Trains, Tesla and People Moving for Workday Commutes

[22522]

A lot of folks are wedded to mass transit for various reasons, maybe because they have seen it work well.

Tesla autonomous vehicles may present a better alternative, particularly in the suburbs where drop off points are distributed.

Can four person autonomous vehicles, perhaps Model 3s, perform the city, or factory, side drop offs with less cost to society than tracked mass transit solutions - trains?

 

[SwTslaGrl]

Perhaps with train/subway systems the tunnel (pipeline) is underutilized when loading/unloading passengers?

If the goal is to maximize the use of a an expensive tunnel perhaps many small side-stations loading pods will fill the pipeline?

 

[Lessmog]

Risking a thread derail by reporting a real one ...
Multiple fatalities reported after train derails in Washington state, spilling onto busy highway

Seems to be between Tacoma and Olympia. I hope noone here has a connection to the accident.

Multiple people were killed and others were injured Monday after an Amtrak train derailed and spilled onto a busy interstate in Washington state.

Those who died were believed to be on the train and not on the highway. Ed Troyer, a spokesman for the Pierce County Sheriff’s Department, said the highway will be closed for a long period, calling it a “pretty horrific” crash.

 

[BioSehnsucht]

Regarding the comparison of trains to the old telephone systems before packet switching...

If platooning can work for Semi, why not people-pods & etc in tunnels? Combined with packet switching and you can get the A to B efficiency of a train and also the C,D,E,F to A and B to W,X,Y,Z of dedicated/smaller transport. You end up with platoon "trains" that exist for as long as there are enough vehicles going in the same direction in the same tunnel at the same time to be called such.

Plus, loading and unloading can take as long as necessary (so if someone who can't move quickly, or is in a wheelchair, etc needs to board, it only holds up the one pod and not the entire system), with either single-pod elevator access or even terminal style access (just ensure the pods are far enough apart that their turning radius allows them to parallel park/unpark, or pull in to loading/unloading parking spaces perpendicular to traffic flow)

There might be some minor losses of efficiencies in some places (i.e., the larger the container you put people in, the better "payload mass fraction" you can get), but I think overall people movers as a packet switched network is a huge step up in overall efficiency.

 

[scaesare]

Regarding the comparison of trains to the old telephone systems before packet switching...

If platooning can work for Semi, why not people-pods & etc in tunnels? Combined with packet switching and you can get the A to B efficiency of a train and also the C,D,E,F to A and B to W,X,Y,Z of dedicated/smaller transport. You end up with platoon "trains" that exist for as long as there are enough vehicles going in the same direction in the same tunnel at the same time to be called such.

Plus, loading and unloading can take as long as necessary (so if someone who can't move quickly, or is in a wheelchair, etc needs to board, it only holds up the one pod and not the entire system), with either single-pod elevator access or even terminal style access (just ensure the pods are far enough apart that their turning radius allows them to parallel park/unpark, or pull in to loading/unloading parking spaces perpendicular to traffic flow)

There might be some minor losses of efficiencies in some places (i.e., the larger the container you put people in, the better "payload mass fraction" you can get), but I think overall people movers as a packet switched network is a huge step up in overall efficiency.

Yeah. The "packet switching" post was mine, and this is the sort model that one needs to consider.

There a number of pieces to this puzzle that will need to eventually be implemented... but the marriage of new modes of transport with high speed autonomous piloting and/or routing capabilities potentially turns "mass transit" on it's head in that it may eventually mean "transporting masses" rather than "transporting en masse".

 

[neroden]

OT

If FSD becomes reality, public train transport will not be able to compete.

This is, of course, ignorant nonsense, because of traffic congestion. I presume everyone who has lived in a big city has experienced an endless traffic jam on the way into downtown during rush hour, while the trains rushed past, on schedule, beating the cars.

