I like Jarrett Walker. He does come at the problem from a certain perspective though -- he wants to solve problems with mass transit. Just like he talks about there being immutable laws of geometry, there are some immutable laws of transit. Mass transit fundamentally depends on having a lot of people who want to get from point A to point B, all at the same time. Since this is never true, you have to relax the requirements. You can make people wait until there are enough of them, you can make them walk to or from a stop, you can make them detour, slow down, stop and then speed up to let others get on and off, and you can make them switch vehicles. Fundamentally those are your options given some set of vehicles, whether they're planes, trains, or automobiles. All forms of transit exist on the tradeoff surface between these possibilities. As population density goes up, it's easier to find lots of trips that are close enough in the five dimensional space of origin, destination, time.
Where I perhaps differ from Jarrett is that I see no problem in having transit options that have one or two people per vehicle. The 2-5 person capacity mass transit vehicle has been a very underexplored space because someone has to own it and someone has to drive it. Services like Uberpool and Lyft Line are exploring this because smartphones make it possible, and this will only increase when you don't have to pay someone who isn't riding to operate the vehicle.
America use to have more rail at one point in time than Europe has today. It wasn't that long ago. We're talking less than 70 years.
If we start building infrastructure in our cities again for rail and express buses, you will see things being to naturally be more compact. Business spring up around rail stations, and people who want to live close to one another fill in those spaces and move closer to the city.
The "America is too spread out" argument is total bullshit. At the very least, make the cities have good transport and walkable areas. It can happen in a very short time frame. We lost it in a short time frame. Northern Indiana has the same population density of Scotland, so it can work if we build it. We use to have trains from Nashville to Atlanta. The famous "Chattanooga Choo-choo" doesn't even have an AmTrak anymore.
Transportation is the largest factor in the ability of people to move up from poverty. This isn't some subjective statement. There are tons of objective studies to support this[1].
The idea that you need to own a car (and insurance an maintenance) in most of America is pretty insane. A car shouldn't ever be a necessity. Associating it with freedom, as the American car companies have, and buses with poverty, is ass fucking backwards.
> All told, these dozen mega-regions span 243 metropolitan areas in the U.S. and Canada, more than six in ten of all U.S. metros. They have a combined population of more than 230 million people, including 215 million from the United States or 70 percent of the U.S. population. Together, they produce more than $13 trillion dollars in economic output, equivalent to three-quarters of America’s total GDP.
Vast regions of the US don't even matter for the purpose of this discussion. When you only look at those regions the problem is no different than in a lot of other countries. That is also the geographical level where public transport makes sense, it's per-region (and different regions may want to/have to solve it in very different ways since their respective situations are different), not one solution for the whole country.
> We use[d] to have trains from Nashville to Atlanta. The famous "Chattanooga Choo-choo" doesn't even have an AmTrak anymore.
Why do you think that is? Because nobody bought tickets from Nashville to Atlanta. Do you think that route was throwing off millions a year in profit to AmTrak and some executive sat back and said "No, I hate poor people let's get rid of that route."? Of course not.
If people bought the tickets but the business was mismanaged, that's one thing. But not enough people bought the tickets.
Also, when you get to Atlanta, the train station isn't where you want to be. The train station is not even at a subway station, despite Atlanta having a subway that's decent if you happen to want to go to one of the few places it serves.
(I'm talking about the station that currently serves the only intercity rail in Atlanta, which runs from New York to New Orleans. I am aware that perhaps the former stations downtown were better-located.)
pc86 you are wrong. I have bought tickets on Amtrak, and Amtrak sucks. I have done the price/time opportunity cost calculation for several destinations, and Amtrak came out a big loser, and I ended up driving. I have taken trains in Europe many times, and the overall difference is astounding. I don't know how to fix it, but American trains just seem to want to fail.
Amtrak is a bit of a special beast. It's a Frankenstein's monster that the government assembled from the corpses of the many independent rail companies that were failing in the mid 20th century.
As for why Amtrak's service is so poor, my personal sense is that Amtrak made a bunch of bad decisions that maximized their bottom line, such as using excessively cushy (and heavy, and low-capacity) rolling stock for the kinds of routes that would be covered by something more like IC service in Europe. In essence, they made a gamble that what rail travelers want is luxury, which translated into 250lb seats with half an acre of leg room and 3" of foam padding, when what they really want is frequent, regular service.
