The 1830 opening of the Liverpool to Manchester railway line was supposed to be a pure celebration of innovation and progress. In the event, it was overshadowed by the death of William Huskisson. An MP for Liverpool, he was hit by George Stephenson’s Rocket about half way up the line — the first in the world built uniquely for steam-powered trains. The accident served to galvanise public interest in the advent of fast, long-distance public transport. It also, of course, drew attention to its dangers.
The tragic collision between an autonomous Uber taxi and a pedestrian in Phoenix, Arizona on Sunday may have a similar effect. It has served as a reminder of how close we are to a world of pilotless mobility. At the same time, the fatal accident — the second in two years of trial runs of autonomous and semi-autonomous vehicles in several US states — has underscored the risks inherent in many new technologies, just as Mr Huskisson’s death did.
Uber, whose brash founder Travis Kalanick has given the company a certain notoriety, has on this occasion been duly sensitive. It has conveyed sympathy to the victim’s family and suspended all testing of its driverless vehicle fleet while investigations into the cause of the accident are pursued.
Two immediate questions need answering. Why did the vehicle and its software fail to spot and avoid the pedestrian? When it did fail, why did the so-called “safety-driver” behind the wheel, not intervene in time to prevent the tragedy?
An accident like this was going to happen. The tech companies, carmakers and start-ups investing billions in developing automated computer systems to replace human drivers may have hoped their technology would be further along the line before it did. At an awkwardly early stage it raises fundamental issues about when it is safe to allow robots of this kind into the public sphere.
Last year about 1.3m people died in road accidents across the world. Extrapolating back to the time when cars were invented and relevant data did not exist, some 50m humans may have lost their lives in accidents. Hypothetically, cars operated by computers should have a far safer record. While they may never be 100 per cent safe, at least they will not get drunk, distracted, nor doze off. But, it is still far from clear how they will interact on the road with humans who do. Nor is there yet sufficient knowledge of the safety implications of humans interacting with automated systems. The Phoenix accident has underscored that too.
In spite of Sunday’s tragedy, safety may yet emerge as the most compelling justification for this new technology. It is not though the main reason why at least 18 companies are investing billions in developing it. As Mr Kalanick has pointed out in the past it is all about cutting costs, and increasing productivity, by getting rid of the “dude in the car”. The day that is widespread, countless millions of drivers will be losing their jobs around the world — a factor that must also be taken into consideration when weighing up the pros and cons of this new technology for society.
Before even getting to that, however, the regulators need to re-examine the ethics of allowing technology still in development to be tested in the public sphere. Some real-world trials are necessary in order to improve the safety of these vehicles. But if they are to win public acceptance, the proponents of driverless cars need to be able to convey a greater understanding of how the technology works and why it could go wrong. If doing so delays the day when human hands are no longer needed at the wheel, so be it.
Bagikan Berita Ini
0 Response to "Driverless cars and the imperative of safety"
Post a Comment