The key thing to know about railways, economically speaking, is that they're all about volume. If you do telecommunications, the word to remember is throughput. Railways are completely economically hopeless for low-throughput applications, and unbeatable for high-throughput applications. I could go into detail regarding the physics first principles reasons why, and I have in the past. (This means railways are essentially urban, or urban-to-urban; the concept of the rural railway is non-viable. Although an urban railway could stop at the edge of the countryside -- this is called a "commuter railway" -- or an urban-to-urban railway could stop in the countryside en route, since it's cheap to do so.)

Apart from the obvious disadvantages, train transport is also unexpectedly energy inefficient.

Before you make incorrect assumptions, look at the age of the vehicles. Railway locomotives (and EMUs) last a startlingly long time, often in excess of 25 years, and the result is that very, very old rolling stock is still running. It looks like Belgium is still running 1980s locomotives. Older rolling stock didn't have regenerative braking; you'll mostly only find it on 2008 or later vintages. The efficiency obviously goes up, *a lot*, as they're replaced with rolling stock with regenerative braking, and it's ahead of cars, again.

Literally everything you can do to make an EV efficient, you can do on a train, plus more, and better. The energy efficiency per person, of course, depends on throughput, which gets me back to the point above.

 

[neroden]

OT

Mijn vlaams is a bit rusty but it looks like they applied a 2.9 factor to the electricity consumption to account for its production and transport losses (primair energieverbruik).

Oh geez, that's apples to oranges, then. It's already >2.5 times more efficient than a one-occupant Tesla, then.

 

[neroden]

OT

Re-reading the PDF, indeed, it lloks like they used a well-to-wheel factor of 2.9. Taking this into account, I still get to an astonishingly high number of 110Wh/km. Let’s say this is competitive with an EV with one passenger, but as soon as you get to 2 passengers per EV it’s no contest.

As noted previously... not really. 3, if you like, assuming that electric trains don't get more efficient, which they are getting.

However, after three decades of active campaigning starting in the 1970s, nobody has been able to get the average occupancy of cars up to 2. Anywhere. It's always between 1 and 2. So the trains are more efficient.

Background info: I looked this up because there are political parties in Belgium that want to expand public transport. Over the last few decades, a lot of train stations have been closed, and the frequency of trains has gone down (a lot). With the above data, if the frequency goes up without a corresponding increase in passengerkm (most likely, because it’s the unpopular timeslots that have been scrapped),

Well, now we're getting into interesting points related to throughput, which I can agree with you on.

Lowering the frequency of trains is generally wrong because there is a strong advantage to a "turn up and go" service (every 10 minutes or less), or a "takt" service (same time every hour), but it often makes sense to shut the trains down at night for maintenance, or to run extremely short trains off-peak.

As for closing stations... Belgium's kind of tiny, isn't it? How many stations did you have to start with? The country's only 170 miles across, and that includes some apparently-rural areas in the Southeast in the Ardennes. Many stations appear to be less than a mile apart; this is subway-station spacing. Are all of these areas fully urbanized? It looks like you had an insane number of stations before the closures started, and you still have a very large number of stations. The spacing of stations you have now would be considered a *substantial expansion in station coverage* over here in the US.

Here in the US we get idiotic suggestions to cut major commuter or urban subway lines, or to cut the railway lines between the centers of cities with million-plus populations. Which is why I tend to fight back against anti-rail propaganda. But rural railways -- i.e. where a two-lane road *doesn't* turn into a parking lot at rush hour -- are really non-viable. Looks like you may still have some in Belgium.

 

[bhtooefr]

Imagine the following...

Car takes dad to work and then returns to run mom to the same. It comes home to run child number 1 to school. It comes home to take child number 2 to the local college for classes. It returns home to plug itself in to charge up before it reverses the process to bring everyone home. 1 car payment instead of 4. 1 insurance payment instead of 4. 1 fuel bill instead of 4. The list goes on and on.

But that 1 fuel bill is 8x the cost - each of those drivers would have taken one round trip each, and the car is taking two round trips per passenger each instead.

 

[Dan Detweiler]

But that 1 fuel bill is 8x the cost - each of those drivers would have taken one round trip each, and the car is taking two round trips per passenger each instead.

