Archive for the ‘Congestion’ Category

I was grooving on this almost Ed Ruscha-style illustration (”27 Onramp Configurations”?) in a new paper from David Levinson and Lei Zhang, “Ramp Metering and Freeway Bottleneck Capacity,” in Transportation Research: A Policy and Practice 44(4), May 2010, Pages 218-235.

The findings were sanguine on ramp metering:

Traffic flow characteristics at twenty-seven active freeway bottlenecks in the Twin Cities are studied for seven weeks without ramp metering and seven weeks with ramp metering. A series of hypotheses regarding the relationship between ramp metering and the capacity of active bottlenecks are developed and tested against empirical traffic data. The results demonstrate with strong evidence that ramp metering can increase bottleneck capacity. It achieves that by:

(3) and increasing queue discharge flow rates after breakdown – the average queue discharge flow rate is 3 percent higher with metering than without (with a 3% standard deviation).
Therefore, ramp meters can reduce freeway delays through not only increased capacity at segments upstream of bottlenecks (type I capacity increase), but also increased capacity at bottlenecks themselves (type II capacity increase). Previously, ramp metering is considered to be effective only when freeway traffic is successfully restricted in uncongested states. The existence of type II capacity increase suggests there are benefits to meter entrance ramps even after breakdown has occurred. This study focuses on the impacts of ramp metering on freeway bottleneck capacity. The causes of such impacts should be more thoroughly examined by future studies, so that the findings can provide more guidance to the development of ramp control strategies. It should also be noted that both types of capacity increases on the freeway mainline are at the expense of degraded conditions at the on-ramps and possibly arterial network. Therefore, without more comprehensive system-wide analysis, the findings of this paper, though in favor of ramp metering, do not necessarily justify its deployment.

The new INRIX scorecard is out — time for all you sabremetricians of the traffic world to drool — and reveals that congestion is creeping back to pre-recession levels. One can, I suppose, take this is another form of economic indicator — like same-store sales or new housing starts — as more jobs and more activity equals more miles driven. Then again, some of the congestion is directly a result of the economic crisis:

“Stimulus spending on road projects nationwide is starting to have an impact on congestion, particularly in off-peak periods. Delays across the country during off-peak periods – mid-days, evenings, overnights and weekends – were up 25 percent. Of the nation’s biggest new work zone slowdowns in late 2009, more than half were directly tied to stimulus projects.

There was the usual mix of interesting data points (hello Philadelphia and welcome to the Top 10 most congested metros!; Friday between 5 and 6 p.m. remained the worst time to be on the road in America), but one particular bit in the section on long-haul freight traffic caught my eye in particular:

INRIX data highlights that the nation’s truck freight network is highly interconnected, with some of its most important links located in places that aren’t immediately obvious (except to fleets and people traveling those roads). Nationwide, less than 5% of road miles have 3 times or more the average density of freight data, and less than 1% of road miles have 4 times or more. Of the most intensely used 1000 miles, California has the most miles of any state (271), closely followed by Arkansas (228); and I-40 has the most miles of any road (314).

I was surprised to see Arkansas pop up as number two in this category — I would have expected the Chicago region or some such — and I couldn’t help wondering, as one always does when one thinks of Arkansas, if there was a ‘Wal-Mart’ effect here? But the simpler explanation is that Arkansas itself is a trucking hub, home to a number of the country’s largest haulers and, it turns out, the state with the highest percentage of private-sector work force employed in the trucking industry. I’m sure there’s a bevy of interesting geographical/logistical reasons of why that came to be (e.g., proximity to rail hubs, the distribution center of Memphis, etc.), but in any case, it’s just one of the interesting tales lurking in the INRIX data.

Via the Moscow Times:

An enormous television screen showing a pornographic film caused a midnight traffic jam along Moscow’s Garden Ring Road as stunned motorists slammed on the brakes to gawk at the writhing naked bodies.

Authorities said they are investigating the incident, which lasted about 20 minutes. “Within three minutes we found it out, and within 15 minutes the screen was shut off,” said the deputy head of Moscow’s advertising committee, Alexander Menchuk, Interfax reported.

I can’t help be reminded of one of James Howard Kunstler’s favorite words: Clusterfuck.

A useful addition to the urbanist lexicon.

