Roads with closely spaced traffic signals...

[english si]

there's three separate signallised junctions between here and the white building

...and at the white building there's three in view heading round - there's actually a shocking large 100yrd gap between lights!

and the first picture is at the other end of the building from here, where there's two different sets of lights in quick succession. Just behind is a pelican crossing (oddly not synched with the rest of the lights - which have been computer controlled throughout the city for years), so another set of lights.

In half a mile, there are 9 sets of lights, 8 of which are road junctions (to Western Esplanade/from Civic Centre Road, from Western Esplanade/to Civic Centre Road, West Park Road/Commercial Road, Commercial Road, Devonshire Road, Grosvenor Square, Bedford Place, Above Bar Street/London Road), one a pedestrian crossing. They also come in threes - three clusters of three lights in ~450ft, with ~500ft gaps between those clusters.

There used to be two sets of lights here - one for Dahlia Road (and a cycle path), the other for University Road. They stopped up Dahlia Road (other than for cyclists) and turned it into one big junction with a Toucan crossing for the cyclists going up Dahlia Road. The distance between the two junctions was about 15ft. This might be the overall winner.

Bare in mind that Southampton is the city with the most signallised junctions per head in Europe and has, as the pioneer (in Europe, at least) of computer-controlled road movements, using ROMANSE, has aggressively pursued a campaign of using traffic signals at every opportunity it can, often at junctions where it would be a simple T-junction with a Give Way elsewhere in Britain, given the traffic levels. I'm surprised that the Mayflower Roundabout is still a roundabout!

[Duke87]

As for proceeding on red, look out... you'll get your picture taken if you do it at the wrong intersection. And unlike in other places that have red light cameras, in New York most of them do not have signs warning of their presence. You just get "within NYC red lights photo enforced" at the city line and from there you're in a minefield.

Are the cameras very visible, if you know what to look for?

Their look is distinct, but no, they are not easy to spot. They are, after all, positioned to take pictures from behind. They're thus a lot easier to spot when you're going the opposite direction from them.

I don't see why anyone would want to drive in NYC anyway. We're talking about the most transit rich city in America. There is no way in hell I would deal with NYC's tiny streets, subpar freeways, and relentless traffic.

Some parts of the city are more car friendly than others. Driving into Manhattan I wouldn't do not on account of traffic but on account of parking is damn expensive and every garage has valets (I refuse to hand my keys to a valet). The exception is if it's a drive-by: only stopping briefly to drop someone off or pick someone up. Then the parking issue is a non-issue.

That said, I would qualify the statement about driving versus transit as such: tranist makes sense for getting around within the city, and it makes sense for heading from outside the city into Manhattan. But if one wants to get from Connecticut to The Bronx or Queens, then driving makes the most sense. The Bronx is particularly easy to get into and out of by car.

As for the other problems, subpar freeways aren't a big deal if you're used to them, and traffic isn't a huge issue either if you know what spots to avoid and how to avoid them. 
And then there's timing. The Gowanus/BQE (I-278) may be notorious for traffic, but I've managed to do 80 mph from the Verazanno to the Triborough slowing down only for curves and to go around people... at about 1 AM Sunday morning.

[agentsteel53]

I remember when I used to drive to New York every other weekend or so, I'd park in Brooklyn where a friend of mine lived and parking was relatively easy to find, and we'd take public transportation and the occasional cab.

the one time I remember finding parking in Manhattan, it was in SoHo and I parked right in front of my other friend's house after searching for parking for precisely 2 seconds.  I figure I'd never be able to do that again!

[SidS1045]

As for proceeding on red, look out... you'll get your picture taken if you do it at the wrong intersection. And unlike in other places that have red light cameras, in New York most of them do not have signs warning of their presence. You just get "within NYC red lights photo enforced" at the city line and from there you're in a minefield.

One must also remember that RTOR is prohibited in NYC...possibly the only major jurisdiction in the country where it's still not allowed.

[Riverside Frwy]

As for proceeding on red, look out... you'll get your picture taken if you do it at the wrong intersection. And unlike in other places that have red light cameras, in New York most of them do not have signs warning of their presence. You just get "within NYC red lights photo enforced" at the city line and from there you're in a minefield.

One must also remember that RTOR is prohibited in NYC...possibly the only major jurisdiction in the country where it's still not allowed.

I've been trying to figure out the reasoning behind that.

[tradephoric]

This thread reminded me of a video i watched on youtube about a city in England that shut out the lights at a bunch of closely spaced traffic signals in the heart of the city:

A previous poster mentioned the city of Southampton uses an adaptive system known as ROMANCE.  I went to Southhampton's ROMANCE website and found a cool link that gives you travel times on some main corridors through the city.  Does the city use the data collected from the intersections to create this map and estimate the travel times along the signalized corridor?

http://www.southampton.romanse.org.uk/
(you have to click on 'Journey Times' tab in the legend)

[english si]

I think it's ANPR (auto number plate recognition) and Traffic Cameras, though I'd imagine that they probably used to do with clever calculations with traffic lights and guesses at traffic speeds, but ANPR would be easier and more reliable.

