I started writing about driverless cars nearly ten years ago. I had read an article in the Evening Standard which suggested they would soon be on the streets of London in trials and would eventually make a massive difference to road safety.
I was initially taken in and thought this was an astonishing development. But then I began to dig a bit into what was needed to enable the mass use of what we now call more accurately autonomous vehicles and found there was far more hype and hope than anything tangible. The picture presented at the time by the supporters and funders of the concept was that this was a completely life changing experience. I remember being at a conference when the enthusiast for the idea – of course a paid consultant – said earnestly that everyone in the room would benefit from this development. He said that within a few years a majority of cars on the road would be driverless and that the quarter of a million annual deaths in road accidents across the world would be massively reduced.
Suburban roads would be greened over because no one would need to own their cars but rather would make use of a massive pool of vehicles. It was therefore going to be a triple revolution. Cars would be electric, autonomous and shared use. All this was patent nonsense as I wrote in a book called Driverless Cars: on a road to nowhere, which has been through a couple of editions.
The problems with this vision, which incidentally has now largely been abandoned even by the most ardent protagonists, are so numerous that it is incredible anyone took it seriously. First, there are a wealth of technical difficulties. Even with AI, ensuring that autonomous vehicles can drive in all circumstances and weather conditions at all times has proved a momentous challenge. Indeed, in their more honest moments, even the bosses of the tech firms admit that it is far more difficult than they imagined and, indeed, may well be insuperable. The idea that driverless cars will be the dominant form of traffic on our roads within even the lifetime of my grandchildren has proved to be a myth. The notion that we will have vehicles that can take our kids autonomously to school and then drive the parents to work and then the grandparents to the hospital has been debunked.
Secondly, there is much resistance to the idea that we will want to share cars. This has been put forward by the supporters of the concept in response to concerns that having lots of driverless cars on the road will lead to increases in the amount of traffic. ‘Ah, they say, but as your car sits in your front driver 90 per cent of the time, it will be better to just share vehicles with other users.’ This has been shown to be totally unrealistic, and is no longer part of their vision. Uber, however, while it has abandoned its own research on driverless cars, is still hoping that autonomous vehicles will enable the company to become profitable. But the company has not explained how owning and maintaing all the cars – since there will be no drivers – will be a cheaper option than having an owner driver who probably earns around £30k per year.
But, thirdly, the biggest hole in the ‘autonomous vehicles for all’ plan is that driverless cars will never be able to cope with certain situations – when two cars meet on a single track road, who will determine which will reverse? – or be entirely secure for its occupants – standing in front of a driverless car will force it to stop so women alone in a vehicle, for example, will be at the mercy of any bad person intent on attacking them. Certainly no VIP will ever use a driverless vehicle.
And so on. None of these points have been properly addressed by the supporters of the idea. Nor have they ever answered the question as to why these vehicles should replace conventional cars. Making them safer than humans – who are pretty good at driving – is a big challenge and possibly insuperable. Moreover, some people – even me – like driving at times. Perhaps not in traffic laden London but on the open roads of rural Italy.
Yet somehow all these lessons learned from twenty years of attempts to introduce autonomous vehilces has passed the government by. There have been numerous failed trials – there was supposed to be tests on whether a convoy of lorries could be led by a driver in the first vehicle and an attempt to run buses over the Forth Bridge with autonomous buses was abandoned because no one wanted to go on them. Yet, while the most extreme versions of this hype have been abandoned, the government does not seem to have learnt any lessons from the fact that all this hype has led to billions of dollars and some twenty years being spent on enabling a few thousand robotaxis to operate in a handful of American cities – and Elon Musk can’t even manage that.
So we now have the road minister, Lillian Greenwood, putting out press releases last week (July 21) on how autonomous vehicle trials are to be speeded up because by the middle of the next decade autonomous cars will generate 38,000 new jobs and ‘unlock’ an industry worth £42bn. None of this is properly explained, but has become an oft repeated mantra. The obvious point that if there really were a driverless car revolution, surely thousands or even hundreds of thousands of drivers would lose their jobs seems to have passed Ms Greenwood and her colleagues by. The government is therefore in danger of speeding up the introduction of a technology that has no clear purpose and whose implications it does not understand. They have been warned.
As a special offer anyone signing up in the next week to my premium substack can have a free copy of my book Driverless Cars: on a road to nowhere. Just send in your address and I will post it out. To sign up to premium, go to the orange button top right corner “Upgrade to Paid “ of this post and choose a premium level (monthly or annual). You dont have to do anytrhing else, I’ll be notified by Substack of all prmium subscribers.