Another delight from Copenhagenize.

Perhaps one could, with Lego Mindstorms, create an electronic congestion charging cordon?

Here’s Peter Gorrie on the subject, and Bern Grush’s response.

Daniel Pink points me to an interesting new study via NBER: “Traffic Congestion and Infant Health: Evidence from E-ZPass,” by Janet Currie and Reed Walker.

The abstract states:

This paper provides evidence of the significant negative health externalities of traffic congestion. We exploit the introduction of electronic toll collection, or E-ZPass, which greatly reduced traffic congestion and emissions from motor vehicles in the vicinity of highway toll plazas. Specifically, we compare infants born to mothers living near toll plazas to infants born to mothers living near busy roadways but away from toll plazas with the idea that mothers living away from toll plazas did not experience significant reductions in local traffic congestion. We also examine differences in the health of infants born to the same mother, but who differ in terms of whether or not they were “exposed” to E-ZPass. We find that reductions in traffic congestion generated by E-ZPass reduced the incidence of prematurity and low birth weight among mothers within 2km of a toll plaza by 10.8% and 11.8% respectively. Estimates from mother fixed effects models are very similar. There were no immediate changes in the characteristics of mothers or in housing prices in the vicinity of toll plazas that could explain these changes, and the results are robust to many changes in specification. The results suggest that traffic congestion is a significant contributor to poor health in affected infants. Estimates of the costs of traffic congestion should account for these important health externalities.

I’ve not read the paper yet (if anyone has a PDF I’d love to see), but one interesting question is whether this is longitudinal as well — were the rates tracked before and after the introduction of EZ-Pass? And would this vary depending upon the number of lanes that actually offer EZ-Pass (roughly half at most NYC-area toll plazas). A provocative thesis in any case, coming on the heels of David Owens’ interesting piece in the WSJ.

Congestion isn’t an environmental problem; it’s a driving problem. If reducing it merely makes life easier for those who drive, then the improved traffic flow can actually increase the environmental damage done by cars, by raising overall traffic volume, encouraging sprawl and long car commutes. A popular effort to curb rush-hour congestion, freeway entrance ramp meters, is commonly seen as good for the environment. But they significantly decrease peak-period travel times—by 10% in Atlanta and 22% in Houston, according to studies in those cities—and lead to increases in overall vehicle volume. In Minnesota, ramp metering increased overall traffic volume by 9% and peak volume by 14%. The increase in traffic volume was accompanied by a corresponding increase in fuel consumption of 5.5 million gallons.

One thing I’d be curious to know about the papers Owens’ cites is whether the introduction of ramp metering simply brought more vehicles back to the metered-facility, and away from other roads they may have been traveling on (perhaps those were covered in the “overall vehicle volume,” but it typically seems smaller roads are not as well measured in those terms compared to highways).

At a conference this weekend, a Disney logistics guy told me that the number of buses Disney operates to ferry visitors around the Magic Kingdom would, if it were a municipal system, make it the 21st largest in the U.S. (Not to mention those other 20 cities don’t have monorails).

Via Engadget.

My latest Slate column considers transportation from an iPhone-centric point of view, with an eye toward ways apps might change the experience for the better. I’d be curious to hear what I left out (I omitted some things for space) or things that are in the works, or apps you’d like to see, etc.

Given that I’m always talking about how traffic can skew our sense of time and perception, I was fascinated by a recent article in the New Scientist that was interested in a simple question: Do the moving walkways at airports actually move people any faster?

This may save energy, but even under ideal conditions of no congestion and no baggage a walkway only makes a small difference in travel time - about 11 seconds for a 100-metre stretch.

Now, granted, this is only a model. But as someone who spends a lot of time in airports, and loves the idea of moving walkways but not often the reality (more on that in a sec), I feel as if there’s something to this. And trying to save travel time at the airport can be a futile, as with traffic: You may blaze down the moving walkway, only to be caught up in a bottleneck at security or the exit doors. And then there’s the reason I so often don’t get on in the first place: I don’t want to have to barge past the people who are simply standing on the walkway, actually going more slowly than normal walking speed (and there’s always a little hiccup of people getting off and on). This is the escalator problem: The technology was designed to move more people more quickly, by augmenting their normal motion, not simply ferrying passive passengers.

