Small circle thing between traffic lights

[stormwatch7721]

What are those small tall things that stand between the traffic lights? this is what I'm asking about: http://goo.gl/maps/UWbVA

[Alex]

What are those small tall things that stand between the traffic lights? this is what I'm asking about: http://goo.gl/maps/UWbVA

Signal detectors.

[stormwatch7721]

Ohh, ok. I didn't even know that.

[M3019C LPS20]

That is a video detection camera. Just an example of vehicle actuation. It is probably, by far, the most popular form in today's world of traffic control.

[hbelkins]

Kentucky uses loops far more than cameras.

[cl94]

PA seems to love them. A lot of lights, especially in northwest PA, use overhead detectors. In Erie, for example, every single light seems to have them, even temporary signals.

New York, on the other hand, uses them pretty sparingly. Here, they tend to be used for older retrofits and wide divided highways (NY 78 in Amherst/Lancaster/Clarence and Walden Avenue in Cheektowaga). The typical format here for new installs and retrofits is inductor loops with one or two additional located 1/10 mile or more before the light on major roads. For a while in the 90s and early 2000s, they used ~0.5x0.5 ft square metal detectors placed in the middle of the lanes. They didn't last well with the harsh climate and resulting potholes and are being slowly replaced with inductor loops. Several of those that remain are a hazard for motorcyclists and cars with the potholes forming around them, actually preventing people from driving over the detector.

[US71]

Kentucky uses loops far more than cameras.

Arkansas is going more towards cameras. Only problem is if you get a gust of wind, it messes up the camera alignment.

[M3019C LPS20]

Video detection is beneficial in some ways. One way, as an example, is that you could adjust the digital loops that a camera focuses on at an intersection if necessary. No need to cut pavement for the installment of magnetic ones.

New Jersey has been in favor of video detection for quite a while, and a lot of intersections that originally used loops now use this form of vehicular detection.

[xcellntbuy]

Video loops are quite popular in south Florida.

[DaBigE]

Pavement loops are still the default around Wisconsin, however WisDOT is beginning to use cameras more now that they're experimenting with adaptive signal controls (WI 100 around Mayfair being one such area). They've also been regularly using cameras for temporary (construction-related) signals. A few other communities around the state have begun to embrace cameras--West Bend being one of them, Madison also has a few sprinkled around the city.

[jeffandnicole]

Video detection is beneficial in some ways. One way, as an example, is that you could adjust the digital loops that a camera focuses on at an intersection if necessary. No need to cut pavement for the installment of magnetic ones.

New Jersey has been in favor of video detection for quite a while, and a lot of intersections that originally used loops now use this form of vehicular detection.

They seem to fail quite often though, sending signals into max length patterns.  Or, I can walk in the ped crosswalk with no other traffic around, and suddenly the signal will needlessly change, because the video detector detected something (me) in its view.

[froggie]

Arkansas is going more towards cameras. Only problem is if you get a gust of wind, it messes up the camera alignment.

They seem to fail quite often though, sending signals into max length patterns.  Or, I can walk in the ped crosswalk with no other traffic around, and suddenly the signal will needlessly change, because the video detector detected something (me) in its view.

Minnesota has been using camera detectors for over 25 years and does not seem to have these problems.

They work pretty good, especially in northern, snowy climates.  They work in/after snowfall, can be easily adjusted as needed (as someone noted earlier), work when pavement is ripped up and don't need to be replaced after a repaving.

[jeffandnicole]

Arkansas is going more towards cameras. Only problem is if you get a gust of wind, it messes up the camera alignment.

They seem to fail quite often though, sending signals into max length patterns.  Or, I can walk in the ped crosswalk with no other traffic around, and suddenly the signal will needlessly change, because the video detector detected something (me) in its view.

Minnesota has been using camera detectors for over 25 years and does not seem to have these problems.

They work pretty good, especially in northern, snowy climates.  They work in/after snowfall, can be easily adjusted as needed (as someone noted earlier), work when pavement is ripped up and don't need to be replaced after a repaving.

