Autonomous vehicles: Governments push technology but issues remain
In mid-August, the UK government pledged a further £100m in backing for the continued development of autonomous vehicles (AVs). It’s now consulting on a ‘safety ambition’ – for self-driving vehicles to be as safe as a competent and careful human driver. This will then feed into road worthiness standards for vehicles and their manufacturers to meet.
The UK is one of many countries creating legislation to smooth the rollout of AVs onto the roads, as governments worldwide embrace the concept. There remain legal and ethical risks, however.
For Jamie Horner, a partner at Keystone Law in the UK, AVs are undoubtedly the future of the car industry. However, they’re not without problems. ‘The plans for AVs are certainly ambitious but as countries such as the US and China have seen, it is not as simple as inputting clever software into a vehicle and will require careful consideration of a range of factors before we see these vehicles on our roads’, he says.
The lack of regulation in this area poses a problem. ‘It is difficult to manufacture something that has no stringent guidelines’, says Horner. Meanwhile, preparing for full autonomy while semi-autonomous and driver-supervised test vehicles are on the roads will present significant legislative challenges as well as raise safety concerns, as driver input, control and responsibility is gradually reduced, he adds.
AVs raise a number of complex legal and ethical issues, agrees Sajai Singh, Chair of the IBA Technology Law Committee and co-chair of the corporate practice at JSA, in Bengaluru. Anything modelled on artificial intelligence (AI) involves gathering user and geospatial data, with the former raising privacy issues, while the latter carries risks associated with national security. ‘Companies/manufacturers would therefore require permissions from the government on the type of technology and data collection activity they would be allowed to use and would need to adhere to data protection [rules]’, says Singh.
"Edge cases’ represent instances when an unlikely and potentially dangerous event occurs […] and go some way to explaining why the safety of AVs is 99.99 per cent rather than 99.9999 per cent
Global Head of Transport, DWF Law
Liability determination is also a key concern. With AVs, the user and the manufacturer are the two major actors on whom liability will rest. Ethical issues could also arise in crash situations in which the AI is faced with the need to make difficult judgements. For example, the AI might need to choose between saving multiple individuals over a single individual – the classic ‘trolley problem’.
To tackle this conundrum and to ensure, as far as possible, that the final decision is made by a human being, legislators in Germany have introduced the concept of a technical supervisor.
‘This is the natural person who can deactivate the motor vehicle during operation and release driving manoeuvres,’ explains Eric Wagner, a partner at Gleiss Lutz in Stuttgart. ‘It is also obliged to evaluate an alternative driving manoeuvre and to release the motor vehicle for this, to initiate any necessary measures for traffic safety and to establish contact with the occupants of the motor vehicle without delay.’
This would not, however, work for what the Society of Automotive Engineers defines as ‘Level 5’ AVs. These fully automated vehicles – which do not require human attention and won’t even have steering wheels – are undergoing testing around the world, but none are yet available to the public. This means, says Wagner, that the German innovation won’t provide a permanent legal-ethical solution to the problem.
The UK government is considering a similar system: that each AV should have an automated driving system entity, which would be liable in the event of wrongdoing. ‘This means the manufacturer who gains authorisation for the car will have to accept some form of direct/vicarious liability,’ says Jonathan Moss, Global Head of Transport at DWF Law in London. ‘Each car would therefore be backed by a separate legal entity that could be subjected to a range of criminal sanctions, which could include infringement notices, enforceable undertakings, or suspension and withdrawal of ADS [automated driving system] approval.’
Insurance products will need to adapt to cover the party that could be liable, says Moss. ‘It seems liability will attach to the product manufacturer; however, how can this work in terms of establishing who was at fault? Insurers also need to consider the scenario where there is some level of control built in; where a passenger could step in and avert an accident and does not do so, who is liable?’ he adds.
Given that most road traffic accidents are currently caused by human error, the hope of AV developers is that they will be ultimately safer and that this, when combined with the increasing amount of data capture available from vehicles, should result in less associated litigation overall, says Brian Wong, a partner at Burges Salmon in the UK.
‘There will certainly also be different forms of litigation in respect of road traffic incidents from AVs – in particular, in respect of enforcement proceedings under the proposed new AV safety framework (or associated aspects such as data protection) and/or civil claims focusing on more product liability aspects rather than human driver negligence’, he says.
The UK government is seeking a wider rollout of AVs in 2025, with some self-driving features rolled out later in 2022. This, says Moss, is very optimistic and is a target rather than a prediction, with the main stumbling block being technology.
‘It is difficult to account for the near-infinite number of variables that an AV might encounter’, he says. Issues such as running red lights and misidentifying tram tracks remain, while adverse driving conditions – such as rain – complicate matters. ‘Most importantly, “edge cases” represent instances when an unlikely and potentially dangerous event occurs’, says Moss. ‘These are difficult to prepare for, and hence go some way to explaining why the safety of AVs is 99.99 per cent rather than 99.9999 per cent.’
The countries at the forefront of the AV revolution – China, Germany, South Korea and the US – are currently operating Level 4 AVs, those that can operate in self-driving mode but allow a human to manually override, as a maximum and then only in limited areas with strict restrictions in force. It must be questioned therefore whether Level 5 AVs can become a reality at all unless infrastructure and legislation move on apace.
Image credit: Mike Mareen/AdobeStock.com