Now, I'm not taking up an ideological position on this, and it may indeed prove to be the case that driverless cars end up reducing pollution by large amounts. But I feel that it is also plausible that we are going to end up being driven by a lot hype into a rather ungreen future or, more mundanely, into accepting a piece of 'modernisation' that makes little difference to pollution outcomes.
Now of course electric vehicles are to me something that represents a great gain, especially as electricity systems move more and more to be based on low carbon energy sources given the increase in renewable energy use. Of course also we need also to plan our environment so that other modes, especially walking, cycling and also buses and trains are given greater priority. We know (or at least I am sure) that these things will reduce pollution and improve quality of life.
But driverless cars and the sort of systems that they will involve are very unknown quantities. In many ways by comparison substituting electric for petroleum based vehicles is a fairly modest change in systems, give or take some changes in fuelling structure (which could have great benefits for balancing renewable electricity with demand). But driverless cars represent a completely new system. This brings me to four potential problems.
First, we do not know how how driverless cars will alter demand for road travel. What is there about a driverless system that would encourage people to travel less by car? Not much as far as I can see. Indeed, 'packages' sold to consumers might offer lower prices for using particular company offerings of ride contracts for driverless cars if they sign up to travel at least x000 miles a year which might actually encourage people to travel more. Alternatively why won't people simply buy their own driverless cars and carry on travelling as usual? After all a lot of people gain their identities form their cars! It will be necessary to offer them an incentive, that is to make things cheaper, and this may mean they will travel more.
Not having a driver of course cuts costs to (driverless) car/taxi companies, but that still doesn't eliminate the costs of buying, servicing and maintaining the cars in the first place. Getting a driverless uber is still going to be very pricey compared to the fuel cost (or marginal cost) of taking that trip in your own car.
Now I'm guessing here. I could be wrong, and miss out something important. I don't know. But what is the point is that the people who are being (no doubt with good intentions) optimistic about the green-ness of driverless cars do not know much more than I do about how this entirely new system is going to work out and interact with consumer demand for travelling by motorised vehicle, driverless or otherwise.
A second factor is that we should be very wary of the modernisation-bandwagon effect. We are witnessing a process whereby this paradigm shift to driverless cars is being dressed up as an inevitable part of modernisation, and its benefits seem to be in a process of being elided with electrification, which, as I have said, is not the same thing necessarily at all. The danger here is that planning systems are given over to this new, 'inevitable march of modernisation' and other important considerations cast aside. This has happened before with urban planning, with everything from road design to high rise flats, to bad effect.
A third factor is that claims made about driverless cars is that they will be used more efficiently than individually owned electric cars. These gains may actually prove to be pretty marginal gains or even non-existent. A problem with new technological systems is that before the practical engineering and socio-technical ramifications become clear the technologies are presented in a utopian fashion (remind you of anything?). Even now we can see some slightly heroic assumptions being made about how cars are going to be made use of all of the time. This on its own may have the perverse consequence of incentivising the companies who own them to encourage people to travel more than they do now. I hear that driverless cars will be much more efficient at braking and using gears etc. Maybe, but I suspect that conventional electric driven cars are being and will be adapted to incorporate at least some of these gains.
A fourth factor is. well, the unknown unknowns. Former US Defense Secretary Donald Rumsfeld famously made the point that there are three types of risk: the risks which can be calculated; the risks that cannot be calculated, and the risks which we don't know about at all (unknown unknowns). In fact he was popularising a economist called Frank Knight who wrote about this in the 1920s as he was discussing the difference between risk and uncertainty. We know about the risks of electric cars and where the uncertainties lie. Or at least we have a much better idea than with driverless cars. But we don't know the unknowns that will almost certainly jump out to bite us in the case of driverless cars. Whether these are minor irritating gremlins, or big monster ones, we don't yet know.
But the point is that there are all sorts of unknowns which really seriously undermine the now widely (and unreasonably) accepted claim that driverless cars necessarily represent a major green leap forward.