Jean-Yves Gilg

Editor, Solicitors Journal

The rise of the autonomous car

The rise of the autonomous car


Following the first fatalities involving driverless vehicles in the US, David Thompson and Dorothy Agnew consider the legal questions raised by this fast-growing technology

Following the first fatalities involving driverless vehicles in the US,
David Thompson and Dorothy Agnew consider the legal
questions raised by this fast-growing technology

By 2025, the driverless technology industry is expected to be worth a staggering £900bn globally and is currently growing by 16 per cent a year.

Now autonomous vehicles are coming to Britain, and a government consultation on changing the law to make the UK driverless-car friendly is currently under discussion. While there has been much heated debate about driverless technology, many legal questions remain unanswered. Who, for example, is ultimately responsible in the case of a collision? The driver? The insurance company? The manufacturer or the MOT mechanic? And will this vary by vehicle type?

Already, in the US, there have been two fatalities involving driverless vehicles, and these incidents have highlighted the complexity of the subject area.

Driverless technology

Autonomous technology has existed for many decades in some form or another, but perhaps the first real taste of what was to come was born with the invention of the self-propelled torpedo by Robert Whitehead in the 1860s and the development of sonar for guidance. The 1940s saw the evolution of self-guiding in the air, once again for military purposes, with the V-1 flying bomb and, after that, the V-2 rocket.

Unsurprisingly, it has been the sea and air where autonomous 'driving' has been most widely adopted as - stating the obvious - both the skies and seas provide relatively direct passage, are highly regulated, and are free from many of the issues that road users face.

A self-guiding car on the road was another matter altogether. Developing a driverless car that was both safe and practical meant the development of special roads which could be used to guide the cars along them, and the early focus was on the infrastructure, rather than the concept, of having an autonomous or 'robotic' car. This infrastructure was first showcased by US designer and futurist Norman Bel Geddes with his Futurama ride for General Motors at the New York World's Fair in 1939. The birth of the digital computer in the 1960s enabled a new era of autonomous vehicles, albeit once again born out of military application. The decade saw the focus change to the vehicle and its robotic capacity, rather than the road or 'freeway'.

In 2004, the US Defense Advanced Research Projects Agency challenged a number of teams working on autonomous vehicles to compete for a $1m prize. This was the first in a series of challenges which accelerated the development of the 'smart' car.

Over the years, the computer systems that make autonomous cars work have become smaller and much more sophisticated, turning from huge installations into small, discrete units. But it was not until the last decade or so that the technology evolved to such an extent that cars and commercial vehicles could 'sense' and 'react' to their environment. But as the accidents in the US have highlighted, real issues still remain.

Today, self-driving vehicles can be found in farmers' fields in the UK in the form of automated combine harvesters, and in Australian pits as mining vehicles or specialised vehicles for the purpose of serving warehouses and factories.

Despite BMW, Volvo, Google, and many other vehicle manufacturers all introducing an increasing number of self-driving features on high-end conventional cars, we have not yet seen the first driverless car in the UK. But it won't be long. Already, in the US, the testing of driverless cars on roads has been permitted in the states of Nevada, Florida, California, and Michigan since December 2013.

Programme of reform

So what's next? How will our laws need to be changed to accommodate the actions of an autonomous machine, and who will be responsible when things go wrong?

A programme of reform to prepare the roads for driverless technology is currently being launched, alongside a major government consultation, which demonstrates a huge step towards the UK embracing the driverless car. The consultation, published in July 2016, is a joint programme by the Department for Transport (DfT) and the Centre for Connected and Autonomous Vehicles (CCAV). Under the new proposed measures, rules will change regarding such issues as the insuring of automated vehicles. In addition, regulations making up the Highway Code will be amended to accommodate the launch of automated vehicles.

A regulatory review was completed in February 2015 and demonstrated that testing automated vehicle technology in the UK was already possible. Unlike in other countries, automated vehicles can be tested on any road in the UK without the need to seek permission from a network operator, report any data to a central authority, or put up a surety bond.