[Duke87]

One must also remember that RTOR is prohibited in NYC...possibly the only major jurisdiction in the country where it's still not allowed.

I've been trying to figure out the reasoning behind that.

Because traffic conditions are such that it's rarely "safe" to make a right on red. And in circumstances where it is, they do have exceptions...

[Dr Frankenstein]

As for proceeding on red, look out... you'll get your picture taken if you do it at the wrong intersection. And unlike in other places that have red light cameras, in New York most of them do not have signs warning of their presence. You just get "within NYC red lights photo enforced" at the city line and from there you're in a minefield.

One must also remember that RTOR is prohibited in NYC...possibly the only major jurisdiction in the country where it's still not allowed.

Because of the large amount of traffic and the pedestrians, making the risk for an incident much higher.

Montreal has the same restriction.

[roadfro]

A previous poster mentioned the city of Southampton uses an adaptive system known as ROMANCE.  I went to Southhampton's ROMANCE website and found a cool link that gives you travel times on some main corridors through the city.  Does the city use the data collected from the intersections to create this map and estimate the travel times along the signalized corridor?

http://www.southampton.romanse.org.uk/
(you have to click on 'Journey Times' tab in the legend)

I think it's ANPR (auto number plate recognition) and Traffic Cameras, though I'd imagine that they probably used to do with clever calculations with traffic lights and guesses at traffic speeds, but ANPR would be easier and more reliable.

I would hazard a guess that there is some sort of sensors out on the roads that do the calculation for a given segment. It's a bit tricky to get a travel time to be fairly reliable on a surface road, especially when there's signals and many intersections or driveways present.

In Las Vegas, the FAST system has radar sensors set up approximately every 1/3 mile on the major freeways. These sensors can do vehicle counts, as well as figure out approximately how fast a vehicle is moving. Using the data from these sensors, an average speed over a third-mile segement can be obtained. These speed calcs are aggregated to display travel times on VMS devices, as well as produce a similar speed map on their website. Similar systems are used by other agencies, and the speed detectors can vary from the radar devices to microwave and even "old-fashioned" inductive loops in th pavement.

[tradephoric]

Oakland County (NW of Detroit) uses as adaptive traffic signals system known as SCATS to control over 630 signals.  The county uses the degree of saturation at each intersection and uses that data to set up a rudimentary real time congestion map for the major surface streets throughout the county.

http://www.rcocweb.org/Pages/Real_Time_Traffic_Map.aspx
(in the legend, click inside the mini county map for the congestion data to display)

The map only shows the spot congestion at each individual intersection but it assumes that the link between two congested intersections is going to be congested as well.

[english si]

A previous poster mentioned the city of Southampton uses an adaptive system known as ROMANCE.  I went to Southhampton's ROMANCE website and found a cool link that gives you travel times on some main corridors through the city.  Does the city use the data collected from the intersections to create this map and estimate the travel times along the signalized corridor?

http://www.southampton.romanse.org.uk/
(you have to click on 'Journey Times' tab in the legend)

I think it's ANPR (auto number plate recognition) and Traffic Cameras, though I'd imagine that they probably used to do with clever calculations with traffic lights and guesses at traffic speeds, but ANPR would be easier and more reliable.

I would hazard a guess that there is some sort of sensors out on the roads that do the calculation for a given segment. It's a bit tricky to get a travel time to be fairly reliable on a surface road, especially when there's signals and many intersections or driveways present.

As I said, ANPR would be more reliable than the complex calculations that other methods (radar sensors, induction loops, etc) would be for average speed. ANPR is also what they also use on the freeways in Great Britain for sites like Traffic England. Having some closed circuit visible light cameras and software that picks up number plates and then calculates the point-to-point time for number plates that appear on two cameras (and then ditches the number plate data, hopefully) would be the easiest and most accurate way, now the software exists to process the video (though it was invented in the 70s), to measure average speeds - especially if there are driveways and side turns and stuff.

Doing a bit of digging (though it hardly took any), ROMANSE does use ANPR for journey times. Traffic signals will use the induction loops (given they are all at junctions, it's not good for journey times as the speed data would be off) as they use SCOOT.

[Brandon]

A previous poster mentioned the city of Southampton uses an adaptive system known as ROMANCE.  I went to Southhampton's ROMANCE website and found a cool link that gives you travel times on some main corridors through the city.  Does the city use the data collected from the intersections to create this map and estimate the travel times along the signalized corridor?

http://www.southampton.romanse.org.uk/
(you have to click on 'Journey Times' tab in the legend)

I think it's ANPR (auto number plate recognition) and Traffic Cameras, though I'd imagine that they probably used to do with clever calculations with traffic lights and guesses at traffic speeds, but ANPR would be easier and more reliable.

