell driverless cars by 2020.

The projections for the future of the industry are startling. Consultancy firm Frost & Sullivan predicts that there will be more than one million driverless cars on the road in the US by 2020, and analytics provider IHS predicts there will be nearly 11.8 million sold globally in 2035, taking the tot al number on the roads to 54 million.

Benefiting society

Driverless cars present an attractive proposition for a number of reasons. A report published by PwC estimates that there are approximately 1.3 million motor fatalities every year globally and some 40 million casualties. Analytics companies, insurers, manufacturers and government bodies seem to share a consensus that at least 90% of car accidents are caused by driver error. Removing the principal cause will have a positive effect.

Volvo states that “By 2020 no-one should be killed or seriously injured in a new Volvo.” Driverless cars can also have a positive environmental and social impact by reducing emissions and easing congestion. Vehicle to vehicle technology (V2V), which involves vehicles communicating with each other, and vehicle to infrastructure (V2I), could allow cars to drive more efficiently by knowing how fast other cars are driving and where they intend to go, and find less congested routes.

Removing the need to manually operate a vehicle could also allow those who would not normally be able to use cars – such as disabled and elderly people – to drive. YouTube has a video clip of Steve Mahan, who is 95% blind, using a Google driverless car in 2012 to go to a drive-through restaurant, a dry-cleaner, and home again, without incident.

Finally, driverless cars could create more free time for the operator of the vehicle. The Department for Transport (DfT) estimates that the average driver spends 235 hours driving per year. According to the DfT, driverless cars could free the driver up to ‘read a book, surf the web, watch a film’.

Impact on drivers

It is likely to be at least a decade before vehicles requiring no driver begin to appear on our roads. Until then all vehicles will require a driver to some extent and “driverless” will mean able to function autonomously in given situations ie Lane Assist. Increasingly sophisticated Advanced Driver Assistance Systems (ADAS) will lead the way towards fully autonomous vehicles but the intervening “hybrid” environment will create challenges for all drivers, transport planners, businesses and legislators.

The common law duty of care on a driver is unlikely to change but the test of “reasonableness” will be applied to new circumstances as technology evolves and drivers rely on it for an increasing number of activities. A vehicle with “driverless” capability will either be in fully manual mode, driverless mode, or in the process of transitioning between the two automatically or at the driver’s command.

There is scope for breach of duty, misuse of the driverless capability, product malfunction, conflict of priorities or simple confusion on the part of the driver or other road users. Consumer confidence in driverless vehicles is key, as Volvo’s recent commitment to accept responsibility shows, but in the UK access to justice is also a matter of public policy.

Therefore it is likely that legislative change will be implemented to achieve greater certainty than the present combination of common law duties of care and product liability legislation already provides. What that change may be remains to be seen, but driverless cars may bring radical change to our legal framework for drivers just as they are set to bring radical change to society.

Impact on insurance

The increasing use of driverless cars has several implications for the insurance industry. It is likely that the level of risk will be reduced significantly, as accidents caused by human error decline. In terms of insurance policies, there is likely to be a shift away from traditional motor policies in favour of product liability policies, as the role and responsibility of the driver gradually diminishes and the technology takes over.

Fewer accidents could lead to a reduction in losses and premiums in the traditional motor insurance market, together with a shrinking in the market itself, as product liability policies play an increasing role. If we reach the stage where road accidents are negligible, a very different insurance landscape could be unveiled. Risk would no longer be spread to the same extent amongst millions of individual drivers, but instead could be aggregated against a much smaller number of car producers and other stakeholders.

Claims

With the advent of automation, the landscape for insurance claims could be altered considerably as the burden of responsibility shifts away from the driver and towards the technology. At present, there are 3 types of product liability claims that could arise against producers of driverless vehicles: manufacturing defects, design defects and failure to warn.


  • There may be a manufacturing defect when, for example, a poorly manufactured sensor develops a fault and fails to operate in the way for which it was created, resulting in an accident

  • A design defect may occur where a sensor fails in certain weather conditions – perhaps above a certain level of humidity – in circumstances where they were expected to perform

  • Failure to warn is the most complicated of these, as it brings into play the division of responsibility between driver and technology. For instance, if a safety feature in a car fails to warn the driver of an impending risk, liability may not simply sit with the manufacturer. With such features, there is the risk that the driver may not properly understand the extent to which he can rely on the vehicle to handle an emergency situation

Legal implications

Several grey areas in law accompany automated vehicles. If driverless cars become heavily reliant on receiving up to date satellite and traffic data, as well as information from other vehicles, what happens when the car is unable to obtain this data? For example, what if a roundabout is constructed overnight and the driverless car fails to download this information, and then drives over the roundabout the following day?

Who is liable – the vehicle operator, the vehicle manufacturer, the software provider, the wireless network provider? And will older vehicles struggle to keep up with software updates, in the same way that older mobile phone models slow down when updated to new software versions designed for newer phones?

Consider also a situation where a pedestrian suddenly and unexpectedly runs out into the road. If the vehicle cannot brake in time it has two options; to swerve to avoid the pedestrian or carry on driving. In scenario one, the car may swerve into oncoming traffic and injure its occupant; in scenario two the car injures or kills the pedestrian.

And what dictates what course of action should the driverless car take? Does it make a decision based on which course of action causes the least loss? Does it calculate how many people are in the car? Or does it simply shut down and issue you with a warning that you need to take over? What if the operator does not have time to take the controls, assess the situation and make a decision? Who will be at fault?

Driverless cars are not a futuristic flight of fancy, they are on the roads now and every year will become more visible as an everyday feature of the road. As a new technology, they will create new opportunities and challenges, and it remains to be seen how the insurance and legal services industries – and society as a whole – adapts.

When I look at the automobile, what I see is that software becomes an increasingly important component of the car of the future.

Tim Cook - Apple CEO