> No. Trains must make every station. A low occupancy pod will only diverge from the high speed lane when it has reached a destination or pickup point.
So you'll have to build twin tunnels through stations (and since they'll require an ADA-compliant evacuation walkway, they'll have to be almost full-size subway tunnels even if you're using smaller personal pods), which will be a fiddle to use TBMs for. You'll have to have a system for diverting or not, and that's going to require maintenance and break down sometimes. And what does the actual mechanics of pods stopping for people to get out look like at busy stations? You'd need giant shunting yards to route pods to free "platforms" (like a garage forecourt only much bigger), and you'd have to build them underground in the places that are most expensive to do tunnelling and remove spoil from, and they'd have a complex layout that would have to be built in a labour-intensive way - all the lighting and ventilation and fire-suppression systems would have to be fitted into that layout (underground stations are already the most expensive parts of building subway projects). Or you'd have pods backing up onto the main line. Or you'd have to batch them together in a, uh, train.
> I don't think anyone disputes that. But this happens because there is no comparable point to point transit system and the existing road network is inefficient.
Tunnels are still only going to go from some places to some other places. Roads are already planned carefully. But there's only so much capacity you can get out of individual vehicles.
> The point of building tunnels is remove those conflict areas and provide direct point to point routes with very high throughput.
It's not going to be high throughput though. Every transportation system has a high throughput when you're just looking at maximum capacity along a plain line going at speed - it's all the fiddly interfaces where the throughput goes down. It'll be the same for a tunnel system (if it's ever built).
Let's not speculate on how evacuation can be handled in a regulation-compliant way. I can certainly imagine a system where trafic is reliably halted when pedestrians are detected on the main lane. You might counter that such a system is unreliable, and until a regulator looks over such a proposal, it's impossible to predict their decision. Let's just agree that there is no fundamental physical reason that prevents humans traveling in vehicles down 3m wide shafts - see for example elevators.
I don't really understand what you mean by maintenance of the system used for diverting from the main lane - it's essentially the same thing as an exit ramp from a highway, but much shorter since the reliability of the automated driver is a built-in design factor.
The lane shift maneuver requires about 1s, so about 15m of widened tunnel at the entrance and exit of the stations. An additional 20m of is required for reliable deceleration and acceleration respectively, which would be handled by dedicated tunnels serving the station. I agree that these interfaces would handle much less vehicles, but it would drive down the throughput of the respective station, not that of the main lane.
Regarding the size of the underground station - why not make it on the surface? You can have much steeper ramps in an automated, controlled system, if for example de-icing is guaranteed near the surface with dedicated heaters, the quality and grip of the tires is controlled etc. Essentially, any down-town parking can be transformed in a very large station by dedicating o few hundred m^2 for these ramps.
> Let's not speculate on how evacuation can be handled in a regulation-compliant way.
If we're going to make a valid comparison between different systems then we have to compare them under the same regulatory regime. A lot of the costs of real-world subway systems are down to regulatory requirements, so comparing an imaginary system that's exempt from those regulatory requirements is going to be very biased.
> I don't really understand what you mean by maintenance of the system used for diverting from the main lane - it's essentially the same thing as an exit ramp from a highway, but much shorter since the reliability of the automated driver is a built-in design factor.
Highways try to avoid having exit ramps in tunnels, because fundamentally you're creating the possibility of a vehicle crashing into the pointy bit (or at best a blank wall). If you're relying purely on automated steering to avoid that, inevitably sooner or later a pod is going to hit it, if only through mechanical failure of the steering system.
> Regarding the size of the underground station - why not make it on the surface? You can have much steeper ramps in an automated, controlled system, if for example de-icing is guaranteed near the surface with dedicated heaters, the quality and grip of the tires is controlled etc. Essentially, any down-town parking can be transformed in a very large station by dedicating o few hundred m^2 for these ramps.
That doesn't solve the problem - the reason tunnelling is expensive in these areas is because land is expensive there. Parking has an absurdly small capacity compared to transit. Even bus stations just can't unload enough people quickly enough in the space available in truly dense city centres, and this pod system is going to need a lot more space per person than a bus station.
So you'll have to build twin tunnels through stations (and since they'll require an ADA-compliant evacuation walkway, they'll have to be almost full-size subway tunnels even if you're using smaller personal pods), which will be a fiddle to use TBMs for. You'll have to have a system for diverting or not, and that's going to require maintenance and break down sometimes. And what does the actual mechanics of pods stopping for people to get out look like at busy stations? You'd need giant shunting yards to route pods to free "platforms" (like a garage forecourt only much bigger), and you'd have to build them underground in the places that are most expensive to do tunnelling and remove spoil from, and they'd have a complex layout that would have to be built in a labour-intensive way - all the lighting and ventilation and fire-suppression systems would have to be fitted into that layout (underground stations are already the most expensive parts of building subway projects). Or you'd have pods backing up onto the main line. Or you'd have to batch them together in a, uh, train.
> I don't think anyone disputes that. But this happens because there is no comparable point to point transit system and the existing road network is inefficient.
Tunnels are still only going to go from some places to some other places. Roads are already planned carefully. But there's only so much capacity you can get out of individual vehicles.
> The point of building tunnels is remove those conflict areas and provide direct point to point routes with very high throughput.
It's not going to be high throughput though. Every transportation system has a high throughput when you're just looking at maximum capacity along a plain line going at speed - it's all the fiddly interfaces where the throughput goes down. It'll be the same for a tunnel system (if it's ever built).