Commuter train systems like Chicago's Metra are much better, but they're unable to extend themselves out to full-fledged intercity service, no matter how successful they might be at it, because Amtrak enjoys a statutory monopoly.
Amtrak is something the government assembled from the corpses of the passenger operations of many independent rail companies. Those passenger operations were (all of them) losing money, even if the rail companies themselves were not failing (some were, some weren't).
> America use to have more rail at one point in time than Europe has today. It wasn't that long ago. We're talking less than 70 years.
The US has a larger rail network than the EU, still. Actually, the US's freight rail system is the envy of Europe, although our passenger rail sucks.
> The "America is too spread out" argument is total bullshit. At the very least, make the cities have good transport and walkable areas. It can happen in a very short time frame. We lost it in a short time frame. Northern Indiana has the same population density of Scotland, so it can work if we build it. We use to have trains from Nashville to Atlanta. The famous "Chattanooga Choo-choo" doesn't even have an AmTrak anymore.
No, it's not bullshit. It is not profitable to have a train route of more than about 5 hours (China is sort of the exception that proves the rule, as their air network is incredibly constrained which eliminates alternatives to trains). That is roughly the straight-line distance between Chicago and NYC, which (as I've pointed out before) is longer than the longest axes of most European countries. And Chicago-NYC is one of the shorter links between major city pairs in the US.
It's also funny that you compare to Scotland, since Scotland has some of the worst availability of rail in Europe.
> Mass transit fundamentally depends on having a lot of people who want to get from point A to point B, all at the same time. Since this is never true, you have to relax the requirements.
There is also induced demand. If you make it very easy to get from point a to point b. Lots of people will want to live at a and work at b or vice versa. Right now we focus most of our energy and resources on making it as easy as possible for everyone to drive their car, so that's what they do.
He criticizes people arguing for smaller cars closer together as being a linear solution to an exponential problem, but mass transit is also a linear solution to the same problem. Build better mass transit, and houses near the stations become more attractive and people move there, which increases usage, and so on. And looking at it from the other end, ever smaller cars jammed ever closer together at the limit tends to the same density as mass transit. (We're nowhere near that now, but I can see us getting much, much closer in an automated driving future.)
The key differences end up being routing flexibility, packet size and latency. Mass transit is worse in all three (less flexible and more waiting time are given with mass transit; I'm taking packet size to be a measurement forced space sharing, so big is bad). Its only real advantage is being at the limit for density, a density cars will likely never be able to match. But because it's still linear, it's only a constant factor improvement.
(Personally, I travel around London by motorcycle and scooter because public transport is ludicrously inefficient; it multiplies journey time by 2x to 5x. Ironically, it works best if you're elite, i.e. live inside the circle line; but then you might as well cycle.)
It takes me 35 minutes to take the train from Fife to central Edinburgh - absolutely no chance of doing it by car in that time in the morning. If I had to drive into the center of Edinburgh in the morning I'd probably give it well over an hour and that would ignore the hassle and costs of parking once I get here.
Edit: And with the train I can sit and sip coffee and try and spot seals and herons and enjoy the splendid views....
Fife isn't city living; that's not relevant to the article.
Mind you, with your self-driving vehicle, you'll be able to sip coffee and spot herons. And I believe packetized self-driving vehicles grouping together into convoys ought to be able to exceed 200mph in the future. That future isn't today, but the article is explicitly forward looking, about geometry that doesn't change.
I think the article is ideologically blind to how mass transit has similar supply and demand problems (e.g. when I take the train home, I often have to skip multiple departures due to overcrowding), and it drastically underestimates the potential for technology change.
Stopping distances aren't about to magically improve in the next decade such that any vehicle can speed down the street at 200mph (or anywhere close to that) safely.
This whole scheme your dreaming up is just not realistic, the faster you go, the more dangerous a road is for everything else on it, from pedestrians and bikers to other cars sharing the road with you.
Making your car drive itself won't magically double road capacity either, as you'd need much more traction & better brakes to support tighter packing of cars in any vaguely safe manner.
> Stopping distances aren't about to magically improve in the next decade such that any vehicle can speed down the street at 200mph (or anywhere close to that) safely.
What makes you think the vehicles wouldn't be coupled together mechanically?
Pedestrians wouldn't be on a road going 200mph; I'd expect that kind of road would be fenced off like rail is today, or in a tunnel, and perhaps even be on tracks.
