The Boring Company Tunnel Event - December 18, 2018

[HankLloydRight]

Even as an A to B route, an few express tunnels on the 405 with ramps every 5 miles could be a big boost to throughput.

Right, but then you have to use each individual's car instead of a skate or a pod, so every car would need the guide wheels installed (would the added weight affect range?), as well as some system check to make sure there's enough state of charge in the car before letting it into the tunnel.

Is Elon going to let any conforming ICE with guide wheels into the tunnel, or only Teslas? If it's only Teslas, that's not going to make a dent in the overall traffic situation in LA.

Or is it going to be a huge fleet of skates that each car will ride on? That makes it very expensive, since each skate is almost the marginal cost of the most expensive Tesla parts -- body, battery, and drive train just without an interior, doors, or a roof.

Also, I don't think TBM in its current form is big enough to drill subway tunnels. And I think Elon is going to run into trouble building several tunnels one on top of another. They tried that in Boston with the Big Dig at the Amtrak train yard, but after drilling the tunnel, they realized the soil was too soft, so they had to install huge freezers and plumbing to keep the ground frozen year-round so it doesn't collapse the tunnel below. Imagine how expensive that is.

 

[Xminus6]

Cost of wheel system (at any number) is less than the cost of a skate. The wheel system allows car usage of the loop to be determined by user demand, not skate count. With wheels, loop capacity is loop size/ design. With skates, loop capacity is (charged) skate count. This new setup allows Boring to focus pods on pedestrians/ cyclists instead of splitting capex on pedestrians vs cars.

Loop demand will be limited by the number of cars that are able to have and have opted to install retrofitted wheels. This will clearly be a much lower number than the number of cars that would use the skate, even if limited only to BEVs. If TBC is working on pedestrian and cyclist pods, it would make more sense to just use that research to also figure out the skate, which was previously intended to be the same platform.

Of course no one currently package protects for loop wheels. However, if the loop system exists, other companies will have an incentive to either have loop wheels as a factory option or include mount points for them. Else, Tesla eats more of their lunch.

I find this extremely optimistic. I've owned 5 EVs, three of them Teslas, and I would never opt to retrofit Loop wheels for our cars. The chances we would use one are so low that it wouldn't make sense. As a result, if we were in LA, which we are fairly often, it wouldn't be an option for us to try it. How many people are going to retrofit wheels to their car for a sight-unseen service that they'll use infrequently? The initial buy-in and barrier to entry make it useless for anyone who isn't along the commute corridor of the Loop.

SOC is an issue with skates or without, only change is the initial charge requires for a trip. AC or heat can drain a pack to 0 just like driving can. Loop needs a method to reject cars on 0-x%. Failed cars can be retrieved by a tow vehicle sent from the 'wrong' direction while traffic takes a parallel route (worst case, a route goes half-duplex).

Come on. You're being disingenuous. You can run a Tesla's HVAC system for a long time when it's sitting "stationary." Having skates that are charged and managed by the tunnel operator is the only logical solution. Relying on each car's SOC and charge system is just silly. You're introducing many more points of failure and variables into management.

Putting the charging on the individual cars vs the skates reduces the quantity of skates needed even further.

For Cars:
Skates: >>1 skate required per car, Boring must have sufficient skates in the correct location for peak traffic.
Wheels: 1 set per car, user purchased

Hyperloop uses partial vacuume tunnels, pods are required for that system.

I'm familiar with the difference between HyperLoop and Loop. But I simply disagree that it makes sense to force every car to retrofit wheels and rely on those peoples' management of their own SOCs to be problematic. Almost nobody would ever be able to try the Loop system with a car without taking the plunge to permanently alter their own cars for it.

I'm as big a Tesla fan as most people on here, which is to say rabidly loyal compared to the rest of the world. I cannot think of what they would need to do to convince me to retrofit my car for this. It's just a poor solution IMO.

 

[mongo]

Right, but then you have to use each individual's car instead of a skate or a pod, so every car would need the guide wheels installed (would the added weight affect range?), as well as some system check to make sure there's enough state of charge in the car before letting it into the tunnel.

Weight would have minimal impact overall. Rolling << aero, wheels << vehicle, sub 5%. SOC requirements exist regarding of skate, else someone gets on a skate at 1% with heater blasting for a 15 minute trip...

Is Elon going to let any conforming ICE with guide wheels into the tunnel, or only Teslas? If it's only Teslas, that's not going to make a dent in the overall traffic situation in LA.

