- reuses technology Tesla have already developed and are familiar with
- increases sales of said tech
- a controlled setting where they can get more data about real-world performance of their tech
- introduces their tech to punters
There's no doubt it will be less efficient, especially since the cars are carrying their batteries but the whole point of this is to cut CapEx by 90+% and eliminating miles of power lines is an important part of that effort.
It's a lower level of importance over this 3 mile trip, but it's not uncommon for power upgrades/maintenance to be a multi hundred million dollar Annual outlay for large transit systems.
A well-executed new metro line development, like the MetroSur line in Madrid, runs at around $60M per mile total cost including everything, all the construction and the trains and the stations and property rights.
If you want to improve on that by 90%+, you can't afford much more than just the batteries and the "skates" for this Dugout line.
But some of the costs of satisfying external stakeholders, like buying land; getting approval from regulators, politicians and the local community; and so on would be independent of tunnel size.
IMHO the main reason not to install a conventional metro line is Musk wants to demo something going 150mph, and I've never seen a metro line going that fast.
It may seem counterintuitive, but higher speeds mean less throughput. Most metro systems are concentrated on trying to move high numbers of people, so bumping train max speeds to 150mph just makes no sense. The sweet spot for speed appears to be around 50-70mph, which most metro trains will do when they're cruising between stations.
But anyways, anyone who is complaining that local mass transit ought to be going 150mph is someone who hasn't looked at why local mass transit can't go 150mph or is someone who doesn't want the "mass" in "mass transit".
It's like the difference between running everything on one very beefy server versus having a horizontally scaling architecture on commodity hardware. Each have pros and cons, so its not as simple as supposing that just because one is less beefy than the other, the total value of running on a supercomputer versus a cluster is >95% value for the supercomputer relative to the cluster.
If you have the goal of being able to scale to an arbitrary amount of traffic, building one very fast and expensive computer doesn't meet the traffic needs as well as building a lot of computers. A lot of things start leading to this conclusion. Economies of scale which come in when doing tons of a particular thing rather than just doing it once. Benefits of automation starting to factor in heavily for repeated projects rather than one off projects. Amortized costs of resources for repeated projects rather than one off projects.
Another way you can approach understanding the differences in benefit of this project compared to a subway is by thinking about above ground transportation. Why are roads good, since we already have highways? Why are lifts at ski resorts good, when we have airports? What benefit is there to a boat, when we have cargo ships? Why have cars, when we have trains? All of these questions are intended to point toward there being some value in a cheaper solution relative to a much more expensive but less practically extensive solution.
Yet another way you can understand the potential benefit relative to other transport costs is to imagine trying to optimize sign up for a site. If you reduce the number of fields in a sign-up form with the goal of making it easier to register, it is alike to reducing the cost of adding additional transport. In one case, the probability of a register increases. In the other case, the probability of a project which increases transport capacity increases. This analogy is imperfect, but I'm trying to give a analogy that technology people will understand through their experience: drive down cost, increase usage.
Finally and I think this is one of the key things that needs to be understood about the proof of concept: its intended to be a first and not a last. The idea in driving down costs in tunneling is that there is an arbitrary amount of untapped transport capacity going unused beneath us. Making use of the tiniest fraction of it will only have a tiny benefit, much like making use of a single commodity computer will only have a tiny benefit. Making use of a hundred thousand tiny fractions of it will have a much more pronounced impact, much like making use of thousands of computers can have substantial benefits. You need to keep in mind the end goal, because a journey isn't its first step: one of the superhuman evaluations of a position supposes that the value of a position is the value of the expectation of where you end up, not where you are
Sure. At some point it's worthwhile to actually built it in order to test the projections. And it looks like that point is now.
Another substantial part of this is that the Dugout Loop vehicles ascend to the surface through an elevator or ramp, which allows the station platforms to be built above ground. The platforms can also be smaller (they only need to fit a couple car-sized vehicles, not an entire train). This reduces cost substantially, as building out a large underground train platform is often one of the most expensive parts of a subway build.
The project will also have to deliver on the same demands wrt. fire suppression as any other tunnel.
The station building iself must be sized according to the number of people per hour, not the train length; the physical platforms that you use to step on/off trains on subways typically extend into the tunnels.