The point is that automation will change the nature of transport in ways that don't fit into the current mass transit / individual vehicle binary that we have today. That binary is based on limitations on current technology.
It's only a constant factor improvement, but it is a massive constant factor improvement. Mass transit easily scales up to the limits of our building construction abilities. You can only make buildings so tall before you start getting diminishing returns on density because you need to have more and more space taken up with elevators, HVAC, plumbing, etc.
Even in an incredibly dense city like Paris, they aren't running out of space to put subway lines. They are limited by by the number of people who need to get places and by the cost of building more subway lines.
Paris doesn't have enough subway lines to handle all transportation needs for everyone, but that's not because they ran out of space to put the subway lines in.
It's likely because they have covered enough people's transportation needs that it wouldn't be cost effective to put in more lines.
Incorrect. New York has lots of people trying to get from point A to point B pretty much all the time, and other American systems with half-decent mass transit (Boston, DC) also see those systems get heavy use, to the point that shutdown of a line is a major disruption. Which is also the universal experience in Europe; Paris has added several mass transit lines in recent years, and has had no problems keeping them full.
Jarrett doesn't have a problem with small-vehicle options either. He just says: In a dense big city that has to be a sideshow and therefore discussing it is a harmful distraction.
(I've seen small-capacity "mass" transit, btw. Never two-person-max. But I've seen ten-person microbuses in several cities. Where I live there are a few that feed minor outlying hubs.)
Can we say Jarrett is visually impaired and can't drive? Let's also say Jarrett is a school teacher and earns less than $40k/year.
Jarrett has very limited options now. This person can live in one of the few (like three? four if you stretch?) cities in the US with decent enough transport to get around, that aren't insanely expensive to live in ... or just have to get lucky and find a house in walking distance of the school and a grocery store and a bus stop to the city.
Jarrett isn't going to be able to afford a Lyft/Uber and is very limited on their ability to even be social.
"has to be a minority option" is an arithmetic necessity of density: you can't have everyone owning cars in a high-density city because it's physically impossible to fit them on the roads.
I suppose Jarrett would say that if a place is dense enough to have more than occasional traffic jams, then it already is too dense that its transport problems can be solved by road construction.
There's way bigger problems for 'the 2-5 person capcity mass transit vehicle' (by which you basically mean 'car') than people owning and driving the vehicle.
Time is completely out of your hands (am I first or fifth in? What route will it take? How long are we going to wait for this person to show up?). Its not solving infrastructure issues massively (cars still need roads - you might make it attractive by having a lane for 3 or more people vechiles though). Its not really a system suited for dense walkable centres (see European cities) - and we are going further that way. And whats the objective benefit vs a train or tram? You are still in a space with strangers.
There are places in the world where labour costs are so low that this doesnt figure. There are also those where shared rides are common (e.g. taxis that run 'routes') - I'm not sure in any of these places they are considered a better solution than public transport at all. Of course they have a place, and I think Walker acknowledges this too, but its not really a game changer like building proper public transport can be.
Re "the 2-5 person capacity mass transit vehicle":
Lots of cities have enough density to support good dense tram / train / bus networks in the centre, during the day. But quite a lot of the people who would use that, also have to go further out sometimes, or need to get home later at night sometimes. If mass transit can't do this too, then they need to make a whole different plan... often meaning a car, and a house further out to afford parking.
This seems like the opening for smaller vehicles. They can perhaps fill in the gaps between lines, and the long waits between night busses, in a way that couldn't be done before, essentially because human taxi drivers are too expensive. It's possible this could help make proper public transport viable in lots more places than it is now.
In a city with great public transport, lots of the person-miles are done on full trains, running back to back all rush hour, in a way that nothing else can really replace.
But for the system to be useful, it has to maintain reasonable frequencies until midnight, out to a lot of points. This results in a lot of vehicle-miles being done almost empty. But they are really essential.
It's the second class which maybe small automated vehicles could replace. Lots of cities only cover the first class, and wonder why so many rush-hour commuters don't use it...
I often think about smaller vehicles. Would it be possible and better to have a rail system with lots of small pods slinging around instead of bus-sized containers?
Each station would need to allow for the pods to pull over without blocking the flow of continuing traffic.
Musk says a Model X derived minibus has "surprisingly high" people density.[1] The standard Model X seats 7, so I'd say at least 10.