No ICE, ventilation is not sufficient. Any autonomous capable EV. Teslas goal is to accelerate adaption on renewable transport. It a person can chop 30 minutes off their commute, that is an incentive to get an EV.

Or is it going to be a huge fleet of skates that each car will ride on? That makes it very expensive, since each skate is almost the marginal cost of the most expensive Tesla parts -- body, battery, and drive train just without an interior, doors, or a roof.

Which is why the wheel kit concept makes sense. Especially since skates can't carry ICE. Need to build an EV and a skate, or just an EV.

Also, I don't think TBM in its current form is big enough to drill subway tunnels. And I think Elon is going to run into trouble building several tunnels one on top of another. They tried that in Boston with the Big Dig at the Amtrak train yard, but after drilling the tunnel, they realized the soil was too soft, so they had to install huge freezers and plumbing to keep the ground frozen year-round so it doesn't collapse the tunnel below. Imagine how expensive that is.

Right, the current TBM is the wrong size. However, the tech Boring is developing can be used on a large diameter one.

Not sure on Big Dig's problems, but a tunnel can be fully stable. Smaller diameter ones are easier to do thus with. Concider concrete culverts under roads, very shallow and support truck traffic.

 

[mongo]

Loop demand will be limited by the number of cars that are able to have and have opted to install retrofitted wheels. This will clearly be a much lower number than the number of cars that would use the skate, even if limited only to BEVs. If TBC is working on pedestrian and cyclist pods, it would make more sense to just use that research to also figure out the skate, which was previously intended to be the same platform.

Boring / Tesla wants to push people into EVs. Boring tunnels cannot support ICE. So the existence of a way for people in supported markets to chop a huge chunk of time and stress off their commute will cause them to concider buying a Loop compatible vehicle. Don't just look at current support (at 0 cars), look at a couple years down the road when the system exists.

I find this extremely optimistic. I've owned 5 EVs, three of them Teslas, and I would never opt to retrofit Loop wheels for our cars. The chances we would use one are so low that it wouldn't make sense. As a result, if we were in LA, which we are fairly often, it wouldn't be an option for us to try it. How many people are going to retrofit wheels to their car for a sight-unseen service that they'll use infrequently? The initial buy-in and barrier to entry make it useless for anyone who isn't along the commute corridor of the Loop.

I wouldn't get a Loop compatible car either, but I have an 8 miles commute in Michigan. We are not the target market. People pay for fast-pass and modify behaviors for HOV lane usage, they will pay for a one time modification to reclaim hours of their life.

Come on. You're being disingenuous. You can run a Tesla's HVAC system for a long time when it's sitting "stationary." Having skates that are charged and managed by the tunnel operator is the only logical solution. Relying on each car's SOC and charge system is just silly. You're introducing many more points of failure and variables into management.

It's not disingenuous, a car can hit 0% on skate just like it can hit 0% driving. Without an SOC limit, both cases are possible. (Can you tell I write software). Force exiting/ deny admittance at 10% SOC.

As to other failure modes: Impose a yearly service check up for cars to reduce failure rates, but a skate is pretty much the same complexity as a car.

I'm familiar with the difference between HyperLoop and Loop. But I simply disagree that it makes sense to force every car to retrofit wheels and rely on those peoples' management of their own SOCs to be problematic. Almost nobody would ever be able to try the Loop system with a car without taking the plunge to permanently alter their own cars for it.

People don't manage the SOC, the system does. Those that live in area serviced could get a demo ride with Boring, ride the people skate, or take a trip with a co-worker.

I'm as big a Tesla fan as most people on here, which is to say rabidly loyal compared to the rest of the world. I cannot think of what they would need to do to convince me to retrofit my car for this. It's just a poor solution IMO.

I'm guessing you don't take the subway either, but that doesn't mean it is pointless. It doesn't need to work for you, it needs to work for the people commuting in LA who travel the 405 and other jammed roads every day.

 

[HankLloydRight]

Any autonomous capable EV.

Ok, so the "wheel kit" will need to be installable and adaptable to any BEV. That's a pretty high hurdle to get other anti-Tesla BEV manufacturers (competitors) to design and build their sub-frames to accommodate the hidden Loop wheel kit. Also realize that nobody has seen exactly what Tesla had to do to install that wheel kit on just one demo Model X. And is that demo wheel kit auto-retractable like in the cute little video? I don't think so.

