While driverless cars seem likely to be the future of motoring, questions often arise as to their safety and it appears that, while there are those who eagerly await their arrival, there are some who fail to truly comprehend the notion of autonomy, in its current form, which can unfortunately make drivers dangerously complacent. There are also concerns as to whether a machine can make the split second decisions needed when driving. Driving is often testing one’s ability to look at the way in which everyone else is using the road, as well as being in control of one’s own vehicle – how can a robot possibly do this? Will a robot, in fact, be better at making these decisions?
A better understanding
Currently the most advanced vehicle on general sale is the Tesla which is around level two on The Society of Automotive Engineers [“SAE”] Levels of Autonomy Scale. This is a car which has partial automation, meaning the car has key automated capabilities but the driver must still be firmly in control. A level zero is, as you would expect, zero automation. A level one autonomous vehicle contains intelligent features that can add a layer of safety and control but by no means is the car capable of carrying out these actions without a human driver. The car can alert the driver to conditions, the environment and any obstructions but again the driver must be at the wheel at all times. Level three should be clarified as a ‘co-pilot’ rather than ‘autopilot’ as the car will assist a human driver who is expected to manage vehicle operation. Level four is considered high automation which is capable of performing all safety-critical driving functions while monitoring environments and conditions in defined use cases. Some experts considered this to be full autonomy but to others it is felt to provide only partial autonomy as it only operates in certain conditions and on certain roads. Level five is fully autonomous, a car that will not exist, experts believe, until at least 2025.
The problem is that while these levels of autonomy break down the reality of autonomous cars and their functions, they are not necessarily reflective of what is currently being advertised. The Center for Auto Safety and Consumer Watchdog states, "The marketing and advertising practices of Tesla, combined with Elon Musk's public statements, have made it reasonable for Tesla owners to believe, and act on that belief, that a Tesla with Autopilot is an autonomous vehicle capable of 'self-driving'".
This is not the reality. While it may be a possible in the future to have a snooze on the back seat while you travel into work by a driverless car, for now these cars are not designed to be left to their own devices. The driver is meant to be alert and able to take over at any moment, and for good reason.
In 2016, a Tesla owner was killed when his car failed to spot a lorry crossing its path. The US National Transportation Safety Board (NTSB) found that Tesla's Autopilot system was partly to blame. The NTSB found that the driver had his hands on the wheel for just 25 seconds in the 37 minutes of driving. Christopher Hart, a member of the NTSB stated after their report was published, "Tesla allowed the driver to use the system outside of the environment for which it was designed, and the system gave far more leeway to the driver to divert his attention to something other than driving."
In the UK, a Tesla driver was disqualified for 18 months and was required to complete 100 hours of community service after placing his car on autopilot and moving across to the passenger’s seat while travelling 40mph on the motorway.
This is not intended as a poor write up of Tesla, its purpose is to illustrate the lack of understanding when it comes to autonomous vehicles - the consequences of such can be deadly. More training should be provided to consumers when they buy a car fitted with these programs so they understand how these programs work and what their car is and most importantly, is not capable of.
The truth is autonomous cars, or their current ‘co-pilot’ predecessors are capable of great things but the vital safety benefits that automation can produce is depleted unless consumers are provided with comprehensive and coherent training as to how the automated technology works.
As stated above, automation should provide vital safety benefits. Automatic braking systems, often found in level one and level two autonomous vehicles, will help to ensure less road traffic accidents however this technology is reliant on the consumer knowing that the technology is there and understanding how it works. Last year Volkswagen stated that their vehicles fitted with autonomous braking were involved in around 40% less accidents. Data from Swiss accidents concluded that drivers were responsible for 90% of accidents, so by removing the human element, we may see a dramatic impact on the number of crashes that occur on the roads1. Autonomous features are being added all the time – lane change warnings, blind spot alerts, and parking assistance. These systems help move a driverless car away from levels one and two and move to next steps where vehicle monitors the environment and even communicate with other cars.
Richard Barlow, chief executive of Wejo, (multi-marketplace, connected car exchange), recently discussed this aspect of cars sharing information between themselves and how the developing vehicle technologies influence the road traffic accident landscape. He stated, “[Sharing connected car data] offers comprehensive insight into driving behaviours and the technology to respond effectively to the needs of fleets, drivers, insurance companies… will be the necessary precursor to driverless cars… enriching the in-car experience… and enhancing safety standards.” 2
The communication between cars over a dedicated WiFi, sharing information about speed, direction and traffic flow as well as road and weather conditions will help to reduce accidents but the communication goes further and allows, in the event of an incident, the vehicle to transit a real-time notification of a collision, similar to an airbag deploy, along with exact details of the vehicle’s location which enables emergency services to respond to that location much more quickly than in traditional events – potentially life-saving especially if in a remote area or left unconscious as a result of the accident. The next step will be vehicle to infrastructure communication. Here vehicles will become connected to smart highways and traffic lights which will help reduce congestion and commuting times.