I'm assuming similar volume-stretching tricks as the Model 3, (pushing the dash forward, glass roof for headroom) in a microbus package.
> All forms of transit exist on the tradeoff surface between these possibilities. As population density goes up, it's easier to find lots of trips that are close enough in the five dimensional space of origin, destination, time.
I think the real value of "big data" (something Jarrett dismisses as "big detail" in the article) is that we can plot every trip in that 5D space, deriving a "typical" density curve for any given day of the week. With that, the fleet management software can optimally allocate vehicles.
Clustering nearby points (trips) in this 5D space should optimally combine trips to minimize detours to individual houses (by grouping in the origin/origin and destination/destination 4-planes), combine trips across town (by grouping in the origin/destination 4-plane), and dynamically adjust the wait time to combine trips to/from low traffic locations (by grouping in the time/origin and time/destination 3-planes). Really it should trade off between all of these factors.
The challenge, of course, is that you don't know everyone's route ahead of time. Good prediction is important.
Essentially you're dynamically adjusting the bus schedule and route in response to spatial and temporal demand patterns. And with 10 passengers and no aisle[2], the maximum passenger density is similar to a bus (the "geometry" problem) So it seems like this can be more efficient than even a fixed bus schedule.
It seems like a decently large (100k+) multi-agent whole city transit simulation could settle the matter one way or the other. Who's plugging autonomous taxi routing into these?
This[3][4] looks promising. Accounts for induced demand. More publications on his website.[5]
[2] "Traffic congestion would improve due to increased passenger areal density by eliminating the center aisle and putting seats where there are currently entryways, and matching acceleration and braking to other vehicles, thus avoiding the inertial impedance to smooth traffic flow of traditional heavy buses." https://www.tesla.com/blog/master-plan-part-deux
I'm reminded of how the Chaos Congress organizers persuaded the Leipzig public transportation agency to run around-the-clock service to the congress: He showed them the curve of devices logged into wifi for the last year. It has a steep decline somewhere at night, which is when people leave.
With all the location data Google & Co have they should try their hand at making plans for public transport systems (if they haven't already).
I wrote about this a while back. Self driving cars are not a solution, at least in America where they won't solve the last leg problem from trains to home (because we have so few trains or even express buses).
Self driving vehicles on dedicated road ways can't even expect to match a tiny percentage of what trains or even express buses can deliver. Singapore already has autonomous fully automated trains. The same tech has been sold to Kuala Lumpar. London has several driver-less trains with more of their underground being switched over to semi/full-autonomous as well.
We've already had two articles on HN about NYC's crumbling rail infrastructure (both the subway and Penn Station). Chicago hasn't expanded their lines in decades. Our two biggest European style rail cities are not expanding, and are actively falling apart.
There is hope though. Seattle is expanding with ST3 and Florida's high speed line just went operational, with additional stops opening over the next few years which will eventually connect Miami with Orlando. Now if California doesn't fuck things up, we might finally have two real high speed lines in this country (and no, I'm not counting the Acela).
You can add Texas to that list. They're aggressively working on building high-speed rail. They will beat California to it and do it at a far cheaper cost. Once they build the first line from Dallas to Houston, I think it's extremely likely they'll continue building more lines between other major cities. You do Dallas to Austin & San Antonio. Then you look at Dallas to Oklahoma City. They'll show it works, that it's financially beneficial, and that it can be done within a reasonable cost.
If Texas can't build a HSR on flat terrain at a cheaper cost than a HSR that must go over/through/around mountains, above valleys and rivers, around world-famous national parks and forests, and is twice as long...quite frankly, they're doing something wrong.
The CA HSR authority has fucked many things up, but you're comparing apples to oranges here.
The outskirts of Austin are much emptier and flatter than what they're building rail through in California. Both of those facts make it magnitudes easier to acquire a dead straight & flat right of way, which is what high speed rail needs!
Seattle's light rail isn't bad, but it's expanding slowly and has a long way to go before it even comes close to matching NYC's breadth of transit service, let alone someplace like Tokyo.
Investment in the NYC lines is continuous, and although much more could be done to upgrade switching infrastructure and renovate some aging stations and train cars, it's still a remarkable achievement.
Seattle's light rail should just about triple over the next 6 years, adding 42 miles of new track, from Northgate to Kent to Bellevue and Redmond.