That seems like an awful lot of technology to have other BEV manufacturers pre-build into their cars, even if it's just the space, mounting brackets and wiring harnesses to accommodate installation after-the fact. I don't know where this $300 number came from, but that makes each wheel only $75? How much do you think it will cost, much less install, a retractable hardened steel wheel frame, guide wheels that can go up to 150mph, and all the extra gears, motors, actuators, and levers to make it fully retractable and hidden from view when not being used? Not $75 each, that's for sure.

It a person can chop 30 minutes off their commute, that is an incentive to get an EV.

So a few people with specially modified Teslas can chop 30 minutes off their commute. Great. Still not solving the soul crushing traffic problems that plague LA.

Also, has anyone thought about this? Let's just say there are multiple parallel tunnels "in 3D!" as Elon says, they're going to have to have the car lifts descend hundreds of feet to get to an open/available Loop tunnel. And the car lifts are single threaded in each direction. One car down, one car up. Another car down, another car up. Just the delay in getting one car from the surface to one of the tunnels, including drive-on and drive-out is going to severely limit how many cars can be in the tunnel at once.

So you've now created a surface level traffic jam waiting for the car lift to get down into the tunnel.. and then what about on the other end? You have five cars all getting off at the same stop.. you've now created a traffic jam getting out of the tunnel waiting for the lift. Then what happens if 10 cars want to get off, but the lift waiting area is full? Does traffic back up into the main loop tunnel? Or are you forced to go to the next available stop? And if that one is overloaded, then what? You're now 5 stops past where you want to go. To solve this problem, Elon would have to bore massive waiting areas underground so the main Loop tunnels aren't affected by traffic waiting to get out. But then you still have to wait for X cars in front of you to get lifted to the surface. One.at.a.time. MORE SOUL CRUSHING TRAFFIC! The more I think about this the more I'm convinced it's a really, really, bad idea.

 

[mongo]

ou

Ok, so the "wheel kit" will need to be installable and adaptable to any BEV. That's a pretty high hurdle to get other anti-Tesla BEV manufacturers (competitors) to design and build their sub-frames to accommodate the hidden Loop wheel kit. Also realize that nobody has seen exactly what Tesla had to do to install that wheel kit on just one demo Model X. And is that demo wheel kit auto-retractable like in the cute little video? I don't think so.


That seems like an awful lot of technology to have other BEV manufacturers pre-build into their cars, even if it's just the space, mounting brackets and wiring harnesses to accommodate installation after-the fact. I don't know where this $300 number came from, but that makes each wheel only $75? How much do you think it will cost, much less install, a retractable hardened steel wheel frame, guide wheels that can go up to 150mph, and all the extra gears, motors, actuators, and levers to make it fully retractable and hidden from view when not being used? Not $75 each, that's for sure.

I don't know how I got to 300 either. One actuator for down, one for extend (could do it all with one instead, or reuse air suspension air supply (Teslas)). The idlers are only on the front wheels and mostly only deal with failure modes. So one unit, two wheels.

So a few people with specially modified Teslas can chop 30 minutes off their commute. Great. Still not solving the soul crushing traffic problems that plague LA.

Or anyone else with a compatible EV who drives. Pods increase commute speed for non-vehicle peeps. And the number of Teslas in increasing.Give people an extra hour a day and there will a lot more loop enabled EVs out there.

If you are parent dropping off two kids at daycare on your way to work in the city, the subway is not your optimal solution...

Also, has anyone thought about this? Let's just say there are multiple parallel tunnels "in 3D!" as Elon says, they're going to have to have the car lifts descend hundreds of feet to get to an open/available Loop tunnel. And the car lifts are single threaded in each direction. One car down, one car up. Another car down, another car up. Just the delay in getting one car from the surface to one of the tunnels, including drive-on and drive-out is going to severely limit how many cars can be in the tunnel at once.

So you've now created a surface level traffic jam waiting for the car lift to get down into the tunnel.. and then what about on the other end? You have five cars all getting off at the same stop.. you've now created a traffic jam getting out of the tunnel waiting for the lift. Then what happens if 10 cars want to get off, but the lift waiting area is full? Does traffic back up into the main loop tunnel? Or are you forced to go to the next available stop? And if that one is overloaded, then what? You're now 5 stops past where you want to go. To solve this problem, Elon would have to bore massive waiting areas underground so the main Loop tunnels aren't affected by traffic waiting to get out. But then you still have to wait for X cars in front of you to get lifted to the surface. One.at.a.time. MORE SOUL CRUSHING TRAFFIC! The more I think about this the more I'm convinced it's a really, really, bad idea.

on/ off tunnels are the shallowest. One minute cycle time for elevator, ramps and spirals (full duplex) are continuous flow. Multiple shallow stops load at the same time, elevators clear the tunnel, X cars depart. That tunnel merges into the trunk line (maybe mid speed with high speed seperate). Later, split back to local that terminates in elevator, ramp, spiral, or directly to parking garage/ lot.

