Why it might be time to apply the brakes on the dream of driverless
July 26, 2018
Simon Walker, Chief Executive Officer, talks to City A.M. about the autonomous revolution.
The automotive industry is undergoing seismic change, fuelled by multi-billion-pound investments in electric, autonomous and robotic technology from manufacturers and governments alike. But while the commercial attractiveness of being first to cross the finish line in the race for driverless remains, it is important to recognise the challenges that come with some of the technologies that sit at the heart of making driverless a reality.
Mechanical glitches that may once have been restricted to brake pads or suspension levels have transitioned to self-parking sensors, connectivity, rear view cameras and electronic handbrakes, to name a few. The introduction of these features has been designed to make driving an easier, safer, and more enjoyable experience, but recent testing has revealed a vulnerability to hacking, with the potential to put the lives of drivers at risk.
Software coding risks in particular were recently highlighted by a Chinese cyber-security lab, which identified 14 separate flaws in BMW’s car computer systems. Moreover, research from Tencent’s Keen Security Lab exposed a range of ways that hackers could compromise the safety of cars, including using infected USB sticks to connect to Bluetooth, thus exploiting the vehicles’ 3G/4G data links.
Whilst 14 faults may sound problematic, it could be the tip of the iceberg of an issue that is much greater. A further unrelated study found that it takes around 200 days for data breaches to be detected, which means there may be additional undetected vulnerabilities present in BMWs and in other connected cars, and that BMW could find itself the face of a far wider problem.
Despite these safety concerns, innovation in automotive technology sees no sign of abating, which begs the question of how the insurance sector will need to evolve to respond accordingly. At what point should risk be transferred from the driver to the manufacturer? In the event of an incident, does blame lie with the technology company behind a particular software? The chain of establishing liability becomes much longer and more complex.
Until we have more clarity on driverless vehicles, we shouldn’t assume that dangers won’t exist. This was most recently demonstrated by the fatality involving the self-driving Uber in Arizona earlier this year. It was therefore surprising to see Carlos Ghosn, chairman of the Renault, Nissan, Mitsubishi alliance, last month claim that ‘all the reasons you have today for a car accident you are going to make disappear’ with the introduction of autonomous vehicles.
This statement of course focuses on the potential upsides of reducing the risk of human error, whilst ignoring the fact that the risk profiles of such vehicles is markedly different, thus opening the possibility of new accident triggers. Driverless cars will undoubtedly reduce the likelihood of human error, but at such an early stage of development we can’t yet rule out risks associated with the technology itself.
Furthermore, driverless vehicles will drive on the same roading infrastructure as current vehicles, and that infrastructure is full of potholes, inconsistent or degraded signage, and of course other vehicles. A fleet of driverless vehicles will not absolve the government of the need to provide safe and well maintained roads.
Investment in automotive innovation should continue to be encouraged, but not at the detriment of safety. Ensuring we balance what is technologically possible with what is socially acceptable means understanding the full extent of the risks associated with driverless technology before customers are handed the keys.
There remains a vast range of differing views on how quickly fully autonomous cars will become mainstream on public roads in the UK; particularly prevalent given the emphasis of autonomous cars within the government’s industrial strategy.
But a disproportionate focus on the future should not take precedence over the importance of striking a balance between the safety and wellbeing of customers.
Regulating a complex new technology, and technologies that go hand-in-hand with that, is difficult, particularly when it is evolving rapidly. Deep and detailed guidance on the considerations and legalities of autonomous vehicles for the vast number of organisations involved in the driverless revolution is vital, as well as ensuring that autonomous vehicles are safe, without inhibiting innovation.