The end of the A380 and California HSR teach us smaller is better

Tags: 
It was just too big to succeed

I have written often about the new economies in transportation that future technology like robocars provide. In my research I've learned something that seems to not be well known in the transportation world -- that often, smaller is better and more energy efficient.

This gelled strongly this week with the end of both the Airbus A380 and the California High Speed Rail -- two giant transportation systems that violated this rule, and tried to be too big.

The normal wisdom is that the more mass your mass transit is, the more efficient it will be. Instead, while transportation does get more efficient as the vehicle size grows at first, as you make the vehicles bigger you create more burdens for the passengers, and they stop riding, and the efficiency and economic viability start going down, even though the physics says they should go up.

It leads to the surprising conclusion that the train and bus may be becoming obsolete, in part thanks to Robocars that eliminate the one big economy of large vehicles --needing only one driver.

I have written up a new version of this lesson, based on the A380 and CHSR in a new Forbes post, found at:

Smaller is better in transportation

Comments

The only thing I have that no one can replace is my time.

We used to call new appliances labor-saving devices. We need to think of transit as time-saving for its users. Yes, it needs to be fuel efficient etc, but to be successful it needs to optimize for peoples time. We will pay a (small) premium for that when it is available.

Mass transit can't optimize for your time, because the essence of mass transit is a group travel together, and a group rarely has exactly the same time agenda. To compensate for that, they give mass transit private right-of-way or other perks to help it compete with the car (or sometimes beat it.)

But it's the private right of way that is the thing you want, not the mass transit.

This is a much improved version of your 'Is green U.S. mass transit a big myth' article that I personally found very educational. In my experience you are exactly right about many environmentalists opposing cars and planes in favor of more rail services. The assumption that such mass transit is the environmentally friendly choice is often based on narrow assumptions that that fail badly when the bigger picture is considered. Efficiency numbers can be subjective (and confusing) depending on what factors are chosen to be included or excluded.

My only other comment is that energy efficiency is a rough proxy for environmental damage and fuel costs. Getting closer to first principles from an environmental POV would probably involve a measure of point to point CO2 emissions per person. Whether deliberate or not, your analysis is far closer to this measure that many of the arguments used to favor mass transit.

(Note; minor typo in your Forbes piece 'spends about $16B to for 13B passenger miles')

Robocars will probably help increase ridership on buses and (especially) trains. One of the biggest problems with using a train is that it's pretty much never door-to-door. There are some, or at least one, commuter train line that works - the Northeast Corridor line. But it's still a hassle. To commute to work you have to pay for parking at the train station, unless you happen to live in one of the (usually very expensive) homes or apartments within walking distance of the train station. Then at your destination you have to pay for some sort of public transportation, unless you happen to work at one of the (usually very prestigious) office buildings within walking distance of the train station. And public transportation may very mean an expensive taxicab, unless you are working in a place like Manhattan with an excellent subway system, or want to deal with the undoubtedly awful bus system.

Robotaxis might fix all of that. For a commute from Princeton to Manhattan, you can now forego the $450/year parking fee at the Princeton Station, and take a robotaxi straight to Princeton Junction. At the other end you might not change anything, the subway system is pretty good, but over time it'll make more sense for businesses to move from Manhattan to North Elizabeth, and you can take another robotaxi from the North Elizabeth station to your workplace there.

Moreover, now that you don't need your car to drive from your house to the train station, you might be able to get rid of it completely now. I suppose you could just take a robotaxi the whole way, but the monthly cost for a ticket between Princeton Junction and Manhattan is $451. That's a 59 mile drive, which includes a trip through the Lincoln Tunnel that costs $12.50 (peak). If you travel that 20 days a month that's $250 just in tolls, which leaves $200 for the 2,360 miles of monthly travel, or less than $0.01/mile. And that assumes you avoid the New Jersey Turnpike as much as possible, as you'd have to add another $13 round-trip in tolls to do that, and wouldn't have *anything* left over after tolls. Granted, maybe your one-person robocar will pay half as much in tolls as a regular car, but it's not going to be free (unless Chris Christie decides to do another "traffic experiment").

Now, granted, I picked what might be the only commuter rail route in the USA that is profitable to run. But I'd argue that one of the main reasons that commuter rail isn't profitable in most places is because of the high expense of getting to and from the train station, and robocars will greatly change that.