8X the cost of 4 vehicles traveling the same distance instead of 1? I don't follow. 2X maybe, but that is far exceeded by the savings of insurance and car payments. Add to that if the one vehicle is electric and the others ICE? It's a no brainer.

Dan

 

[abasile]

The main challenge with this scenario is that the overall distance driven is double what it would have been. Indeed, worries have been expressed that autonomous vehicles may lead to more traffic congestion.

In urban/suburban areas, there's the likelihood that "transportation as a service", aka. fleets of autonomous taxis, will reduce car ownership. Then again, it may prove cheaper for a family to primarily utilize the one or two cars that they own, and just have those cars make extra roundtrips, thus raising the concern for congestion. I'm assuming that many families will prefer to own at least one vehicle, in order to guarantee immediate availability in all seasons and to ensure that they have the means to evacuate in the event of a disaster scenario, etc.

Still, even with some added congestion, I do agree that this scenario is probably more appealing to most Americans than mass transit.

Imagine the following...

Car takes dad to work and then returns to run mom to the same. It comes home to run child number 1 to school. It comes home to take child number 2 to the local college for classes. It returns home to plug itself in to charge up before it reverses the process to bring everyone home. 1 car payment instead of 4. 1 insurance payment instead of 4. 1 fuel bill instead of 4. The list goes on and on.

I think the average American will see this as a more appealing scenario than taking public transport of any kind. Of course add to this the ability to help pay for itself in the Tesla network when it is not being used by the family and it is another win/win.

Dan

 

[neroden]

From Market Action (there's probably a better thread than this, but screw it, and quoting two posts out of the discussion):

Apparently here --> Trains, Tesla and People Moving for Workday Commutes

 

[neroden]

The way I view is that public transport makes sense in very dense traffic situations, like peak hours between cities. Expanding the service (as proposed by some political parties) to a wider geographical coverage and a higher frequency creates a poorly utilized service, and robotaxis are a better solution to provide mobility to the entire population.

(Yeah, I am not moving my reply to a different thread.)

OK, the key thing I learned from studying the history of railways: it's all about throughput.

-- higher *frequency* improves your efficiency and financial performance, at least until you hit market saturation -- and in a dense urban area, that takes a long time. NY, London, and even Boston subways are crowded at all hours of the day, and replacing them with taxis would really not work -- the congestion would be awful at noon, not just at rush hour.
-- higher *geographical coverage* is usually a mistake; you lose the economies of scale. You have a whole new track, whole new stations, whole separate trains, etc.

Obviously some geographic coverage is valuable.

Case 1 for geographic coverage: if you're missing out a giant urban center (like, suppose there were no trains stopping in La Defense) then connecting it is probably a good idea.

Case 2 for geographic coverage: You want network connectivity, because there are network effects.

If you have

* city 1 -> suburb line ...short gap... suburb <- city 2 line
and you convert it to
* city 1 - suburb - short rural area - suburb - city 2 line
Then that's an improvement, because of network connectivity; you'll get a bunch of extra passengers and have practically the same costs.

But extending rail lines out into the countryside in general is a bad idea, especially if they dead-end in the countryside. Rural railways don't make sense, because the market isn't there for the throughput. Railways are all about very high throughput on a single line. That's where they are unbeatable.

I have repeatedly advocated for higher frequency rail, and against higher geographic coverage of rail (except where entire dense cities or dense neighborhoods are unserved). There is a strange lobby in part of the US which wants rail to the exurbs and is simultaneously unwilling to beef up downtown subways -- it's had a particularly malicious influence on Seattle -- and this lobby is, economically, incorrect.

So I'm going to distinguish between those two points: higher frequency vs. higher geographic coverage. I'm not sure what the political situation is in Belgium. But basically, if there's advocacy for higher frequency on the mainlines through the high-rise areas, it is correct. If there's advocacy for more stations in the Ardennes, it is incorrect.

