A time of great change for the UK’s transport network it is not far off, but will we be ready for the practical and legal implications?
Opinions about timing vary, but global analytics provider IHS has forecast that 230,000 autonomous vehicles will be on the road worldwide by 2025 and over 11 million by 2035, 20% of the vehicles will be in Western Europe. Add to this the UK Government’s desire to be at the forefront of technological developments with driverless vehicles, with the economic and social benefits offered by their arrival, and it will be an exciting and challenging decade or two ahead for road users, and those in the automotive and transport industries responsible for safe and effective implementation.
In February this year the Department of Transport published “The Pathway to Driverless Cars” and gave the go-ahead, with £19 million of financial support, for testing of driverless vehicles in four locations (Bristol, Greenwich, Coventry and Milton Keynes) without requiring legislative changes. Instead the test vehicles must observe a Code of Practice which is expected to be published imminently, as the current law can accommodate this.
The pertinent questions in future are going to be when and in what respects does a driver – if there is one – owe a duty of care to others, and what duties of care are in turn owed to him by others on whom he relies in an aged of advanced technology? And where will the chain of liabilities take us?
To better understand the potential of driverless cars, and the practical, legal and insurance issues, one must first distinguish between “partially/highly” automated and “fully” automated vehicles.
The levels of automation
Both “partially/highly” automated and “fully” automated vehicles will be capable of operating without human intervention in defined circumstances, the partially or highly automated vehicles must have a driver able and ready to assume control if needed. Only fully autonomous vehicles will be truly “driverless” and any occupant will be a just a passenger, assuming there is an occupant.
A useful guide to the different levels of automation has been published by SAE International (see: http://www.sae.org/misc/pdfs/automated_driving.pdf ); it outlines six levels, where zero represents a fully manual vehicle and Level 5 a fully automated vehicle. Some cars on the road today are at Level 2, with the driver monitoring the dynamic environment at all times but with on-board systems such as emergency assisted braking, lane warning and assisted parking that can play an active role. Over time Levels 3 and 4 will gradually reduce the involvement of the driver until Level 5 is achieved.
The ability of partially or fully automated vehicles to use the road network safely is crucial to their successful introduction. Realistically, a staged approach seems inevitable; motorways will be safer than narrow country lanes and city centres better than rural villages to adapt. In future, major town and city road networks may have to be redesigned on an automated grid system if road based public transport systems are to function effectively, but as readers will be aware, the concept of public transport is also likely to evolve significantly. It was reported earlier this year that Ligier Group and Robosoft have created a driverless 10-passenger vehicle which learns its route and, by means of sensors, will adapt its speed to pedestrians and other vehicles around it.
Since it will be impossible for driverless cars to develop and perform in isolation from other technologies, the wider issue is that of “connectivity” in general. Thus “smart roads” and “smart infrastructure” are of equal significance.
In the United States vehicle to vehicle (V2V) communication is seen by the National Highway Safety Traffic Administration as a key issue. Vehicle to infrastructure integration (VII) is the subject of extensive research by some major motor manufacturers (Ford, General Motors, DaimlerChrysler, Toyota, Nissan, Honda, Volkswagen, BMW), with IT suppliers and U.S. Federal/State transportation departments, through the Vehicle Infrastructure Integration Consortium (ViiC).
In the UK, the recently announced testing in Bristol (“VENTURER”) will develop the technology for driverless vehicles to communicate with smart city infrastructure and one of the consortium partners is The University of Bristol’s CSN group, with expertise in advanced communication and wireless sensor networks.
In the future, driverless vehicles will re-invigorate and re-develop the motor and transport industries in all directions. Many individuals and businesses will still want to drive independently, or be unable to adapt quickly, so it is unlikely there will be an overnight switch and an extended hybrid environment will develop over the next 20 to 30 years. Car ownership may even decline over that period as society changes, too.