Entry/ exit points would be designed for traffic flows, system can be like ATC and stagger arrivals to reduce backups. Portland and other cities gate cars at on ramps, same principle.
Worst case, they time divide access, soul destroying traffic 2 days a week, tunnels 3. Or it causes people to shift their commute hours while still saving time.

 

[HankLloydRight]

The idlers are only on the front wheels and mostly only deal with failure modes. So one unit, two wheels.

Are you sure about only two front wheels needed? That just doesn't sound right. Maybe on a straight-line tunnel, but anything with curves you're going to need four guide wheels.

.Give people an extra hour a day and there will a lot more loop enabled EVs out there.

But at what cost? The Loop certainly isn't going to be free. So people have to (1)Buy a Tesla, (2) upgrade to guide wheels, and (3) pay tolls for the Loop. You make it sound like everyone and anyone will do that, but in reality, most people will take the lower cost solution over convenience. But the end cost is going to be very, very high for both the vehicles and the tolls to use the loop.

on/ off tunnels are the shallowest. One minute cycle time for elevator, ramps and spirals (full duplex) are continuous flow. Multiple shallow stops load at the same time, elevators clear the tunnel, X cars depart. That tunnel merges into the trunk line (maybe mid speed with high speed seperate). Later, split back to local that terminates in elevator, ramp, spiral, or directly to parking garage/ lot.

"on/ off tunnels are the shallowest." - what? How are you going to go from level 10 to level 1 to get back to the street? What you're describing is a very complex web of main and interconnect and ramp tunnels on and off each main line.. it's not just about boring a few parallel tunnels anymore, it's now about the tunnels and all the interconnects and ramps and spirals and elevators. And who's going to pay for all of that?

Entry/ exit points would be designed for traffic flows, system can be like ATC and stagger arrivals to reduce backups. Portland and other cities gate cars at on ramps, same principle. Worst case, they time divide access, soul destroying traffic 2 days a week, tunnels 3. Or it causes people to shift their commute hours while still saving time.

"Stagger arrivals" means TRAFFIC and more waiting. That's not a solution. And where is Elon going to get all this prime real estate to put the car lifts in the most convenient places? And if not there, people will still have to find some other method of getting to their final destination.

 

[mongo]

Are you sure about only two front wheels needed? That just doesn't sound right. Maybe on a straight-line tunnel, but anything with curves you're going to need four guide wheels.

Why? The rear wheels follow the front, just like cars currently do. The side wheels are only for safety and to prevent losing paint if the steering algorithm is a tad off. For rear guides to touch the walls would mean either the curve was made too tight, or the back end slid out.
Boring demo was only front guides.

But at what cost? The Loop certainly isn't going to be free. So people have to (1)Buy a Tesla, (2) upgrade to guide wheels, and (3) pay tolls for the Loop. You make it sound like everyone and anyone will do that, but in reality, most people will take the lower cost solution over convenience. But the end cost is going to be very, very high for both the vehicles and the tolls to use the loop.

Highways aren't free either, traffic itself is an economic cost.
Say it stays at 10 million per mile with a 30 year life. 330k/mile/year or 1k per mile per day. Following distance of 3 seconds, 20 cars a minute, 1,200 an hour. 6 peak hours a day: 7,200 cars per mile per day. Cost: $0.13 per car per mile per day. Monthly cost: $4 a mile. Less than the coffee they drink on just one trip.

Of course, pods would also contribute.
At $1 a mile ride cost, you need a whopping 1,000 riders a day (per mile) to cover install costs. Over 4 hours peak travel period, that is 250 riders an hour or 32 pods at 8 people per pod per mile.

Apply cost, ridership, and system size factors as needed, but it seems financially feasible to me.

"on/ off tunnels are the shallowest." - what? How are you going to go from level 10 to level 1 to get back to the street? What you're describing is a very complex web of main and interconnect and ramp tunnels on and off each main line.. it's not just about boring a few parallel tunnels anymore, it's now about the tunnels and all the interconnects and ramps and spirals and elevators. And who's going to pay for all of that?