We also maintain our rail, unlike NYC. Chronic underfunding of rail maintenance (in favor of unneeded station upgrades) is tanking reliability on light rail lines in NYC and DC currently, and its only going to get worse so long as critical maintenance is delayed.
>And so this is why it’s tricky and this is why in your example of a bridge, if you widened the bridge but lots of people want to cross it, you’ll end up with a wider bridge that’s exactly as congested as it is now.
Widening the bridge does increase throughput; people who used to be unable to use the bridge are now doing so. There's a benefit there, it's just siphoned off from "reducing congestion" to "people are no longer forced to stay home rather than deal with congestion".
Wenover productions does a great video on how increasing traffic lanes makes traffic worse in every city. People who normally wouldn't drive, will now drive thinking they'll be less congestion.
Atlanta had some major closures recently and the city migrated around the problem, people took different paths to work and more people decided to get on public transport.
The fact is car lanes can never reach a fraction of the capacity of a train running at regular intervals. Even if you had nothing but self driving cars running bumper to bumper at 120kph and all filled to a capacity of 4.2 people, on a four lane road, you wouldn't even approach 8% of the capacity of a single rail line running at 2 minute intervals! Singapore's trains are automated. They're already self driving and run at <1.5min intervals during peak hours.
There’s something wrong with your numbers. 120kph for two minutes is 1km, or 250 4-meter cars, bumper to bumper. Four lanes, 4.2 occupants, that’s 16,800 passengers per two minutes, which is a full order of magnitude more than a 12-carriage train packed with standing passengers[1] - and the people in the cars are all sitting down.
People don't generally give up and stay home, they find ways to go about their lives. Driving to downtown San Francisco is awful and congested, so people simply take the train (because it exists in a useful capacity).
Widening that bridge has an opportunity cost. LA ramped up road infrastructure at the expense of useful public transit and urban densification, and it's a stressful and sickening experience to get around there.
leisure trips account for a small fraction of the kind of trip we're talking about
they APTA's study shows something like 15% of public transit trips are social or shopping/dining related[1]. i suspect the proportion would be similar (or lower) wrt road traffic.
His point is that if you make going by car more attractive by simply widening the bridge for everyone, people will switch from public transport to private cars. So his solution would be more "make a separate bus line on the bridge, so that they are more attractive to use than a private car".
Inducing demand is not just from mode-shift (towards cars in this case), but also from development. Take the example of a big city on a big island (e.g. Montreal), a larger bridge may encourage sprawl off-island rather than infill development on-island.
A choke-point like a bridge isn't just a sort of barrier (geographic, social economic...), but also an opportunity to manage travel patterns and thus development.
Facilitating more trips absolutely should be the goal of any transit system. Getting yourself from point A to point B provides you value - else why would you go? "More people sitting in congestion" just means that the facilitated trips are limited by capacity and density rather than geometric distance.
Yet having more people sat in congestion is going to have a negative affect on societies overall wellbeing. The problem is we look at it from the perspective of the immediate economic benefit and ignore the costs of poorer health outcomes and increased environmental degredation.
On further thought overnight, I agree with you more. There's a combined goal between zoning and transit to meet human needs, and obviating the need to make a trip on the zoning end is just as good as facilitating a trip on the transit end.
The overall systems approach is better in total, too - there are tons of broad cross-cutting concerns here, and transit is one piece of a larger and more complex system.
One point that doesn't come up os that serious operational inefficiencies and "friction" exist in how current fixed route services are delivered in most cities. Some due to political challenges (see WMATA funding woes). Some are due to older technologies that lead to lower capacities (see MTA signals), or sluggish transit agencies that have networks that have not kept up with changes to the city (see Dublin Bus).
Public transport needs to work better for the public too.
Ah, Vantucky just north of Portland, OR has the same BS "concerns". Ditto for the rich part of Bellevue, they've blocked Eastlink Light Rail from using any of the existing rail that is already government owned, cause "oh dear it'll bring the baddies!". They're still pissed off by the rails to trails project though...
Where I perhaps differ from Jarrett is that I see no problem in having transit options that have one or two people per vehicle. The 2-5 person capacity mass transit vehicle has been a very underexplored space because someone has to own it and someone has to drive it. Services like Uberpool and Lyft Line are exploring this because smartphones make it possible, and this will only increase when you don't have to pay someone who isn't riding to operate the vehicle.