Have you walked around intersections or sat at them when no other traffic is around to notice?  And while it's certainly not every intersection, if it's one or two that someone travels thru each day, the problem gets old quick.

[roadfro]

One downfall of video detection is weather can mess with the detection. The cameras actually detect changes in contrast within the detection zone drawn on the image. Depending on sensitivity, light vehicles may not be detected during snow, or heavy snowfall can obscure some vehicles. Catching the sunrise/sunset in view of the camera at certain times of day can cause false calls--Las Vegas had to put extended visors over some cameras when they were first experimenting with video detection (in its infancy).

Also, camera placement is crucial. At certain angles, large vehicles in one lane can block detection of vehicles in other lanes, which can be detrimental if using the cameras for certain applications (i.e. ramp meters or automated vehicle counts).

[tradephoric]

An ongoing problem with video detection is false calls due to shadows.   Under the right weather conditions (IE. sunny morning), shadows can cause a video detection zone to flicker and over-count.  If the intersection is adaptively controlled, these false calls can lead to poor intersection operation (IE. left turn phases cycle when nobody is present due to shadows from the adjacent lane, main-corridors gets squeezed because of faulty detection flickering on the side-street, etc.). 

Now, the use of forward looking infrared (FLIR) cameras may be able to address the shadow issues since they focus on the heat signature of the vehicle.  Of course, FLIR has its own set of problems.  One is cost, but the other has to do with appearance.  A new state of the art traffic operation center would probably want to showcase crisp and clear video as opposed to grainy looking black and white video. 

[tradephoric]

If you watch the FLIR video it's not hard to spot "faults".  No manufacturer is going to claim 100% detection accuracy, but detection accuracy should be a prime consideration when choosing what type of detection to use at an intersection.  Inductive loops are still the most accurate type of detection IMO. 

 

[roadfro]

An ongoing problem with video detection is false calls due to shadows.   Under the right weather conditions (IE. sunny morning), shadows can cause a video detection zone to flicker and over-count.  If the intersection is adaptively controlled, these false calls can lead to poor intersection operation (IE. left turn phases cycle when nobody is present due to shadows from the adjacent lane, main-corridors gets squeezed because of faulty detection flickering on the side-street, etc.). 

Now, the use of forward looking infrared (FLIR) cameras may be able to address the shadow issues since they focus on the heat signature of the vehicle.  Of course, FLIR has its own set of problems.  One is cost, but the other has to do with appearance.  A new state of the art traffic operation center would probably want to showcase crisp and clear video as opposed to grainy looking black and white video. 

I hadn't heard of FLIR, but that seems pretty neat. (I like on the video image that it simultaneously shows the indicates the state of each signal phase.) I wonder though how well FLIR detects bicycles, which would not have nearly as high of a heat signature as a vehicle... Video is sensitive enough to detect bikes and even people--I've seen it used to automatically activate warning beacons at trail crossings whenever a ped or bike approaches.


I've not known a traffic command center to use the detection cameras to monitor video feeds. More often than not, at least in Nevada installations, the detection camera image is only seen by the controller or on a monitor within the cabinet for the purposes of setting up detection zones. There's usually a separate color camera mounted at one corner of the intersection with full tilt/pan/zoom features. That's what the first video looks like.

[Roadrunner75]

How common is emergency vehicle preemption systems at signals now?  I know there are a couple of different methods.  I remember the old "flash the highbeams" trick did actually work at one intersection near my home town 20 years ago, but a lot has happened since then...

[hm insulators]

Induction loops in the pavement in Phoenix and also the Los Angeles area.

[roadfro]

How common is emergency vehicle preemption systems at signals now?  I know there are a couple of different methods.  I remember the old "flash the highbeams" trick did actually work at one intersection near my home town 20 years ago, but a lot has happened since then...

I think it's one of those that some areas use but others don't. In Nevada, every signal in the Las Vegas area, and most (if not all) signals in Reno/Sparks and Carson City have preemption sensors.

By the way, most likely the "flash the highbeams" trick didn't actually work, rather it coincided with what the signal was already doing. Most preemption signals wouldn't pick up headlights from the low positioning at the stop line (and some require rapid light flashes in distinct patterns likely unable to be easily duplicated by drivers).