Tunnels do not need to stay at the same depth, near surface tunnel can curve to the side and descend to join with the trunk, just like freeways do now. You would transition back to the surface before exiting. Or trunks are to the side at the same level, that works too. Key thought: if there are depth levels, put the faster tunnels deeoer.

Really, loop can be thought of as roads with cheaper overpasses and no/ less intersections. Can also be thought of like a moving sidewalk with multiple lanes/ speeds.
Still, no reason they couldn't do a deep spiral, other than land requirements.

"Stagger arrivals" means TRAFFIC and more waiting. That's not a solution. And where is Elon going to get all this prime real estate to put the car lifts in the most convenient places? And if not there, people will still have to find some other method of getting to their final destination.

2 parking spots == elevator site. Who owns the streets of LA? Parking garages would be able to add access. New buildings with parking could integrate access at construction. Parking lots/ decks aren't free now.

There are also the people driving through to get somewhere else, loop could have ramps outside the city proper to bypass the bulk of city traffic.

If people are driving, they drive to their final destination from the exit. If on pods, they walk from the terminus. Pods have priority so the walk will be a known length.

 

[HankLloydRight]

The entire purpose of the guide wheels is that you don't need ANY autonomous driving technology. It's no different than rail. The guideways (curbs) and guide wheels steer the car in the direction it needs to do. No autopilot needed.

So, we'll just have to agree to disagree. I think it's a terribly horrible idea with dozens of fatal flaws that will cause it to never see the light of day.

Just because it comes out of the mind of Elon Musk doesn't mean it's a good idea.

 

[mongo]

The entire purpose of the guide wheels is that you don't need ANY autonomous driving technology. It's no different than rail. The guideways (curbs) and guide wheels steer the car in the direction it needs to do. No autopilot needed.

So, we'll just have to agree to disagree. I think it's a terribly horrible idea with dozens of fatal flaws that will cause it to never see the light of day.

Just because it comes out of the mind of Elon Musk doesn't mean it's a good idea.

The wheel are there for positive (fail safe) control only.

Unless the guide wheels are hard mounted to the front wheels and placed far enough forward to pivot them (with a steering rack that can be back driven), they won't steer the car. Instead the front tires would be scrubbing sideways continuously.

You slso need autopilot for system interface and speed control.

There currently exist vehicle tunnels that people drive through. How is this idea less practical? How does requiring the cars be EV and the addition of positive position control introduce "dozens of fatal flaws"?

 

[Xminus6]

Boring / Tesla wants to push people into EVs. Boring tunnels cannot support ICE. So the existence of a way for people in supported markets to chop a huge chunk of time and stress off their commute will cause them to concider buying a Loop compatible vehicle. Don't just look at current support (at 0 cars), look at a couple years down the road when the system exists.

I understand the need to limit it to EVs but in reality it will be limited to newly-built Teslas. The logistics and mechanics of retrofitting those wheels are not going to be so easy that it'll cost only $300. Literally, what can you do on your car that costs $300? You can probably get your hood wrapped and that's it. Nothing that you would need to entrust with the safety of your vehicle is a $300 retrofit.

I wouldn't get a Loop compatible car either, but I have an 8 miles commute in Michigan. We are not the target market. People pay for fast-pass and modify behaviors for HOV lane usage, they will pay for a one time modification to reclaim hours of their life.

On the contrary, our family would definitely benefit from something like the Loop. I was really looking forward to it when it made more sense to me with skates. If they ever build an East Bay to SF specific Loop then we would look into it for sure.

It's not disingenuous, a car can hit 0% on skate just like it can hit 0% driving. Without an SOC limit, both cases are possible. (Can you tell I write software). Force exiting/ deny admittance at 10% SOC.

But if a car runs out of charge on a surface street it just gets stuck in one of many lanes or is pulled over onto the shoulder. Unless Loop has many tunnels on the same routes and can easily re-route cars between those tunnels, a car breaking down or running out of charge in the Loop is a massive failure point. It'll shut down the whole system. It's like when a BART train gets stuck in the city. Since it's the singular path for every train line, it shuts the whole system down in that direction.

As to other failure modes: Impose a yearly service check up for cars to reduce failure rates, but a skate is pretty much the same complexity as a car.

