> Maybe automation [...] can cut down on [...] headway.
Indeed. One can think of a BC tunnel as a narrow rail tunnel containing a single highway lane. An express lane restricted to networked self-driving non-truck vehicles. Headway buys fewer vehicles sharing the damage and acceleration of a lead accident vehicle's decel - because with networking, they needn't ever exceed it. One might literally do full-speed bumper-on-bumper.
It's a really pretty point in design space that the Boring Company is targeting. But the commentary I've seen has pervasively been confused. Which may be for the best - industry disrupting startups are often better off remaining disregarded for as long as feasible.
> Indeed. One can think of a BC tunnel as a narrow rail tunnel containing a single highway lane. An express lane restricted to networked self-driving non-truck vehicles. Headway buys fewer vehicles sharing the damage and acceleration of a lead accident vehicle's decel - because with networking, they needn't ever exceed it. One might literally do full-speed bumper-on-bumper.
How do you do stations? If you have the station on the mainline, your entire train of vehicles has to stop at the station just like, well, a train, so what you have is just a less safe version of a train. If you have turnouts for stations, you have the issue of switch fouling preventing you from bunching vehicles too close together, and high-speed turnouts have to be long to deal with turn radii. But the real killer with turnouts is trying to get a train to merge back onto the trunk at speed--with little-to-no space between cars, a slight mismatch in speed is going to equal a high-speed collision.
> How do you do stations? [...] high-speed turnouts have to be long to deal with turn radii [...] to get a train to merge back onto the trunk at speed
It's a highway. Paying with tunneling to avoid surface real-estate constraints. And needing fewer lanes for similar throughput due to density and speed. With minibuses. That are self driving. And don't have to worry about human drivers. And can be street legal.
So how would you do X? Well, how would you do it now, with a highway and minibuses? What if you could restrict it to networked automated minibuses? What if they were tightly integrated with infrastructure?
So for stations? Now: highway ramp to/from a lot, platform curb, or street. Automated: similar, but with tighter non-human margins. Integrated: so many possibilities... including elevators?
Consider a subway station. A short station. For a single minibus. Minibus and mini platform and elevator. Everyone gets off the minibus, crosses the platform, and gets on the elevator. Platforms are expensive. Why have a platform? Ok, so minibus and elevator. Everyone steps off the minibus, directly into the elevator. But the minibus isn't all that heavy, so why craft an extra human space? Ok, so you could have the minibus get on the elevator. At the "top", of ramp or elevator, in a mall, or lobby, or platform, or garage, or street, the minibus could play elevator cab, or PRT, or kiss-and-ride Lyft, or lot rideshare, or neighborhood minibus trundling off down the street. And continuous elevators for tightly-integrated automated vehicles could have surprising throughput. The artist sketch[1] of a point-to-point Chicago-O'Hare station has them playing PRT.
Indeed. One can think of a BC tunnel as a narrow rail tunnel containing a single highway lane. An express lane restricted to networked self-driving non-truck vehicles. Headway buys fewer vehicles sharing the damage and acceleration of a lead accident vehicle's decel - because with networking, they needn't ever exceed it. One might literally do full-speed bumper-on-bumper.
It's a really pretty point in design space that the Boring Company is targeting. But the commentary I've seen has pervasively been confused. Which may be for the best - industry disrupting startups are often better off remaining disregarded for as long as feasible.