But the model above actually has an empirical counterpart, notes the magazine.

“Moving walkways are the only form of transportation that actually slow people down,” says Young.

The blogosphere has gone nuts over the idea that Google Mobile Maps will be showing real-time traffic maps, with information generated by the very people (at least those with Android phones, and a few other devices) navigating that traffic (yes, this is what Dash does/did, on a smaller scale). We can crowdsource our way out of congestion!

As Wired notes, “The new service takes this data from everyone and combines it, using Google’s big brain, to give a pretty accurate picture of traffic conditions, which are then piped back to your device.”

Question is just how “pretty” is pretty. As Roadguy notes, the reality looks better on screen than on the street. Google’s Big Brain is still developing, apparently.

Michael Frumin was intrigued by a report on 2008 subway passenger counts.

Just to get warmed up, chew on this — from 8:00AM to 8:59 AM on an average Fall day in 2007 the NYC Subway carried 388,802 passengers into the CBD on 370 trains over 22 tracks. In other words, a train carrying 1,050 people crossed into the CBD every 6 seconds. Breathtaking if you ask me.

So he began wondering what New York City would have to look like without that subway capacity — or, say, if every New Yorker decided to drive where they were going.

At best, it would take 167 inbound lanes, or 84 copies of the Queens Midtown Tunnel, to carry what the NYC Subway carries over 22 inbound tracks through 12 tunnels and 2 (partial) bridges. At worst, 200 new copies of 5th Avenue. Somewhere in the middle would be 67 West Side Highways or 76 Brooklyn Bridges. And this neglects the Long Island Railroad, Metro North, NJ Transit, and PATH systems entirely.

And that’s not all of it.

Of course, at 325 square feet per parking space, all these cars would need over 3.8 square miles of space to park, about 3 times the size of Central Park. At that point, who would want to go to Manhattan anyway?

Reading Frumin’s post, I was reminded of the early, Utopian visions, as sketched by people like Bauhaus stalwart Ludwig Hilberseimer, of cities “built for the motor age,” which would seamlessly blend great agglomerations of people with smooth, huge highway networks that always seemed to be largely empty, as in the image above. What these plans never acknowledged is the point raised by Frumin: The actual infrastructure required to move all those people by car to their massive towers, not to mention such questions as what they would all do once they got out of their cars (if they even desired such a thing), where they would park, etc. etc.

On the last point, Norman Bel Geddes, writing in the seminal text Magic Motorways, thought parking provided an easy answer to the congestion question:

There is one method, however, which does point the way to a future solution. It is the construction of parking space directly underneath or actually inside of heavily frequented buildings. The newest building unit in New York’s Rockefeller Center, for example, is provided with six floors in which over 800 cars can find parking space by means of ramps. The same idea has been incorporated, even more dramatically, into Chicago’s Pure Oil Building, in which the interior spaces of thirteen floors are reserved for tenants’ cars 300 of them.

How providing more supply would lead to long-term solutions to the congestion problem, particularly as all those drivers poured out of their massive garages at 5 p.m., was a question the modernist visions were never able to answer.

Of course, Hilberseimer’s early visions were admittedly a bit dystopian, as even an automobile city proponent like Le Corbusier was moved to note:

A wretched kind of “modernism” this! The pedestrians in the air, the vehicles hogging the ground! It looks very clever: we shall all have a super time up on those catwalks. But those “R.U.R.” pedestrians will soon be living in “Metropolis,” becoming more depressed, more depraved, until one day they will blow up the catwalks, and the buildings, and the machines, and everything. This is a picture of anti-reason itself, of error, of thoughtlessness. Madness.

And while the city pictured at the start of the post never materialized, that modernist dream of the (non-congested) automotive city never died, and its DNA carried on through GM’s “Futurama,” on through fantastic visions like Geoffrey Jellicoe’s “Motopia,” (pictured above, with its rooftop roads) through more serious (and taken seriously) tracts like Colin Buchanan’s “Traffic in Towns,” and into built places like Cumbernauld.