But it would be centrally managed. When I ride the subway or take a bus, my ride isn't disturbed when my, or some other random person, doesn't maintain their subway car or municipal bus. Centrally maintaining the skates and their SOC make much more sense since they manage the system.

People don't manage the SOC, the system does. Those that live in area serviced could get a demo ride with Boring, ride the people skate, or take a trip with a co-worker.

I'm guessing you don't take the subway either, but that doesn't mean it is pointless. It doesn't need to work for you, it needs to work for the people commuting in LA who travel the 405 and other jammed roads every day.

On the contrary I rode the NYC subway and BART for many years and I commuted up and down the 405 for many years as well. Again, the strength, and weakness, of those systems it he centralized management. If everyone who rode on the subway was responsible for their own subway car it would be a disaster. It would essentially be an underground version of the 405 with less escape routes.

I just don't see how mixing system managed vehicles and random consumer vehicles together is a good way to assure clear and expedient travel. The benefit of the skate system is that it centralizes the management and allows them to build to their specific need. Also, slower EVs will slow down the whole system. Driving a Chevy Bolt on retrofitted bumper wheels will slow down everyone in the tunnel. Hardly any EV right now can sustain the purported 150mph for any reasonable distance. So will everyone just have to slow down to 60mph as some people are trying to preserve their charge?

If they want to build a Hyperloop at some point in the future, the pod issue is going to have to be resolved anyway. So why not work on a logical solution to pod-light now?

 

[mongo]

I understand the need to limit it to EVs but in reality it will be limited to newly-built Teslas. The logistics and mechanics of retrofitting those wheels are not going to be so easy that it'll cost only $300. Literally, what can you do on your car that costs $300? You can probably get your hood wrapped and that's it. Nothing that you would need to entrust with the safety of your vehicle is a $300 retrofit.

If that is the way it is, that is on the other OEMs, not Tesla. Same as the Supercharger network, others can join, they just choose not to.
$300 was a wild guess, but it need not be super expensive. A tow hitch is arguably more critical than guide wheels and costs $400, so it might run $1k. $80 $17 a month over 5 years. Having a guaranteed available skate would have to cost more than that.

Sure, skates are great from a system control point of view. However, they would also be the limit to system capacity. Only 1,000 skates? Then only 1,000 cars at a time. Everyone heading out of town? Need to wait till the skates return to reuse. Not efficient. It also removes the battery and motor manufacturing capacity from the EV side of things.
Say a skate is 40kWh and an EV is 80kWh. 960 kWh gets you 12 EVs, or 8 EVs plus skates.

On the contrary, our family would definitely benefit from something like the Loop. I was really looking forward to it when it made more sense to me with skates. If they ever build an East Bay to SF specific Loop then we would look into it for sure.

Gotcha, so you are in the passing through case. I can see why skates would appeal to that use case. Other than trying it out ahead of time, any reason you would not want to have a loop express lane available to take instead of a surface route?

But if a car runs out of charge on a surface street it just gets stuck in one of many lanes or is pulled over onto the shoulder. Unless Loop has many tunnels on the same routes and can easily re-route cars between those tunnels, a car breaking down or running out of charge in the Loop is a massive failure point. It'll shut down the whole system. It's like when a BART train gets stuck in the city. Since it's the singular path for every train line, it shuts the whole system down in that direction.

Loop would have more than one track in a direction (worst case, go half duplex in one tunnel) So a single failure does not have the same impact as with a subway failure or Amtrak's crossings issues. Reversing a vehicle may be an issue...

If you take a bus, you are at the mercy of all other drivers on the route and their cars. On the subway, you have points of delay also. Someone standing in the door will stop the train and delay those behind it , along with people on the tracks. It is not a common thing, but it is a thing.

Sure, in the idealized case, there would be no breakdowns, but transport systems already operate with acceptance that a crash stops traffic.
Two lane roads are all single point of failure, Mackinaw bridge is 5 miles of no gas stations, Overseas Highway has a 20 mile stretch of no stopping points. Admittedly less traveled, but usable. Interstates are one pile up or fire from stoppage.

A properly maintained car should be as reliable as a skate. Possibly more since personal cars are likely treated nicer than 3rd party car movers.

I just don't see how mixing system managed vehicles and random consumer vehicles together is a good way to assure clear and expedient travel. The benefit of the skate system is that it centralizes the management and allows them to build to their specific need. Also, slower EVs will slow down the whole system. Driving a Chevy Bolt on retrofitted bumper wheels will slow down everyone in the tunnel. Hardly any EV right now can sustain the purported 150mph for any reasonable distance. So will everyone just have to slow down to 60mph as some people are trying to preserve their charge?