I would hazard a guess that there is some sort of sensors out on the roads that do the calculation for a given segment. It's a bit tricky to get a travel time to be fairly reliable on a surface road, especially when there's signals and many intersections or driveways present.

As I said, ANPR would be more reliable than the complex calculations that other methods (radar sensors, induction loops, etc) would be for average speed. ANPR is also what they also use on the freeways in Great Britain for sites like Traffic England. Having some closed circuit visible light cameras and software that picks up number plates and then calculates the point-to-point time for number plates that appear on two cameras (and then ditches the number plate data, hopefully) would be the easiest and most accurate way, now the software exists to process the video (though it was invented in the 70s), to measure average speeds - especially if there are driveways and side turns and stuff.

Doing a bit of digging (though it hardly took any), ROMANSE does use ANPR for journey times. Traffic signals will use the induction loops (given they are all at junctions, it's not good for journey times as the speed data would be off) as they use SCOOT.

Around Chicagoland, two methods are used.  IDOT tends to use induction loops buried in the expressways, and has started using a system more like ANPR.  The Tollway uses I-Pass data.  They track I-Passes along a given section of tollway to determine speed and congestion.

[hm insulators]

Well, the one way avenues are nicely timed to have the lights progress up the block at about 30 MPH (the speed limit). And indeed, if traffic permits, you can expect to drive 20, 30, even 40 blocks up 1st Ave without hitting a red light.
Park Avenue and other two-way avenues, naturally, do not and cannot have such convenient light timing. There you will hit a red light every 5-10 blocks because, while the greens do at least seem to cluster together, they can't be made to progress with traffic since traffic goes both ways. I don't know what the precise strategy is. But I do know that NYC traffic signals are controlled by boxes at the site of the signal itself (some of which are mechanical - you can hear parts click and clank in there as the light changes!). They don't have a central computer controlling things. So, whatever coordination there is between signals is all still done the old-fashioned way: by having guys go out and set the up one by one onsite.

"Click and clank as the light changes"? Man, what is this, the 1940s?

[roadman65]

NYC also has signals that are at one way streets without any cross traffic that warrants its need.  For example: Central Park West has stoplights on inward one way streets away from Central Park on the east side of this road and they are timed to turn red with the others (one way outward onto this street with lights) spite there is no reason to stop anybody for anything.  If a pedestrian wants to cross the street at anytime a simple crosswalk button can be installed.

[Alps]

NYC also has signals that are at one way streets without any cross traffic that warrants its need.  For example: Central Park West has stoplights on inward one way streets away from Central Park on the east side of this road and they are timed to turn red with the others (one way outward onto this street with lights) spite there is no reason to stop anybody for anything.  If a pedestrian wants to cross the street at anytime a simple crosswalk button can be installed.

You don't understand NYC. Pedestrians are ubiquitous. The traffic lights are absolutely about them, and given the volumes present 24/7, NYC is fine with a fixed-timing system that allows crossings at every block.

[Kacie Jane]

ANY central business district is going to have traffic lights at nearly every block.  I know Chicago and Milwaukee do for sure.  I bet most of NYC - especially Manhattan (midtown and lower in particular) have such.

Since this thread is back at the top, I'll post the following...

You mention that most CBD's will have lights at every block, but what about places where signals are even more frequent than that?  Downtown Seattle has several signalized mid-block crosswalks.  Heading south on 5th Avenue from Pine Street to University, not only will you hit a signal at each of the four intersections, but also at three pedestrian crossings.

This doesn't meet the criteria of the original post though, as it's a one-way street, and generally well-timed.

[roadman65]

Kirkman Road in Orlando, FL has way too many signals on it.  The spacing between them is 1/10 of a mile through the Metro West/ Valencia College area.  During rush hour, get prepared to be stopped at every one of the signals (or 90 percent of them anyway) from FL 50 to FL 482. Long waits at LB McLeod and Vineland as well as the Major Boulevard entrance to Universal because these are other arterials with 4 movements (3 at LB McLeod cause its a three way intersection) and a long crosswalk at the theme park.

[Mark68]

W Burnside in downtown Portland...not only are there lights at every cross-street, there are no left turns. Most of the cross-streets are one-way couplets, so if you're driving westbound on Burnside and want to turn left to go south, you need to turn right (on a northbound one-way), then right, then right again.

Also, E 13th & 14th Aves in Denver are a one-way couplet that have some "pedestrian" lights mid-block, and they seem to turn red at random, methinks to keep one from exceeding the speed limit. There never seem to be pedestrians when these lights turn red. These aren't the only streets that have these either, but these are ones that seem to annoy me the most.