The legal and insurance framework
All of the above supposes the legal and insurance framework is in place for all stakeholders, not just drivers, to be protected. Stakeholders will include vehicle manufacturers, OEMs, software designers, communications providers, employers, fleet owners, ancillary service providers and statutory bodies with regulatory or public service obligations. This will include highways and transport system engineers in relation to V2I and VII applications.
The ability to obtain motor insurance for driverless vehicles will be a decisive factor in general uptake and market penetration, placing the insurance industry in a central role. Obvious risks for owners and drivers include software reliability and cyber security as well as residual human error. For many years to come ordinary vehicles will be on the roads too, the road network will need to allow vehicles of significantly different levels of sophistication to interact safely.
Accidents will significantly reduce over the long term, but not disappear and new and different levels of risk will have to be considered depending on the vehicle type and environment. Identification of the appropriate standard of care will be challenging, as will establishing a breach of duty. The argument that a driver “lost control” inherently requires (a) a driver and (b) that he/she had a measure of control to start with.
The general view is that the existing regime of strict manufacturer liability will continue to apply. However, this will require the vehicle (or a part of the vehicle which is sufficient to constitute a product in itself), to be proved defective; vehicle manufacturers will strongly resist any such allegations as it is likely to bring about significant reputational damage. If vehicle and software manufacturers do not standardise the in-car controls, standards of driver-training will increase in importance where a person remains in control of the car as the controls are likely to vary across models. Once the individual owner ceases to have any direct input beyond programming a destination, liability for an incident, and the degree of risk associated with the individual vehicle, will rest elsewhere.
The wider effect of driverless (and semi-automated) vehicles
Manufacturers are unlikely to be the only ones with sophisticated vehicle and traffic management systems in play. Assuming that in future decades V2I and VII become core components of the transport network it is important to consider the ways the Highways Act be amended or interpreted to deal with a systems failure outside of the vehicle and where will that chain of liability lead? Public sector insurance policies will undoubtedly change to reflect the impact of driverless vehicles in a myriad of different settings but it’s too early to do more than speculate on those changes.
An obvious area for discussion is emergency services – how will emergency vehicles physically interact with highly or fully automated vehicles and where will liability for any incident lie? Assuming driverless emergency vehicles are unlikely for a long time to come, the duty of care will remain with the driver – probably assisted by V2V communication which will inform driverless vehicles of its approach and allocate priority. Should an incident occur then technology will assist with determining apportionment of liability but the financial cost of a claim may be higher if automated vehicles are expensive to repair or replace, offsetting the prospect of reduced premiums.
There are also implications for public sector bodies where an employee’s role is vehicle based. Employers’ primary obligations are unlikely to change but the ability to assess and manage the risks for employees and members of the public as they interact with an autonomous vehicle will be more challenging. More positively, V2I/VII clearly has the potential to improve safety for junctions, road and lane departures, speed management, commercial vehicle operations and unmanned railway crossings, to name a few.
Driverless cars will also collect and use vast amounts of data, much of it personal to the driver. Google has spent a substantial amount of money testing its driverless car technology but it has also invested over $250 million in ride-sharing business Uber, presumably to obtain data on urban transport patterns. Access to such data will be of interest to insurers assessing risk and responding to accident claims and suspected fraud. Technical data contemporaneous with an accident will help resolve liability by means of expert interpretation and the legal costs currently associated with many contested cases will largely disappear.
Cyber security will be a major consideration for all stakeholders and the consequences of a breach are likely to be very damaging to individuals, businesses and data controllers. Hacking, for whatever purpose, represents an unresolved threat to the safe operation of autonomous vehicles.
There are many changes ahead and the development of driverless cars is being watched closely by many. A necessary full review of the relevant civil and criminal legislation is promised by the summer of 2017; the potential for change in the practical application of our law is considerable, even if fundamental principles such as a driver’s duty of care are unlikely to change. The transformation of the vehicles we see on the road will be gradual but the infrastructure to accommodate such a shift is being built, and rightly so. While the devil will be in the detail, there will a great deal to do to ensure not only effective implementation but also customer awareness of the knock on effects driverless cars will bring.
Nick Rogers, partner and head of motor - BLM
0+
