A new kind of vehicle has taken to the roads, and people aren’t sure what to make of it. Is it safe? Can it cope with other road users? Will it require a radical overhaul of the transport infrastructure? The questions that are being asked today about self-driving cars were raised a century ago when the first motor cars roared onto the roads. So how can the concerns raised during the first road revolution help us think about the second?
Safety is the main concern at the moment, after the first fatality in a self-driving car, which took place in May. Joshua Brown, the owner of a Tesla Model S, was killed when he set the car to “Autopilot” mode and it ran into the side of a lorry which its sensors had failed to spot. Autopilot is still in “beta” (ie, it is an unfinished prototype): drivers are supposed to keep their eyes on the road and take over if anything looks amiss. In this case, that did not happen. Tesla notes that this is the first-known fatality in 130m miles of its cars driving themselves; the American average for road deaths is one every 94m miles, and the global average is one every 60m miles. In other words, Autopilot is already safer, on average, than human drivers. But the company, nonetheless, has been criticised for using its customers as guinea pigs for an unfinished and imperfect technology. Never mind that in America alone, around 90 people die in road-traffic accidents every day: Autopilot, it seems, is expected to be flawless.
Exactly the same debate took place a century ago. The first deaths in car accidents attracted much attention – church bells tolled in Memphis, black flags were flown in Detroit and stone memorials erected in Baltimore. The car’s defenders hit back: an Italian car magazine argued in 1912: “Horses, trams, trains can collide, smash, kill half the world, and nobody cares. But if an automobile leaves a scratch on an urchin who dances in front of it, or on a drunken carter who is driving without a light,” then people blamed the scourge of motor cars. Yet people gradually came to accept road deaths. Self-driving cars are being held to an impossible safety standard today, but eventually the benefits of safer – if not perfectly infallible – vehicles will become apparent.
There are similar parallels in the infrastructure conundrum. Cars, moving much faster than carts, whisked up clouds of dust on unpaved roads. The obvious solution was to pave the roads, but the idea of paving all roads seemed absurd. Brian Ladd, the author of “Autophobia”, a history of opposition to cars, writes that “it was hard to imagine that such expense could ever be justified, and discussion turned to speed limits and even to outright bans on automobiles.” But as the merits of cars were embraced, the cost of paving roads seemed a reasonable price to pay and the entire road network was redesigned around the needs of cars. Today the cost of replacing traffic lights and roundabouts with management systems that allow cars to glide through junctions without stopping seems prohibitive – but not when you consider the amount we have spent on infrastructure for cars over the past century.
Just as early “horseless carriages” looked similar to carriages, but without the horses, early self-driving cars look like ordinary cars. They will probably change shape altogether when they no longer need pedals and steering wheels. Will they be pods? Shared minivans? Will they be owned by individuals, or summoned from a pool when needed using an app, or a mixture of the two? Where will they be deployed first – and will all this take five years, or 50, to materialise? The idea that our lives will be changed by what is still an imperfect, speculative technology, seems unlikely. But so did the idea of mass car ownership in the 1890s.
As we grapple with the implications of self-driving vehicles, we can learn a lot from the era when cars first took to the roads a century ago. Zooming down the road ahead, we should take an occasional glance in the rear-view mirror.