Will AVs be a disaster?
Autonomous Vehicles embody advanced AI, and are likely to radically reshape the future of traffic. AVs can improve accessibility by providing a reliable mode of transport for those who may have difficulty driving due to age or disability. This would allow these individuals to access facilities and services that might have been challenging to reach before.
EF responded:
“[Autonomous Vehicles (AVs) will not help accessibility for those who do not have access to a car, it will gridlock our network and give more of our precious public space to machines. We in the transport profession have been fighting single occupancy travel for a long time and in Australia 40% of people have a license. To then suggest that AVs can open up car travel to the remaining 60% where they may also be capable of zero occupancy travel boggles the mind at the potential increase in VKT, therefore carbon emissions and therefore air pollution.
We should be advocating for less cars on our roads and giving back our public space to people.”
My previous post was mostly positive rather than normative (though I was optimistic about the future, a rare state for me lately).1
I believe we should have relatively less space for cars and more for people. I think I have said that. But that doesn’t mean we shouldn’t try to automate cars, or keep disabled or people who cannot now drive imprisoned in their homes for lack of effective mobility. Public transport does not serve most people on most trips in most of the world. It can serve more, it will never serve everyone. Nor will bikes. Nor will walking over longer distances. Even if sustainable modes can serve all independent travellers in cities, not everyone is in cities, and not everyone can travel without assistance. Furthermore automated cars facilitate more space for people instead of vehicles, because they are much more capable of staying in their lane, lanes can be narrower.
On energy and CO2 emissions: AVs will in general be EVs, and energy will in general be renewable. That won’t make AVs pollution free (tires, brakes), but it will be better.
Complaining about congestion and gridlock is a bit of nonsense. Obviously roads should be priced, and if they are not, AVs will have no incentive not to drive empty. But because they will have no incentive not to drive empty without pricing, and without pricing there would be gridlock, AVs will be regulated or priced to discourage that kind of behaviour once a critical mass is hit. The Hell Scenario is self-negating, it is only a question of when.
Finally AVs can help more by increasing safety and allowing right-sized vehicles that are either privately owned (as people won’t need to own their own tanks if driving a small vehicle, even a bicycle, is perceived as safe), and in denser areas, appropriately-sized vehicles can be shared and delivered to users on demand.
I wrote a book on this. I don’t need to regurgitate everything I have ever written every time I write something new — a link is sufficient.
I'd try approaching this from the policymaking prospective.
As soon as the technology is there (assume it is possible), in my opinion, we are very likely to end up with the following interests becoming dominant in the decision-making framework shaping:
- car companies willing to sell more cars,
- banks willing to extend more car loans at an unprecedentedly low rates (to private owners or private shared fleet operators),
- construction companies eager to build more roads than ever,
- politicians willing to get rid of the headache of dealing with all the 'political difficulties' associated with development of (and running) any mode of transportation that is not cars, and willing to sell to their constituents something that is shiny and innovative,
- lots of people being genuinely happy with cars becoming more available to them (both in terms of lower costs, and zero barriers presently associated with individual suitability for driving)
All these pressures are not new, are already there, and will be emphasized with cars becoming more affordable ('affordable', car loans involved). Another round of urban destruction and mass automobilisation under these circumstances appears to be a scenario far from being impossible.
There are, of course, more positive futures for humanity to pursue - ones where the AV technology gets its first full deployment across all modes of public transportation (where it should be computationally easier by default) which enables further service expansion and street space redistribution towards sustainable mobility, and, eventually, phasing out cars in cities entirely (or almost entirely).
However, there's no guarantee that would happen - too many policymakers are oftentimes more motivated with big (and quick) money rather than anything else, and I don't see enough lobbyists working on this version of our future.
All tech comes with opportunities. It is not the tech nor the worst case scenario that interests me. It is the governance, regulation and policy that maximises the public interest while balancing the need for vendors to get a return which together delivers the best public value outcome. AVs can solve many problems. And as San Fran has found, if allowed on streets before capable, can cause many problems. With something as incredibly complex as automating a vehicle in a mixed use environment there will be multiple false starts. The Gartner Hype Curve is close to the trough of despair for investors in AV, but every learning and failure is a gain towards something that can be made socially useful. And it is always important to point out that access to a car is necessary for many people to thrive in their daily lives simply because the economy has shaped over a century around presumptive automotive access. No government can afford to close all those gaps with route-based PT, but it could potentially overcome inequity through access pricing and a transport card for those in need with a fleet of competitive AVs to buy service from. The tech is a very long way from ready, even in a dry and well maintained city like San Diego, so we have time to get the public value settings right...