With multiple tracks are different speeds the impact of a slow EV is minimized. Force Bolts to stay in the lower speed local tunnels.

All the vehicles are managed, the drivers will not be setting their speeds, the central system will via wireless link.
Even in the skate scenario, you still need an override system for the cars using it. You also need the car to get off the skate at the end. Having skates does not eliminate the issue of a car failing while in the loop, it just concentrates it at the end points (assuming the system even allows the skate to move if it loses telemetry with the car).

If they want to build a Hyperloop at some point in the future, the pod issue is going to have to be resolved anyway. So why not work on a logical solution to pod-light now?

They are different use cases until lots of people commute via hyper and still need cars. However, hyperloop can transport cars in pods for added cost.

LA level traffic is more people over short and more diffuse routes.

Edit: math error

 

[HankLloydRight]

Some interesting observations in this article:

Elon Musk's Loop tunnel relies on century-old roller coaster tech

 

[mongo]

Some interesting observations in this article:

Elon Musk's Loop tunnel relies on century-old roller coaster tech

Interesting but incorrect when applied to Loop. If the track side wall are 1 foot further apart than the guide wheels, then AP has a 1 foot margin of error before getting nudged. Cars track better than that now. With side range detectors (maybe the exiting ultrasonics), the side wheels would normally never touch at all.

Roller coaster car guidance: Purely side wheels
Loop Guidance: Autopilot steering with wheels as fail safe

Roller center of gravity: high due to people
Loop: low due to pack and drive train

Roller track curves: flat(ish)
Loop track curves: banked (in future), no issue with yaw induced motion sickness
Article mixed old school pure side wheel coasters (yaw issues) with modern wheels on all sides of pipe high speed no yaw issue coasters.


Roller: requires up-friction wheel to avoid leaving track "Leap-the-Dips"
Loop: Car would not leave path to being with, due to limited incline and no hills (see normal driving on similar terrain). No need for up-wheels (see also COG)

 

[HankLloydRight]

Roller coaster car guidance: Purely side wheels
Loop Guidance: Autopilot steering with wheels as fail safe

Absolutely not true at 150 mph. Not even 100mph. I can assure you that the guide wheels will be the primary steering method on loop tunnels, not AP.

What about non-Tesla BEV cars without AP?

 

[HankLloydRight]

then AP has a 1 foot margin of error before getting nudged.

"Before getting nudged". LOL Do you realize what "getting nudged" is like traveling at 150 mph?

Even the few demo rides people got were "a bit shaky" at 45mph on a straight shot run.

 

[mongo]

Absolutely not true at 150 mph. Not even 100mph. I can assure you that the guide wheels will be the primary steering method on loop tunnels, not AP.

What about non-Tesla BEV cars without AP?

Elon said (and the slide showed) EVs with autonomous driving. No non-self driving (steering + speed) cars.

How can you assure that AP will not be primary steering? In the presentation he said they bracing and prevent deviation on mechanical failure. They can't be totally rigid, maybe spring loaded dampers, but if you don't have active steering, its is going to kill the tires. 1 degree steering misalignment is 92 feet of sideways slide per mile if travel.

"Before getting nudged". LOL Do you realize what "getting nudged" is like traveling at 150 mph?

Even the few demo rides people got were "a bit shaky" at 45mph on a straight shot run.

Forward speed has nothing to do with how hard a sideways nudge is, lateral speed does. That is tied to the angle of deviation, that angle will not be very large in a system designed for high speeds. If the wheels are spring and damper setups, then the lateral correction is progressive. Ideally with sensors to adjust steering before the limit is reached.

Keep in mind, that was a demo on their first tunnel. The sunken track approach was not in the original presentations and so may have been a latter add on. As such, it is not currently tuned or optimized for that setup. Elon even stated that they ran out of time to smooth the guides and later tweeted that the real tunnels would have banked curves. He also mentioned they did the run at over 100 MPH.

 

[HankLloydRight]

EVs with autonomous driving. No non-self driving (steering + speed) cars.

Oh, so now it's not just AP, but autonomous? So we're back to Teslas only allowed. How is any other manufacturer going to produce FSD cars nearly as good as Teslas? What if they're only half as good? Is there going to be a SAT test the cars have to take before being allowed into the Loop? <sarcasm>

How can you assure that AP will not be primary steering?

