I was recently quite critical of self-driving cars and today discovered a new angle on why I’m not terribly excited about them. A recent study showed that automated vehicles would nearly triple the capacity of existing roads:
If all the vehicles in the U.S. communicated with each other and used automated braking, existing highways could handle 273 percent more cars, according to a new study. The increase would come from smaller gaps between cars, as self-driving systems are able to react much more quickly than human drivers, so they don’t need as much lead space.
If that increased capacity sounds like a good idea, think again:
Any good research out there on the sprawl-inducing potential of automated vehicles?
— Michael Rhodes (@michaelprhodes) September 12, 2015
If the policy of building a ton of publicly-funded roads incentivized sprawl, increasing the capacity threefold will only encourage it all the more, since traffic expands to fit capacity. This can be mitigated by tolls but those are generally only effective in higher density areas.
The simple fact is that building infrastructure on the assumption of individual automobiles is a terrible idea where everyone actually loses, even if current policies allow the illusion to the contrary. Making roads more efficient will probably only increase consumption, which could in turn prop up the infeasibility of the suburbs well past the point where its infrastructural bankruptcy should become obvious. The pile of reasons why the current obsession with self-driving cars frustrates me just got a little higher.