Developing technology not only helps reduce accidents but is also important post-RTA. As Shirely Woolham states, “Most of the developing technologies either speed up claims process… or eliminate accidents altogether…on-board cameras and telematics make liability resolution easier than ever before, the latter already being used in criminal proceedings to provide evidence… Technology also provides rapid assessment of the extent of damage and cost of repair.”3
This technology is the foundation for fully automated cars which make driving safer and easier for consumers. While we do not have accidents free roads just yet, the possibility may one day be available due to the technology currently being created. However, the point remains, consumers must be made fully aware how these systems so that they be used effectively. Further, the development of technology also means that MOT tests and testing centres will need updating to reflect the changes. Repairers will need up-skilling and be provided with right equipment in order to fix faults. Moreover, in-car technology, including car to car communication, along with car to infrastructure communication allows governments to track the car and in turn its driver. Updates confirming current location and where the car is going – even predictive analytics similar to a smart phone, allows the government to either know or predict our every move while driving. While we may, potentially, be able to trust in the cars technology to assist us while driving, the fact that this technology may come with the price of constant state surveillance, arguably this removes the notion of freedom associated with driving. Worryingly, car autonomy may render us ‘safe’ but drivers may be left ‘powerless’ as a result.
Unsurprisingly, the burning question generated by autonomous cars is responsibility. Not just the responsibility of manufacturers to ensure that consumers know exactly what they are driving and how it is used but also the responsibility of the ‘drivers’/owners themselves.
Currently if there is an accident, the driver at fault is penalised. In a level one, two or three autonomous car a driver will continue to be at fault – as noted above due to the fact that they must continue to operate the car while being assisted by autonomous functions but what happens when the car is in control? Does the driver cease to be a driver, are they now simply a passenger and who then takes responsibility?
This aspect presents interesting liability debates. How can a driver – or rather passenger – be held accountable for any accident if they could do nothing to prevent it and if they are not responsible then who is? Moreover if they were injured themselves in the event of an accident how do they make a claim for compensation?
Chris Grayling, the Secretary of State for Transport confirmed in November 2017 that; “We are creating a new compulsory insurance framework that covers the use of automated vehicles and when the driver has legitimately handed control to the vehicle. This will ensure that victims have quick and easy access to compensation.” The government proposes to extend compulsory motor vehicle insurance, Part 6 of the Road Traffic Act 1988, in that the owner must also ensure that there is an insurance policy in place that covers the manufacturer and other entity’s product liability.
The Association of Personal Injury Lawyers (APIL) suggests that, Part 6 of the RTA 1988 should be amended to make clear that accidents involving automated cars are covered by compulsory motor insurance. Automated vehicles should already be covered by motor insurance policies without the need for additional product liability insurance. APIL suggest that in circumstances where the accident involves a car with automated technology, the injured party should bring a claim against the motor insurer in the usual way, and be compensated under the normal car insurance policy on a strict liability basis. It will then be up to the insurer to recoup damages from the manufacturer in the relevant circumstances. They suggest that the model should mirror the Employer’s Liability (Defective Equipment) Act 1969 which imposes strict liability on the employer who recovers damages from the manufacturer when an employee is injured at work.
This process is far preferable to a separate product liability policy which would require a claimant to bring their claim under the Consumer Protection Act 1987 against the manufacturer which involves proving the product was defective with a long-stop limitation period of ten years after the product was put in circulation.
The government agrees, for their intention behind the extension of the RTA 1988 is to emphasize that if there is an accident, the compensation route for the individual remains within the motor insurance settlement framework, rather than through a product liability framework against a manufacturer. Gayling confirmed that at current, if there is an accident involving an automated vehicle, “victims might have to take vehicle manufacturers to court, which would be time consuming and expensive [and] undermine the quick and easy access to compensation [currently in place]…’ It also means that insurers can recover costs from the liable party, which in the majority of cases is anticipated to be the manufacturer.
While this all appears to be packaged nicely, as automated cars do not exist as of yet in this way, i.e. level four and five on the SAE scale, it is unknown as to how the technology will fully work and regulating in this way could diminish the benefits that automation may eventually provide. The regulation may not work at all and may need to change entirely within the next 3-5 years – it is something that is difficult to settle until vehicle technology has evolved further.
While we should embrace driverless cars for the advances in technology they provide and the benefits to drivers there remain many issues that must be ironed out. Without the correct training, clear regulations and compliance in place how can we ensure overall safety for consumers and pedestrians alike? State surveillance, autonomous ethics, insurance and the liability of driver - the list of issues is endless. More work needs to be done to ensure autonomous cars are a safe everyday reality.