Are self-driving cars really all that?

The self-driving car holds much promise. But it might not be taking you to any destination just yet.

The self-driving car has been invented, but is it likely to be widely adopted by as a mode of transport? Will we really buy self-driving cars? Or will we hire them like taxis, or even hop on and off them (or in and out of them) like buses? What would have to change in the way we live our everyday lives in order for us to adopt this strange interloper? And what sort of other things and services might we need to support travelling around in driverless cars?

Some of these questions go so far into the future that we can't possibly know the answers. So let's start with what we do know, and that is, perhaps surprisingly, the technology behind a self-driving car.

The concept of the self-driving (or, more alarmingly put, the driverless car) relies on the clever assembly of many existing technologies. There are several versions of the driverless car in existence, so let's focus on one of them. Google's car is perhaps the most widely covered. Developed through their long collaboration with Stanford University, it cleverly combines a raft of technologies that most of us, in one way and another, are fairly familiar with.

The car uses data gathered from Google Street View with artificial intelligence software, inputs from video cameras installed inside the car, a light detecting and ranging sensor (LIDAR) on the roof, a radar sensor on the front and positioning sensors on the rear wheels. Interestingly, the self-driving car looks suspiciously similar to any other car (only with a few extra gadgets). This is likely to change as the physical constraints presented by these technologies are overcome by the skills of designers that reconfigure interior car spaces as meeting rooms, cafes or perhaps even playrooms.

Not only does the current Google car look very similar to any other Toyota Prius, but the idea is that it actually behaves like one too. Making use of our existing infrastructure, roads, parking facilities, fuel and service stations, the car is designed to replicate the capabilities of human drivers - without human input.

This could be all for the good, and bodes well for its early adoption. Research shows products that fit in to our everyday ways of doing things are much more likely to be adopted. We need little convincing of this when we consider how Apple introduced a tiny computer into our pockets where a mobile phone used to be.

We took to the iPhone like ducks to water because it fitted in with how we were already trying to live (emails on the move, taking and sharing photographs, synchronised and shared diaries and so on). Will the self-driving car be such a welcome fit? As a society we remain much in love with the car, despite persistent efforts to persuade us to make more frequent use of public transport. With a car we can simply jump in and go wherever we want, whenever we want, with no prior arrangement or planning. This flexibility and convenience is what makes us love our cars and if the driverless car can offer this then perhaps it's in with a shout.

The logic behind the driverless car is also most appealing. By relying on technologies we can tame or even eliminate the human error that lies behind so many horrific road traffic accidents. But change begets change.

If we remove drivers, what else might we surreptitiously lose, gain or need, to make our new automobile system work as well as or better than our existing one? We might lose the need for driving tests. So we may lose the structure which provides both education about rules of the road and the framework for developing the capabilities and knowledge that underpin licensing and traffic regulation requirements.

And what is missing? If we break down now, we call the breakdown services and they come and fix our car by the roadside. When a driverless car malfunctions and takes us somewhere we don't want to go, or stops on the highway and won't move, what kind of breakdown service will we need to rescue us? With the increased sophistication of the technologies of driverless vehicles we will need to access different kinds of services to keep us all motoring.

Similarly, as the sophistication of the technologies increases, will the price increase too? This matters because it affects how the market unfolds. If prices turn out to be prohibitive for mass market consumption, then it is likely to be only the wealthy, businesses (maybe taxi firms, or other public transport providers), or even governments, that first adopt these technologies.

Even with government support and intervention there is no guarantee of success. Take the French government’s ill-fated attempt to introduce Aramis, a driverless light rail car, to Paris in the 1980s. Despite being championed by the French defence company Matra and supported by the government, and despite the successful development of prototypes (boasting some of the most advanced and reliable technologies of its time) the project failed to take off. The experience is described in a wonderful book Aramis or The love of Technology by Bruno Latour.

Could the driverless car meet the same fate? We do not yet know. What is more certain is that there is much work that needs to be done to make the market for the driverless car a reality. There are policy implications associated with developing the right infrastructure, creating a new automotive support system and the right markets for self-drive systems. Manufacturers will need to understand the services and capabilities they need to supply with these new vehicles. Businesses will have to develop new models that connect them with other businesses to form networks of support and they will have to work to imagine, make and shape markets for these new technologies as they unfold. The self-driving car holds much promise. But it might not be taking you to any destination just yet.

Dr Katy Mason is Reader in Marketing & Management at Lancaster University. She is co-author of the paper "Self driving cars: A case study in making new markets", which is part of a series of reports on market making for the Big Innovation Centre.

The Google car. Photograph: Getty Images

Dr Mason is Reader in Marketing & Management at Lancaster University.

Getty Images.
Show Hide image

The problems with ending encryption to fight terrorism

Forcing tech firms to create a "backdoor" to access messages would be a gift to cyber-hackers.

The UK has endured its worst terrorist atrocity since 7 July 2005 and the threat level has been raised to "critical" for the first time in a decade. Though election campaigning has been suspended, the debate over potential new powers has already begun.

Today's Sun reports that the Conservatives will seek to force technology companies to hand over encrypted messages to the police and security services. The new Technical Capability Notices were proposed by Amber Rudd following the Westminster terrorist attack and a month-long consultation closed last week. A Tory minister told the Sun: "We will do this as soon as we can after the election, as long as we get back in. The level of threat clearly proves there is no more time to waste now. The social media companies have been laughing in our faces for too long."

Put that way, the plan sounds reasonable (orders would be approved by the home secretary and a senior judge). But there are irrefutable problems. Encryption means tech firms such as WhatsApp and Apple can't simply "hand over" suspect messages - they can't access them at all. The technology is designed precisely so that conversations are genuinely private (unless a suspect's device is obtained or hacked into). Were companies to create an encryption "backdoor", as the government proposes, they would also create new opportunities for criminals and cyberhackers (as in the case of the recent NHS attack).

Ian Levy, the technical director of the National Cyber Security, told the New Statesman's Will Dunn earlier this year: "Nobody in this organisation or our parent organisation will ever ask for a 'back door' in a large-scale encryption system, because it's dumb."

But there is a more profound problem: once created, a technology cannot be uninvented. Should large tech firms end encryption, terrorists will merely turn to other, lesser-known platforms. The only means of barring UK citizens from using the service would be a Chinese-style "great firewall", cutting Britain off from the rest of the internet. In 2015, before entering the cabinet, Brexit Secretary David Davis warned of ending encryption: "Such a move would have had devastating consequences for all financial transactions and online commerce, not to mention the security of all personal data. Its consequences for the City do not bear thinking about."

Labour's manifesto pledged to "provide our security agencies with the resources and the powers they need to protect our country and keep us all safe." But added: "We will also ensure that such powers do not weaken our individual rights or civil liberties". The Liberal Democrats have vowed to "oppose Conservative attempts to undermine encryption."

But with a large Conservative majority inevitable, according to polls, ministers will be confident of winning parliamentary support for the plan. Only a rebellion led by Davis-esque liberals is likely to stop them.

George Eaton is political editor of the New Statesman.

0800 7318496