[Roadrunner75]

How common is emergency vehicle preemption systems at signals now?  I know there are a couple of different methods.  I remember the old "flash the highbeams" trick did actually work at one intersection near my home town 20 years ago, but a lot has happened since then...

I think it's one of those that some areas use but others don't. In Nevada, every signal in the Las Vegas area, and most (if not all) signals in Reno/Sparks and Carson City have preemption sensors.

By the way, most likely the "flash the highbeams" trick didn't actually work, rather it coincided with what the signal was already doing. Most preemption signals wouldn't pick up headlights from the low positioning at the stop line (and some require rapid light flashes in distinct patterns likely unable to be easily duplicated by drivers).

Yes, I'm aware that in most cases the highbeam thing is more of an urban legend and the vast majority of these are coincidence.  However, this was 20 years ago and a lot of technology has been developed and adjusted.  The signal, which is adjacent to a firehouse, almost worked too well when flashing at it after dark, and it created a ridiculously short yellow interval (probably around 2 seconds) for the cross-street to the point where it would clearly cause someone to run the red.  For awhile I had to cross through the intersection frequently so I got to test it for the preemption versus standard detection, and it was very consistent with the high beams to the point where I was worried about causing an accident (or getting a ticket).  Then one day it suddenly stopped working for the beams, and I figured that they finally adjusted it or just removed/replaced the system.  My friends caught on and used it as well while it lasted - with one who became convinced that all signals had it, leading to him barreling through a few red lights...

[jeffandnicole]

How common is emergency vehicle preemption systems at signals now?  I know there are a couple of different methods.  I remember the old "flash the highbeams" trick did actually work at one intersection near my home town 20 years ago, but a lot has happened since then...

I think it's one of those that some areas use but others don't. In Nevada, every signal in the Las Vegas area, and most (if not all) signals in Reno/Sparks and Carson City have preemption sensors.

By the way, most likely the "flash the highbeams" trick didn't actually work, rather it coincided with what the signal was already doing. Most preemption signals wouldn't pick up headlights from the low positioning at the stop line (and some require rapid light flashes in distinct patterns likely unable to be easily duplicated by drivers).

Yes, I'm aware that in most cases the highbeam thing is more of an urban legend and the vast majority of these are coincidence.  However, this was 20 years ago and a lot of technology has been developed and adjusted.  The signal, which is adjacent to a firehouse, almost worked too well when flashing at it after dark, and it created a ridiculously short yellow interval (probably around 2 seconds) for the cross-street to the point where it would clearly cause someone to run the red.  For awhile I had to cross through the intersection frequently so I got to test it for the preemption versus standard detection, and it was very consistent with the high beams to the point where I was worried about causing an accident (or getting a ticket).  Then one day it suddenly stopped working for the beams, and I figured that they finally adjusted it or just removed/replaced the system.  My friends caught on and used it as well while it lasted - with one who became convinced that all signals had it, leading to him barreling through a few red lights...

Being that this was at/near a firehouse may have something to do with it.  While some states (PA, DE) install pre-emption devices at nearly every signal, NJ never does.  Per their guidelines, they won't either, unless someone else (county, township, hospital, firehall, etc) pays to have them installed.  So when someone is flashing their lights at most intersections in NJ, there's nothing there to preempt the signal no matter what vehicle approaches the light.

[andrewkbrown]

I do believe some very early emergency vehicle preemption devices worked with just flashing light, such as headlights until people starting catching on to it. Thereafter places started using Opticoms, a small strobe light transmitter placed on the roofs of emergency vehicles and receiver on the traffic light assemblies. Then newer ones began using infrared transmitters. Others use a specific siren tone to preempt the traffic signal.
It varies as to what the traffic signal does when preempted by an emergency vehicle. Some give a green to the oncoming emergency vehicle, while all other sides get red. Some change all directions to red, which allows an emergency vehicle to proceed, while defeating the attempt for someone using a black market preemption transmitter.