It should be relatively easy to _program_ AVs not to take shortcuts through side roads.
Regulating AVs in this way would be hard. It's a hard problem to formulate in words, as a regulation needs, and there would be lots of edge cases. And I'm not sure it would hold up in court. And there would be opposition.
That's probably true, but assuming there's a small number of AV companies that want to maintain good relations with local authorities, regulating isn't necessary, just the threat of regulation should be enough. Especially the risk that if it's regulated it might be regulated badly, gives a good incentive to them to avoid it being regulated in the first place.
All true. On the other side, though, there will be extensive requests like this, and robotaxi firms won't want to concede even one, given the possibility they will then be overwhelmed by them; and not using side streets will open them to accusations of jamming up arteries; and human drivers will start using side streets in response to jammed-up arteries, and you can't regulate THEM from doing it, which will set up a fairness defense for the robotaxi firms to use...
I can see arguments on both sides. It's hard to predict which will win, though in my judgment the congestion-everywhere argument seems to have more force.
One countervailing effect should be that expanding bus service (and making that service safe, pleasant and reliable) should be immensely easier and cheaper than it is now. In fact this should in part already be true since I can’t see any technical limitation to Waymo licensing their driver for buses, and the constrained routes are ideal for AVs.
It should be relatively easy to regulate AVs not to take shortcuts through side roads in a way that you couldn't regulate private cars.
I'm less sure than you are.
It should be relatively easy to _program_ AVs not to take shortcuts through side roads.
Regulating AVs in this way would be hard. It's a hard problem to formulate in words, as a regulation needs, and there would be lots of edge cases. And I'm not sure it would hold up in court. And there would be opposition.
That's probably true, but assuming there's a small number of AV companies that want to maintain good relations with local authorities, regulating isn't necessary, just the threat of regulation should be enough. Especially the risk that if it's regulated it might be regulated badly, gives a good incentive to them to avoid it being regulated in the first place.
All true. On the other side, though, there will be extensive requests like this, and robotaxi firms won't want to concede even one, given the possibility they will then be overwhelmed by them; and not using side streets will open them to accusations of jamming up arteries; and human drivers will start using side streets in response to jammed-up arteries, and you can't regulate THEM from doing it, which will set up a fairness defense for the robotaxi firms to use...
I can see arguments on both sides. It's hard to predict which will win, though in my judgment the congestion-everywhere argument seems to have more force.
If I was going to bet, that's the one I'd bet on.
One countervailing effect should be that expanding bus service (and making that service safe, pleasant and reliable) should be immensely easier and cheaper than it is now. In fact this should in part already be true since I can’t see any technical limitation to Waymo licensing their driver for buses, and the constrained routes are ideal for AVs.
AV technology certainly has a great deal to offer buses, and to transit systems generally; I wrote about this last year: https://www.changinglanesnewsletter.com/p/progress-and-public-transit-part-436
THE END OF DRIVING has a fair bit to say about this subject as well!
Thanks for pointing to the article which I’m not sure I’ve read, book is on my list!