“Kill the street,” Le Corbusier once intoned, the old “donkey paths.” The new cities would do away, as the historian Stephen Marshall puts it in his excellent book Streets and Patterns, with things like the pub on the corner. “There would be no pub on the corner, since no building would interfere with the requisite junction visibility requirements. There would be no crossroads, since these would be banned on traffic flow and safety principles. Indeed, there would be no ’streets’: Just a series of pedestrian decks and flyovers.”

And as the following video (sent to me by Eric Boerer at Bike Pittsburgh) from Pittsburgh, circa 1955 shows, the modernist dreams had some serious propagandistic muscle behind it; the irony of this video (and, I must say, the supposed congestion horror depicted here looks pretty tame) is that just about everything that’s proposed here is the sort of thing that, half a century later, would be seen as a nightmare from which cities were trying to awake. I don’t know the city, and I’m not sure if those waterfront highways were built, for example, but it’s hard not to see Le Corb and Broadacre City all over that image of the tall tower, surrounded by acres of parking — my initial thought was, where would you go for lunch? It’s the sort of mundane question the motopians never paused much to consider as they drafted their gleaming tomorrows.

Mark Wagenbuur has put together a fascinating video (thanks to David Hembrow) on the evolution of a Dutch street (in Utrecht) over time; of particular interest is the creeping automobilization of the street in the 1970s-80s, only to see a subsequent reversion to historical precedents (or what we now call “complete streets”).

I’m currently in Texas, and just heard an item on the radio about a curious new law: That it’s illegal to use a (hand-held) cell phone in a “school zone.”

And, as an article by Ben Wear (who was on my panel back at the Texas Book Festival last year) in the Statesman notes, cities like Austin now have to (or don’t, it’s still a bit up in the air) post signs alerting drivers to the presence of this law, otherwise police cannot enforce.

Robert Spillar, the City of Austin’s transportation director, said the city has not set aside money for the signs. Nonetheless, it will begin installing them this fall, starting with elementary schools. It could take two years to get them all up, he said.

“I don’t see how we can not put them up,” Spillar said. He said he isn’t sure the mere presence of signs will change driver behavior, and said some sort of education program might be necessary to get the message across. “It’s an unfunded mandate that has our backs against the wall. We can’t enforce it if the signs aren’t up.”

This is the first I’d heard of such a particular distinction being made in a particular zone, and I’m having trouble seeing the reasoning, or the safety impact. The first thought that jumps to mind is that a driver on a cell-phone is hardly likely to pick out a “no cell-phone” sign, much less expeditiously hang up their call as they approach. The second is that signs warning of “school zones” themselves, while a bit better — particularly when backed up flashing lights — than the ubiquitous (and absolutely ineffectual) “Slow Children” signs that are not officially recognized by engineers, tend to be little regarded as well, at least based on various tests in which drivers were still found to be routinely exceeding the speed limit; typically it’s the parent bringing their kids to the very same school. The entire concept of “School Zones” is a bit wanting, really, prone to driver and legal confusion, not to mention that it raises that eternal question: One is supposed to drive slowly and attentively on this stretch past a school, but it’s then OK to accelerate to higher speed a block later (a block on which there may be just as many children)?

And then, on the cell phone issue, we’re again making odd distinctions: We’re admitting that cell phones are a hazard to use when driving around groups of children at schools, but somehow OK when driving among groups of pedestrians or cyclists or children on the blocks in front of their homes — or in fact every other car on the road? And that it’s OK for drivers to zip past schools while talking on their hands-free-not-brain-free unit?

And then there’s the aesthetic blight of all the extra signage — more signs for drivers to ignore — not to mention all the money going to put the signs up, just so a law can be enforced; it seems rather ridiculous that if a state law is passed declaring it illegal to use cell phones in a school zone, one would have to expensively repeat that statement at every already marked school zone. After all, we don’t feel the need to erect signs announcing that driving while impaired is illegal, in school zones or anywhere else.

As always, any experiences or technical clarifications welcome.

Reader David notes that when bottlenecks develop at intersections in Ghana, or traffic grows abnormally congested, it’s not uncommon for people to spontaneously take matters into their own hands. The video is of an American friend (perhaps the advertised ‘D.J. Mayonnaise Hands’?) of his who decided to pitch in; I’m not sure he’s accomplishing much, traffic-wise, but I’d give his technique an ‘A.’

silly

I had come across the above slide, via a post at Kottke’s blog, and it is taken from a talk by a Harvard University researcher named Lant Pritchett. I was intrigued by the progression Pritchett had theorized in the way that once-seemingly controversial issues (his slide illustrates changing attitudes over interracial marriage) had, over time, simply become part of the normal state of affairs. Now, clearly this is not always a linear, teleological dynamic, but it’s interesting to try and think of other examples where it applies (a woman’s right to vote, recycling, smoking is bad for you, etc.).