The website linked specifically says that it will not have tracks. The concrete itself acts as the 'guide' for the skates without needing any type of other tracks (such as traditional metal train tracks).
>The project will also have to deliver on the same demands wrt. fire suppression as any other tunnel.
This is also address by the website. Fire suppression isn't needed because the tunnel is made only of concrete. There is no concern about other materials (asphalt, high-voltage electrical, etc) catching on fire.
>The station building iself must be sized according to the number of people per hour, not the train length; the physical platforms that you use to step on/off trains on subways typically extend into the tunnels.
I'm not sure what you mean. Train stations are definitely sized according to train lengths. In my hometown, a significant roadblock to expanding our public train system is that we started off with short trains (and short stations to match), and it is very expensive to extend the length of the stations now that we use longer trains.
Wrt. fire suppression: Tesla can state whatever they want, but they will either have to fulfill the demands set forth in NFPA 502 for fire suppression in tunnels, or they will never be allowed to open to the public. Civil engineering is very strict on these things, for a very good reason.
For an underground train, you can have the train platform extend well beyond the actual station building. If you've ever taken the subway/tube in NY or London or similar you know what I mean.
If not, look e.g. at this Google Maps of the Gloucester Road tube stop:
https://goo.gl/maps/KQwuSz6RCUE2
The station building is the small square labeled Gloucester Road tube stop. But the platform extends all the way under the shopping mall (Waitrose's etc) out to the intersection between Cromwell Road and Ashburn Place.
Why wouldn't it need a fire suppression system? Because once a battery pack has started catastrophic failure, they are all dead anyways? The proposed system has more need for fire suppression than conventional subway, not less.
The article clearly states the tunnels with have a tunnel ventilation system, which is standard fire suppression strategy for underground tunnels for subways.
> electrical systems
They're already laying power cables for the tunnel lighting system which is presumably for their autonomous driving system with cameras, which now has a single dependency on their tunnel lighting system.
The rest, sure, I agree can be cheaper comparatively to a subway tunnel.
> steep inclines/sharper angles
Why would you want steep inclines (more than 5% standard for metros) if you're already underground and use lifts to ascend? The electric skates will still have a kinematic envelope that the tunnel has to adhere to, especially if they're for standing passengers. I don't see how this differs substantially from subway tunnels to allow steeper inclines/sharper angles.
Standard subway systems have to have much more in regards to fire suppression than just ventilation. The entire tunnel has to be protected from fire damage with insulation, some tunnels have sprinkler systems, etc. These wouldn't be needed in the dugout tunnel.
>They're already laying power cables for the tunnel lighting system which is presumably for their autonomous driving system with cameras, which now has a single dependency on their tunnel lighting system.
A power cable for low-voltage lighting is much, much, much different than the electrical systems required for powering a subway train. It's like the difference of running an extension cord to your lamp versus laying mains wiring for an entire house.
There's also no reason to believe that the vehicles would be dependent on the lighting system. Many cars on the market today already come with sonic- or LIDAR-based cameras that do not need lamps to detect obstructions. Tesla's own autopilot on its cars works just fine on unlit roads. There are also automated subway systems all over the world that don't depend on the tunnels being lit, and they don't seem to have any issue avoiding collisions.
>Why would you want steep inclines (more than 5% standard for metros) if you're already underground and use lifts to ascend? The electric skates will still have a kinematic envelope that the tunnel has to adhere to, especially if they're for standing passengers. I don't see how this differs substantially from subway tunnels to allow steeper inclines/sharper angles.
You wouldn't want them, but not being bound to the grade and angle restrictions of a tracked metro train means that you have more flexibility of where your system can go if you need to. It makes it easier to route around existing tunnels, piping, sewers, etc without having to make huge detours or being forced to use elevators.
An example of this is given in the linked website where they talk about the eastern terminus of the tunnel - it uses a ramp (from the looks of it, a ramp with a steepness that a subway train would never be able to use). If space allows for such a ramp, it's probably cheaper than using an elevator.
No they don't. There's no insulation for underground tunnels for subways; just steel liners and reinforced concrete, at least according to NFPA 130. No sprinkler systems. So no difference here.