 

[neroden]

Alternative Fuels Data Center

In US, transit buses operate at less than 25% capacity. This makes them less fuel efficient than private passenger autos on a passenger-mile per gallon of gas equivalent, 39.7 pmpGGE for cars vs 30.1 pmpGGE for transit buses. Demand response (includes taxis, Uber and paratransit) are much less efficient at 8.5 pmpGGE. It is unlikely that robotaxis will ever be as fuel efficient at private cars on a passenger mile basis because any sort of demand response vehicle must travel some fraction of miles without any passenger.

The rule of thumb for demand response is that over half their trips are passenger free. Half is considered good. In practice, except on a few specialized routes, you almost never manage to get significant trip-chaining, so it's out empty from the base, do trip with passenger, back empty to the base.

The primary current limitation on private cars in big city downtowns is, interestingly, parking. It's an unsolvable limitation; you can't build enough parking without destroying the city (songs about this include "Big Yellow Taxi" and "My City Was Gone"), and parking is very expensive in any case (usually heavily subsidized by taxpayers). This has led to the displacement of private cars by Uber-like services or taxis. The robocar fanatics suggest that the private cars will simply drive back to base, which makes them as inefficient as taxis...

 

[neroden]

Getting the right vehicle fit to meet specific demand is important. I think the tendency in public transit has been to error on the side of larger occupancy capacity than needed. This may be economically motivated by rationing the cost of the driver. Public transit folk ridiculed Musk for suggesting an autonomous transit vehicle for small occupancy. But if occupancy is less than 25% of capacity, something is really wrong with the size of the vehicle.

*Sigh* It's not. It's slightly on the low side, but not much. (Now, 10% of capacity, that would be bad.)

This is a complex topic, but the things you need to know to start with...

-- Average occupancy of private cars is 1.59 people per vehicle, and the vehicles typically have capacity of 4 or 5 (sometimes 7). If we assume a vehicle capacity of 4.5 (probably an underestimate), that's 35% occupancy. This is overstated because sometimes the driver really is just ferrying other people around (taking kids to school etc.)
-- Fixed-route transportation has a peak segment. If have near-zero occupancy at the ends of the route, and a full bus / train in the middle of the route, you have average occupancy... of 50%. This is fairly normal, and about as good as it gets. There's always an effort to design routes with "anchors" on the ends to bring the occupancy up, but you'll always have a peak segment.
-- The marginal costs of running larger vehicles are minimal. Yes, part of this is leveraging the cost of the driver, but also of the wheels, the drivetrain, the crumple zones, the fixed energy costs of operation (eg heating), etc. Think about it -- how much more does a stretch limo cost to operate vs. a regular limo? Not much.
-- The marginal costs of running smaller vehicles are severe. When you fill up your bus or train, you have to dispatch extra vehicles on short notice, which is sometimes impossible -- or turn away customers, which creates ill will and leaves customers with few alternatives (remembering that this happens at *peak travel* when cars are hopelessly stuck in congestion). Mass transportation fleets are sized for the peak, and they have to be.

I think smaller vehicles will become much more economical.

You are, to put it bluntly, wrong.

At most, it might make sense to run shorter trains more frequently (Vancouver SkyTrain) rather than longer trains less frequently (subway systems with drivers).

 

[neroden]

Yes, peak time capacity is important. If you replaced 100 40-passenger buses with 400 10-person vans, you'd have the same peak capacity.

Measure the roadspace before you say that. You're incorrect.

That said, buses are essentially ineffective as mass transportation during peak hours unless they have exclusive bus lanes -- and if they do, you are better off putting in trains. Buses are really only good for shuttling people to the railway stations, IMO.

 

[bhtooefr]

8X the cost of 4 vehicles traveling the same distance instead of 1? I don't follow. 2X maybe, but that is far exceeded by the savings of insurance and car payments. Add to that if the one vehicle is electric and the others ICE? It's a no brainer.

The one bill is 8x the cost of the 4 previous bills, so yes, that comes out to 2x (assuming of course that the one car replaced four cars of equal efficiency).