I was also interested in what areas of traffic safety and the larger culture of traffic to which it might apply — seat belt usage, for example (or the idea of laws for same), driving while drunk, motorcycle helmets (or helmets in hockey and other sports), etc. And I found myself reaching for the concept in a recent column for Reclaim, the magazine of NYC’s Transportation Alternatives (of which I’m a member; if you think, by the way, that this makes me some anti-car radical, I’m also a member of AAA). The column was prompted by some recent commentary in the press, in light of the recent closing to traffic of a few blocks of Broadway in Times Square, that the NYC DOT was running a series of “elitist” reforms.

Whether this would in and of itself be a bad thing is another issue altogether — for all kinds of civic reforms we now take for granted and that make cities livable places began as the work of progressive “elites” — but I took exception with the idea that programs meant to benefit pedestrians and transit users, who represent by far the majority mode of Manhattan, were “elitist” policies causing harm to some disenfranchised majority of New York car users. But I am interested also in the reception of this and other projects via Pritchett’s evolution; in certain quarters of the media, they have been branded in the “silly” and “controversial” vein, though as this “Q Poll” indicates (the poll found early support for the Times Square project, support that might rise if the media didn’t always frame the story so negatively, or if the project’s benefits were explained to more people), we might already be moving closer to obvious.

In any case, the essay is here, or after the jump.

(more…)

Kirai (a geek in Japan) notes that there are special uni-directional pedestrian schemes in Tokyo.

In Shinjuku there are even some sidewalks with rules concerning pedestrian traffic. For example, this sign is indicating that on the right lane from 9 in the morning until 6 in the afternoon it is a one way lane.

But on the left lane the direction changes depending on the time. These “extreme” rules are needed only in districts like Shinjuku where more than three million people commute by everyday.

I wonder if this explicitly signed scheme happens anywhere outside of Japan?

Vis a vis the recent discussion at the Transportation Experts blog on the question of whether car VMT in the U.S. should be reduced as a matter of federal policy, I was curious about this factoid over on the Rocky Mountain Institute’s website.

Improve public transportation, they say. Develop housing near mass transport nodes. Form carpools at the office. These are all effective and viable measures to address the average American business commute, and we should indeed do all of these things. But what if our business commute isn’t necessarily where we have the most influence? What if it’s our kids’ activities driving us to drive more — our child miles traveled (CMTs)?

According to the 2001 National Household Travel Survey, the average vehicle travels 3,956 miles for family and personal business. In 1969, that average was 1,270 miles. We’ve tripled our family business mileage, but VMTs for business commuting only increased 36 percent during the same period. Looks like our family miles are to blame.

Is this the end for people fumbling for dropped change on the floor of the car?

Reports the WSJ:

This weekend may mark the beginning of the end for toll-booth operators and plastic coin baskets, two institutions long associated with holiday traffic and highway congestion.

On Saturday, an authority that runs the E-470 toll road near Denver is ditching its coin handlers and going entirely cashless.

One curious thing about electronic tolls; they’re more expensive.

It is unclear whether cashless toll roads will have higher toll rates than ones offering a pay-with-cash option, but some theorists say higher rates are likely. Amy Finkelstein, an economics professor at the Massachusetts Institute of Technology, has analyzed 50 years of data for 123 toll roads. In a paper to be published in the August edition of the Quarterly Journal of Economics, Prof. Finkelstein suggests electronic tolling results in rates that are 20% to 40% higher than they otherwise would be.

One reason, she speculates, is that “when tolls become less visible, it’s easier to raise the tolls.” (but is it also that electronic tolls tend to be built on new, more expensive facilities, or ones more prone to congestion?)

Do economists have a word for this phenomenon? Something about transparency? Price elasticity? But it seems a strange anti-thesis to the anchoring effect, with no frames or anchors at all.