> LV vs HV
Yes, true. HV cables and other supporting systems are more expensive. So the only difference here is batteries, which are far more combustible than cables in concrete.
> LIDAR cameras
Correct me if I'm wrong, but to my knowledge, Tesla has removed LIDAR-based cameras in their lineup.
> There are also automated subway systems all over the world that don't depend on the tunnels being lit, and they don't seem to have any issue avoiding collisions.
Yes, I know. And those trains communicate via access points placed on the tracks, each of which are run with fibre and power cables. Now that bare tunnel is no longer bare.
> incline
Sure, it's more flexible. It still doesn't address the claim that this will help make those tunnels cheaper.
Not sure where you're getting that info. I just looked up NFPA 130 and it specifically calls for multiple types of insulation (for cables as well as concrete and steel) as well as automated water sprinkler systems, and even standpipes for firefighters.
>Correct me if I'm wrong, but to my knowledge, Tesla has removed LIDAR-based cameras in their lineup.
Tesla doesn't use LIDAR but they do use ultrasonic sensors. Tesla's existing autopilot systems works just fine on completely dark, unlit roads. I don't see why these skates would be any different.
>Yes, I know. And those trains communicate via access points placed on the tracks, each of which are run with fibre and power cables. Now that bare tunnel is no longer bare.
Even with fibre and low voltage power for access points (if even needed), you're still leagues away from the cost and complexity of a full tracked third-rail system.
>Sure, it's more flexible. It still doesn't address the claim that this will help make those tunnels cheaper.
I'm not sure what you mean. Flexibility is cheaper. Being able to use a tunnel that is 500 feet long at a 10% grade is cheaper than having to build a tunnel that is 1500 feet long at 3% grade.
You're moving the goalpost and earlier insinuated that tunnels for subways require lining and other fire suppression systems. They do not. The automated sprinklers are needed for stations, but not the tunnel itself. Same for standpipes.
The insulation for cables are a given, because you will need it for ALL tunnels. Not just HV cables, but ALL cables, including fibre and low-voltage cables.
Thus, no difference.
> Tesla doesn't use LIDAR but they do use ultrasonic sensors. Tesla's existing autopilot systems works just fine on completely dark, unlit roads. I don't see why these skates would be any different.
Does Tesla's existing autopilot system using LIDAR/ultrasonic cameras have automated collision protection that does so without having to communicate with the other cars behind it traveling at 120mph to 150mph? No. You will need to modify the system to do this, and you will need to add the mesh network to the tunnels, among other things I described.
> Even with fibre and low voltage power for access points (if even needed), you're still leagues away from the cost and complexity of a full tracked third-rail system.
Agreed. Tracks and third rail are not cheap, comparative to not using them at all.
> I'm not sure what you mean. Flexibility is cheaper. Being able to use a tunnel that is 500 feet long at a 10% grade is cheaper than having to build a tunnel that is 1500 feet long at 3% grade.
How? If the total length of the overall tunnel is the same, the cost of materials are not substantially different.
I haven't moved the goalposts whatsoever. My original comment, which you replied, to, specifically said that a significant part of the cost savings of this project involves the differences in station construction. Stations have always been part of the discussion.
I think you may be confused by the terminology. "Tunnel" does not mean only the trackway where trains travel. Subway stations can also be part of the "tunnel". This is why in my earlier comment I specifically said that "some tunnels require water sprinklers" - because the tunnel parts that are stations require them, even though the tunnel parts that are trackway do not.
>The insulation for cables are a given, because you will need it for ALL tunnels. Not just HV cables, but ALL cables, including fibre and low-voltage cables.
Sure, and the fact that you need significantly less cabling for this tunnel means that you will save on having to insulate those cables.
>Does Tesla's existing autopilot system using LIDAR/ultrasonic cameras have automated collision protection that does so without having to communicate with the other cars behind it traveling at 120mph to 150mph?
Yes? I'm not sure what the point of this question is. Tesla's current autopilot functions every day relying only on cameras and no communication with surrounding vehicles. If anything, the fact that you're removing the variability of cars on the road and will only be interacting with other skates makes this less of a problem.
>How? If the total length of the overall tunnel is the same, the cost of materials are not substantially different.
The total length of the tunnel isn't the same. In my above example, 500 ft vs 1500 ft.