Matthew Yglesias had a post up on autonomous cars reminding us of the eventual promise of a world where that is the norm. I largely agree with him that as envisioned it does represent a game changer. His focus is how it will change ownership, land use, and public policy taking the change in to consideration.
The problem autonomous cars solves for me is the driving coordination problem represented by traffic. On a percentage basis much of the road infrastructure during most times of day doesn’t face this problem. For some stretches of road at specific times though you have far too much traffic. Normal urban commutes suffer badly from this. Barring a change in living/work patterns that eliminates the problem, I’m up for us solving this problem and I don’t think we need to wait until such time as the autonomous cars are the norm to do it. The problem I think is how to optimize the shortest time to make a trip given whatever flexibility I can afford in terms of leave time, route, and manner of driving. There seem to be three components that tie in to this. 1) the technical piece 2) the incentive piece and the 3) herd behavior piece.
The technical piece looks to be pretty straightforward and off the shelf. Smart phone penetration rate is high and getting higher. Most of what is needed as far as location and user interface is right there. The coordination piece is a bit more interesting. Do you come up with a better solution from a central process that directs all the action or some decentralized rules that allow various nodes to interact with each other to come up with the rolling solution? This part of the problem benefits from similar problems that have been solved.
The incentive piece is a lot less clear. How do you incent drivers such that they will alter their individual behavior? Time savings as long as it actually manifested would at least be a weak incentive. If you could regularly save 10 minutes of commute time by some adjustments in driving behavior there would be some folks that would bite. 10 minutes might represent a pretty steep goal depending. I’m thinking you would need something stronger than that. The kinds of things that pop to mind are free bridge tolls, or a break on gas tax or some other kind of tax reduction incentive. Based on X minutes of driving within parameters or based on say leaving X minutes early/late, you accumulate incentives. You could flip that on its head and try to figure out how to disincent not participating but that might be a tougher nut.
The last piece is the herd behavior. What percentage of participants do you need to achieve where the entire herd benefits? This piece might be less problematic than at first glance. For some herd items like immunity, you need a pretty high percentage for the herd to benefit. Given that roads have limited capacity, relatively few participants are needed in some circumstances to get a result. If you have 4 lanes and a car in all 4 lanes going about the same speed, effectively nothing is going faster than that. This aspect might drive some folks crazy but I wonder how long before everyone would systemically adjust to the new norm. Given that it only needs to be enabled to manage particular volume situations there are plenty of times for folks to drive however they want.
This may all be hopelessly naive and impractical but at the root I guess I dislike the idea of waiting for some event to happen where everything improves when some improvement may be viable right now.