3 ms·
> any aspect of either mass transit or personal transportation will be the same after that. That is utter nonsense. The physics of how many cars you can fit on
by discodave 11y ago
> any aspect of either mass transit or personal transportation will be the same after that.
That is utter nonsense. The physics of how many cars you can fit on a road, how fast you can get from A to B and the distribution trip patterns won't change that much.
- CamperBob2 11y agoReally? What makes you think that'll be the case, once the cars are talking to each other? Get rid of every stoplight in your town and see what the "physics" looks like then. The basic capacity of a busy road is not significantly less than that of a train, and we already have thousands of roads. We just don't use them very intelligently. Or, to put another way: a fleet of networked autonomous vehicles is a train. Put money into making that happen, and I'll vote for it. I'm definitely not voting for $50 billion choo-choo trains, or whatever 1890s-era solution the urban transit planners in Seattle are dreaming up this week.
- chipsy 11y agoSimple geometry argues that you can fit more people on dedicated train cars than a fleet of independent vehicles. You don't get 100% of the benefit of stop-free transit in the real world because there will be pedestrians - you need a dedicated right of way. The argument for independent vehicles was made with rail PRT too, and it only ever reached a niche campus-and-airport market. What you get from autonomous vehicles is improved flexibility, and that improves all transit modes while only slightly changing the dynamics that make rail vs. bus vs. car interesting.