 

[bhtooefr]

That said, buses are essentially ineffective as mass transportation during peak hours unless they have exclusive bus lanes -- and if they do, you are better off putting in trains.

If you're thinking in terms of quarterly or annual budget, and have an existing pervasive road network, it can be cheaper to buy a bus fleet and use paint and signage to mark off bus lanes, than it is to deploy rail.

Also, Columbus, OH does something interesting - interstates along bus routes have the breakdown lane signed as a bus travel lane in addition to breakdown.

 

[NicoV]

This is, of course, ignorant nonsense, because of traffic congestion. I presume everyone who has lived in a big city has experienced an endless traffic jam on the way into downtown during rush hour, while the trains rushed past, on schedule, beating the cars.

This may be true in the USA, but not in Belgium. We only have couple big cities, and a large part of the population lives in between the cities.We do have a lot of traffic jams, but in general, most routes are far faster by car than by public transport even taking into account the traffic jams. The main problematic points are the ring roads around the 2 or 3 biggest cities, but during school holidays there are almost no traffic jams there. Lots of people are switching to electric bikes to avoid those traffic jams, of course, that’s the most energy efficient solution.

ooks like Belgium is still running 1980s locomotives. Older rolling stock didn't have regenerative braking; you'll mostly only find it on 2008 or later vintages. The efficiency obviously goes up, *a lot*, as they're replaced with rolling stock with regenerative braking, and it's ahead of cars, again.

No, we have modern locomotives with regen capability. The quoted energy consumption is not high because of inefficient equipment, but because of low occupation outside peak hours.

But rural railways -- i.e. where a two-lane road *doesn't* turn into a parking lot at rush hour -- are really non-viable. Looks like you may still have some in Belgium.

Actually most of the two lane roads are ok during rush hour. It depends on what you call rural. We have relatively small cities (few cities have high rise buildings) spaced closely together, like maybe 20km between city centers. The suburbs of those cities, almost reaching each other, but we call the space inbetween rural. Train transport is very good between those cities, and during peak hours very efficient, 5 times more efficient than a car according to their own statements. But that’s comparing an electric train (with regen!) to an ICE.
My point is that switching to EVs closes the gap for the rush hour situation, and is better energy wise for practically all other cases, with added benefits. While in the past individualised transport was economically not feasible due the driver and fuel cost, and people HAD to use public transport, that role will be overtaken by EV robotaxis.

 

[Buckminster]

There is no doubt that here in the US there is a certain stigma regarding the use of mass transit in many parts of the country. Especially with the love affair we have with suburban living. Unlike most of the rest of the world, as a whole, we just don't like using mass transport for whatever reason. Call it arrogance, call it independence, whatever. It is what it is. Where I think autonomy will have a huge impact here in the states is the ability to truly be a one car family instead of 2,3 or 4.

Imagine the following...

Car takes dad to work and then returns to run mom to the same. It comes home to run child number 1 to school. It comes home to take child number 2 to the local college for classes. It returns home to plug itself in to charge up before it reverses the process to bring everyone home. 1 car payment instead of 4. 1 insurance payment instead of 4. 1 fuel bill instead of 4. The list goes on and on.

I think the average American will see this as a more appealing scenario than taking public transport of any kind. Of course add to this the ability to help pay for itself in the Tesla network when it is not being used by the family and it is another win/win.

Dan

Having spent a bit of time in the US including staying at relative's homes - I think I can be confident in saying that the root cause is that you live in extremely low density housing areas commuting into enormous cities with many centres. The chances of being able to walk to public transport that takes you direct to your place of work (without waiting for changes of transport) is close to 0%. Walking to a Walmart from the store next to it involves walking a mile by a 4 lane freeway without a sidewalk. I take public transport (train into London and then the subway) regularly. I doubt I would in the US much and it has nothing to do with the stigma or quality. Robotaxis will potentially increase public transport across the world as it makes the changeover a great experience. No cost of parking or finding a space.