Self-driving taxis are coming to London – should we be worried? | Jack Stilgoe

5 hours ago 6

At the end of the 19th century, the world’s major cities had a problem. The streets were flooded with manure, the unintended consequence of dependence on horses as the major form of transport. In this sea of filth, the infant car industry smelled an opportunity. The Horseless Age, a US car magazine, claimed in 1896 that, with the spread of motorcars, “streets will be cleaner, jams and blockades less likely to occur, and accidents less frequent, for the horse is not so manageable as a mechanical vehicle”. The streets did eventually become cleaner, but not safer. Cars brought huge benefits to society, but also huge challenges. By the end of the 20th century, cars and motorbikes were implicated in more than a million deaths a year around the world, as well as contributing to pollution and suburban sprawl.

This story is often told to show that the inevitable march of innovation brings both solutions and problems. However, there was nothing inevitable about US cities becoming dominated by cars. As the historian Peter Norton describes in his book Fighting Traffic, it was a direct result of lobbying by the US car industry. It campaigned for the removal of public transport, the banning of jaywalking and the redesign of streets. The advent of the car in the US is a useful cautionary tale as we consider the introduction of self-driving cars into our lives – especially in the UK.

Waymo, the US self-driving car company, has said that, by the end of 2026, Londoners will be able to ride in their vehicles with nobody behind the wheel. Their sales pitch is that their vehicles won’t get drunk or distracted, improving the safety and efficiency of how we get around. And if the computing and hardware becomes cheaper, a self-driving ride may eventually cost less than a regular taxi.

For the past few years, the company has operated about 1,500 driverless cars across the US. Their safety record so far has been good. Vehicles have not been involved in the sort of life-threatening situations that would turn public opinion against them, but they have got into all sorts of scrapes that are annoying for a city’s day-to-day operation – impeding emergency services, causing traffic jams and, in one case, running over a much-loved San Francisco cat. Policymakers are right to question how much of this accumulated US experience is relevant as the cars launch in London.

London is different from US cities. Much of its road layout predates the arrival of the car. Our roads are more chaotic, with the balance of power tipped a little less towards vehicles. Jaywalking is permitted in London. In 1966, the police tried to crack down on it, but gave up after three months. Our zebra crossings are far from machine-readable; they are designed for micro-conversations between drivers and pedestrians. In some parts of London, such as Exhibition Road in South Kensington, there are experiments in “shared space”, tearing up the old infrastructures that used to force buses, bikes and cars to stay in their lanes. This introduces uncertainty that slows down drivers but makes navigation harder for artificial intelligence.

Another issue is that London has been extremely successful in reducing the number of cars in its centre, while increasing the number of buses and bicycles. If self-driving cars pull people away from active travel and public transport, they hinder rather than help the city’s wider transport strategy.

Big questions remain. How many mistakes, crashes and snarl-ups will people in the UK be willing to tolerate? Should these vehicles clearly indicate when they are driving themselves? Will the technology’s benefits be widespread or concentrated for a lucky few? When they go wrong, will we be able to work out why? Will there be pressure to change the rules of the road to suit the technology?

Having run surveys of public attitudes, my sense is that the public will have high expectations for safety, transparency and equity. We are embarking upon a fascinating social experiment. The minister for local transport, Lilian Greenwood, has said that “government must be in the driving seat”. In the UK, cities will have a say in regulating the technology, which wasn’t the case when it arrived in places such as San Francisco.

If London is to become a laboratory, we need to make sure that the knowledge isn’t hoarded by the companies running the tests. As the transport scholar David Zipper recently pointed out, it is hard to assess the safety of Waymo’s technology when we only have Waymo’s own data to go on. Self-driving car companies have in the past been reluctant to share data with crash investigators and have been cagey about how many people are involved behind the scenes as remote operators.

The first time you are driven by a self-driving vehicle, it briefly feels magical. Like the Wizard of Oz, the companies involved would rather we didn’t look behind the curtain. But if we want to use technology to improve transport, without storing up more problems for the future, we shouldn’t be pushed around by the hype.