In Britain in 1865, the legislative response to the increasing introduction of self-propelled vehicles on our roads was the Locomotive Act (sometimes known as the Red Flag Act). Amongst a number of provisions, it stipulated that self-propelled vehicles needed to be proceeded with a man walking 60 yards ahead carrying a red flag to warn other road users of the vehicles approach.
Wind forward a century and a half, and little red flags have been waving in my mind with the increasing coverage of efforts from Google, Volvo and others to develop self-driving cars. I have no doubt that the technology is feasible. I do, though, have serious misgivings about how as a society we will react to such technology.
First of all, the concept of a self-driving car poses really interesting challenges for our existing motoring laws. If a self-driving car crashes, who is liable? The owner? The manufacturer? The software developer? And how does insurance provision evolve in such a world? If self-driving cars are self-insuring then much of the consumer insurance industry goes up in a puff of smoke. I’d like to see the claim that they collectively put in for that one…
There are then a stack of Red Flag-type legal challenges ahead where the new technology will probably be subjected to greater scrutiny because it’s new which will be regarded by the providers as both draconian and unfair. That legislation will limit the usefulness of the new technology – and you only need to look at the viable top speed of modern motor cars set against the legal speed limits on open roads to see how long lasting that kind of thing can be.
The final big challenge that I see is what I would see as the “Ikea design challenge” – how do you design self-driving cars to be useful in a worlds where there are mostly human drivers, a mix of human and machine drivers, and then eventually a majority of machine drivers (assuming that the Clarksons of this world won’t let their cold, dead grip leave the steering wheel)?
This challenge – which is crucial to ensuring that self-driving cars don’t become the 3D TV of motoring – is being seriously underplayed if not ignored by the tech companies at the moment. Benefits of self-driving cars (reduced accidents, and especially increased road use density through shorter distances between vehicles travelling at speed) seem to be ones that will only really come to life at the point when machine drivers are in the majority. Given that the average lifetime of a motor car in the UK seems to be over a decade that point is a long way off, and will be even further if self-driving cars are limited in ways that make the costs outweigh the benefits.
It’s also questionable whether increasing capacity of road traffic does anything other that just generate more demand – any user of the M25 can testify to that.
The usual tech industry response to such societal challenges is to ignore them, release the technology, and then leave it to the rest of us to sort out the muddle. But the auto industry is so heavily regulated (particularly on issues of safety) that that “suck it and see” approach just won’t be viable. Short term, mixed mode benefits must be the ones that come to the fore for this to work. Self-parking, lane sensing, aides to the driver are probably the right approach, but may well be seen as technically inferior by the likes of Google.
Maybe the only route for the auto industry to take is to up the complexity of the modern car to such an extent that they become nearly impossible for a mere mortal human to drive. Oh – I see what you did there…