I'm as sure as you're sure this is a good idea. It's TRIVIAL to have a mechanical steering link actively steering the front wheels EXACTLY where they should be, rather than extremely indirect/passive steering system relying on millions of lines of AI/NN code for AP/FSD to steer the car inside a tunnel for a car going 150mph. Just a single AP/FSD failure inside the tunnel or misreading the guidelines or track, and the car starts ping-ponging against the track or walls inside the tunnel, and you have a disaster and people die. Or you could have an active direct link system that steers the car with almost no chance of failure and no buggy software.

 

[mongo]

Oh, so now it's not just AP, but autonomous? So we're back to Teslas only allowed. How is any other manufacturer going to produce FSD cars nearly as good as Teslas? What if they're only half as good? Is there going to be a SAT test the cars have to take before being allowed into the Loop? <sarcasm>

You're losing me. AP is a Tesla term, autonomous is a general term. FSD is major overkill, loop is a complete knowledge system without signage or pedestrians. Any EV with self-steering and self-speed control (plus any need telemetry hooks) would work in Loop. Otherwise you are back to distracted drivers causing accidents. It is the simplest form or platooning possible (well not the absolute simplest since there are exits/ entrances of n>1).

I'm as sure as you're sure this is a good idea. It's TRIVIAL to have a mechanical steering link actively steering the front wheels EXACTLY where they should be, rather than extremely indirect/passive steering system relying on millions of lines of AI/NN code for AP/FSD to steer the car inside a tunnel for a car going 150mph. Just a single AP/FSD failure inside the tunnel or misreading the guidelines or track, and the car starts ping-ponging against the track or walls inside the tunnel, and you have a disaster and people die. Or you could have an active direct link system that steers the car with almost no chance of failure and no buggy software.

Attaching into a rack and pinion steering system is not simple.The links are not made for those forces, nor for something to attach to them. They also do not have a method to convert torque to lateral motion. You would need to add a linear cross vehicle slide and attach that to the rack in non-bennding moment way for displacement. Attaching to the wheels directly would also be difficult since you would need a bearing, matching lug pattern, and to extend the idle wheels out far enough to produce the needed torque (they would also not be foldable). In both cases, you also need a rack that doesn't mind being back driven a lot. A purely mechanical system also fails at junctions unless Loop uses RR style switches.

Electric power steering is software based, yet there is no concern with two cars passing each other at > 100MPH and a couple foot clearance. ABS and traction control are software based. If it decided to lock up one wheel, you're much more in trouble on the open road than on Loop. If FSD has HW/SW sufficient for open road use, it is sufficient for Loop where the vehicle is trapped inside a guide path. I'm sure you've seen the X rollover test, how are people going to die (even if the wheel/ guide slop was sufficient to allow ping ponging)?

The software need not be FSD level as I mentioned above. Could run steering off a simple PID if you wanted. Outfit each spring loaded damped idler wheel assembly with a displacement sensor and set the control loop to equalize those two (plus redundant) readings. An add on idler wheel assembly could also use in track inductive or magnetic guidance like AGVs/ AGCs in factories use. Sub inch cross track precision. RFID tags can provide exact location information.

SpaceX is running control loops at speeds much faster that 150MPH...

You do give me an idea for a variant. A mini skate that handles guidance and telemetry. The vehicle's front wheels sit on the skate while the vehicle's rear wheel provide the motive power. Skate would be 3-4 wheel with a drive motor and steering for pre-positioning and also allow it to recharge from the vehicle's pack via regen.
Only vehicle mods needed are software and a data port/ interface. The skate would latch onto the front wheels (or a standardized coupler).
This eliminates the steering from the vehicle systems and the need for a large pack and recharging from the skate systems. Still have the problem of Loop capacity/ availability limited to skate count along with dimensional issues.

 

[HankLloydRight]

Yes, there are options, most overly complex ones or risky ones that's going to take years of building and testing new technology (and the associated costs)... which brings me back to my main point. What do we need all this complex and expensive technology for when we solved this problem 115 years ago with subway cars on railroad tracks?

As I've said before, let Elon dig really cheap, really big tunnels. Then just run subway trains in them. Or if he wants to be a little fancy, mag-lev trains (not Hyperloop). That will actually solve the 'soul crushing traffic' problems that plague him and L.A. There's absolutely no way that any Loop system will come close to the passengers per hour that subways have been moving for over a century. Never.