The next step was to publish a code of practice which clearly and simply set out how testers must comply with UK law, such as obeying relevant road traffic laws and ensuring test vehicles are roadworthy, a suitably trained driver or operator is ready, able, and willing to take control if necessary, and appropriate insurance is in place. Importantly, proposals also include helping to fund research and development. In February 2016, the DfT and CCAV announced the winners of the first £20m competition from the £100m Intelligent Mobility Fund, which is being match-funded by >> >> the automotive industry to help facilitate the development of new automated vehicle technologies. They will be launching the next competition, worth up to £30m, in August.

Question of liability

Another major step to embracing driverless technology is to start tackling the domestic regulatory issues that could prevent UK citizens and businesses from taking advantage of safe and approved driverless cars as they come to market.

The legal challenges are numerous, and we currently do not have the right solutions for what is effectively unknown future technology. While too much regulation at an early stage could stifle progress, equally we need clarity before the inevitable accidents and, without wishing to seem the purveyor of doom, fatalities occur.Ultimately, the sector with the most to lose - or potentially gain - is insurance, and the Association of British Insurers (ABI) has been very vocal. In response to the government's proposals, James Dalton, the director of general insurance policy at the ABI, confirmed that 'the insurance industry is working with government, vehicle manufacturers, regulators, the legal community and through the industry's research and repair centre on this potentially life-changing and life-saving technology'.

Interestingly, the ABI clearly supports the use of autonomous vehicles, stating that human error accounts for around 90 per cent of road accidents and that the potential safety implications of autonomous technology are significant.

If human error were removed, then in due course there would be significant reductions in all types of accidents, whether resulting in injury or not, leaving liability arguments in only a small number of cases. In short, what the Legal Aid, Sentencing and Punishment of Offenders Act has not fully achieved in terms of whiplash claims, the driverless car could.

Autonomous emergency braking, for example, where brakes are applied automatically if the driver does not respond in time, has been proven to lower the rate of collisions that result in personal injury claims by around 20 per cent. It is therefore unsurprising that many insurers are already enthusiastic about the new automated technology.From the liability and insurance perspective, we must consider the degree of automation available in the specific vehicle. For example, there is a world of difference between the Tesla examples, where the driver is still in charge, and the proposed Google vehicle, in which the driver may play no part. With vehicles that are fully independent of driver involvement, manufacturers and sellers are likely to have to accept liability for failures leading to accidents.

Tesla makes it clear that the driver needs to be able to assume full control, and it therefore appears that the driver is not absolved of responsibility should there be an accident.

However, this raises questions about the advantages of using a fully automated vehicle. What if, for example, the same level of control is needed as when one engages cruise control, i.e. you have to be physically connected to the vehicle's controls and capable of taking immediate emergency action?

If this is not the case, what will be the acceptable timeframes for action? Will 'black box' style monitoring be required?

Cybersecurity issues

There are other legal issues that need to be addressed, including data protection and cybersecurity. Telematics, the use of technology to gather and analyse information about the operation of a vehicle, has been widely adopted on a voluntary basis to help manage insurance premiums. However, such technology might need to be mandatory to monitor vehicles, and possibly to establish liability in the event of an accident. This raises questions relating to data and privacy laws.

The ABI has issued guidance to help promote compliance with the Data Protection Act 1998 among insurers regarding telematics. The guidance emphasises that consumers are given clear and comprehensive information to ensure that they fully understand how their personal telematics data will be collected. The issue is that such guidance only apples to insurers, so will similar rules need to be applied to vehicle manufacturers?

As all readers will know, technology is fallible. Cybercrime - whether for financial gain or out of malicious intent - is one of the fastest growing areas of fraud. Sadly, in Nice, we recently witnessed a horrendous act of terrorism where a vehicle was used as a weapon, and in such a situation liability is unlikely to ever exist. However, if a similar event happened in the future where an individual or multiple driverless system were hacked, would liability also not exist?

Another area which is rarely debated is that of the 'moral behaviour' of a driverless vehicle. Consider a scenario where a child steps out in front of a vehicle and a collision is unavoidable unless the vehicle swerves and thus potentially hits a pedestrian on the pavement.

Legally, we have many unanswered questions, and, in reality, many will remain unresolved
until the scenario actually occurs. The legal profession must be actively involved in the
debate, raising the potential issues before
they occur.

David Thompson, pictured, is the managing partner and Dorothy Agnew a partner at Moore Blatch @MooreBlatch