It is a common view among transit planners to view robocars as "the great last mile solution for transit." That is, as you might guess, a super myopic transit oriented view. It's like horse breeders seeing cars and thinking what a great way they will be to move horses around.

But yes, they certainly would make use of the train a lot better, and in fact, without them the train has a much harder time competing. But you still need to go to the train station and travel on the train's schedule and get from the drop-off point to the train platform and be there early enough not to miss it, unless the train is every 5 minutes.

What trains actually have that's great is not the train, it's the private track. If you have a dedicated right of way, you can offer superior service. The answer is to find ways to offer dedicated paths to more flexible vehicles like buses (That's what Bus Rapid Transit is) and vans and even carpools.

I actually started off with the thought of what if we got rid of trains completely, and used the rights of way for robovehicles. Then I realized that there is probably a lot of freight that won't be easily switched away from trains. So the question became, if we have to keep the tracks anyway, is there a reason not to use it for commuter traffic?

For lots of people commuting relatively long distances to a central location (which is what the commuter portion of the Northeast Corridor traffic is), being on a schedule isn't bad. Most people work on a schedule anyway.

I still think you're overestimating the ability of robotaxi cars and vans to serve the needs of all, or even most, commuters. Doing this would require lots of cars and vans that are used for commuting and pretty much nothing else. So you wouldn't get anywhere near the 50,000 miles a year per vehicle that you estimate, and you'd have to park the vehicles most of the day (close to where they're going to be used, if you want to minimize miles driven around with no passengers). There's also the issue of loading and unloading. On the marginal (marginal, not average) commuter vehicle you'd basically have all the expenses of personal ownership of a vehicle, plus all the overhead of running a taxi company. The trick, therefore, would be to make sure prices are high enough so that *everyone* doesn't switch over their commute, this way you can have fewer cars in the fleet that have higher miles per year, thereby both lowering expenses *and* increasing profits. (This, of course, is just another way to arrive at the common rule that *marginal* revenues equal *marginal* costs.)

I'm sure some people with a myopic transit view, but I don't think I'm one of them.

Freight doesn't run on commuter rail, subway and light rail lines. Those lines could be hybrid if you paved them with the rails embedded as is done for street cars. However, because trains block the track when they stop, there are complications in doing that.

Yes, commuter vans would get much less use outside of rush hour, but so do most commuter's cars that sit in parking lots all day. The vans would continue to provide service for small groups and offering low cost shared transport on longer trips during the off hours. Some would sit idle. You could use them to carry very small groups, and this is certainly wasteful of energy, so you create a model of the right balance of vehicles and what helps the most. During rush-hour, you can of course also serve smaller groups of commuters in 4 person vehicles, and some in 1-2 person vehicles. You only have to calculate how many people absolutely must share to get the capacity you need. Above that you get a balance -- having more vans improves the efficiency at rush hour, compared to sedans, but it means more vehicles sit idle or run less efficiently off peak. You find the optimum, or the market finds it.

The commuter van should still get in 8 hours of operation per day, as that's the two rush hours, and possibly more at the lunch peak. If they average 40mph (the national average, but they would do more if mostly on highways) that's over 300 miles/day for a van or almost 70,000 miles/year -- if they get full use for all of rush hour. They won't, but they can still end up with a lot of usage. If operating in pure city traffic it's less of course. Unless we start offering that private right of way to them, which we should, rather than having trains use it. If they have to go on congested city streets they only pull off around 100 miles/day, but with heavier wear. Of course, one big point of vans is that they are non-stop -- at least no passenger pick-up stops, they still have to stop for traffic -- and thus take less wear.

The average commuter van might get 8 hours of operation per day (though I'd argue that 4 hours is more realistic). The *marginal* commuter van won't. The marginal commuter van is the one you have to buy to handle the highest peak of rush hour. If the morning commute requires 1,000 vans from 6 to 7, 1,500 vans from 7 to 8, 2,000 vans from 8 to 9, and 1,000 vans from 9 to 10, the marginal van is the 2001st van.

The marginal van is the van that you have to charge customers for, unless you want to have shortages. If you don't charge for the marginal van, demand will outweigh supply.

In picture form: https://en.wikipedia.org/wiki/Marginal_cost#/media/File:Costcurve_-_Marginal_Cost_2.svg

Add new comment