Insensitive Traffic Lights

trafficlight

(This article from remote MDRCC member Jerry Mann)
It is very common for traffic signal sensors to fail to detect bicycles, and even motorcycles.  When no car is present to trigger the light, this results in bicyclists and motorcyclists being compelled to treat red lights as stop signs.  The problem is so widespread that North Carolina, at the urging of motorcyclist lobbying groups, has passed a law allowing a motorcycle to legally run a red light that won’t change after waiting three minutes, if safe.  The law does not specifically apply to bicyclists, however.  This creates confusion as to how bicyclists are allowed to handle a light that won’t change.Even making it legal to treat the red light that won’t change as a stop sign is often not a solution.  When traffic volume or sight lines truly require an intersection to be signalized, often either the cyclist can’t get a break in traffic allowing the cyclist to go, or the cyclist can’t see if any cross traffic is coming.  In such situations, the signal really has to work, or a very unsafe condition is created.The problem of non-detection exists primarily because personnel in charge of adjusting the sensors have lacked a reliable way to arrive at a setting that detects small vehicles, yet does not produce false detection from large vehicles in adjacent lanes.

Another problem is a general lack of understanding about bicycle detection.  Some people incorrectly think that traffic sensors can’t detect non-ferromagnetic materials like aluminum (in fact aluminum is easily detected) or don’t understand where the bicycle should be positioned for sensing (for a conventional quadrupole, it should be over the center wire).

There are design tradeoffs between detection of long, high vehicles and short, small ones.  Modified quadrupole layouts have been developed that allow more sensitivity to small vehicles, over a wider area than sensors that are common in most locations today.  However, most existing quadrupole detectors will in fact detect even the lightest bicycles when positioned over the middle wire, without producing false detection, if the sensor control circuitry is adjusted properly.  Unfortunately, DOT personnel have not had a convenient way of arriving at such a setting.  They consequently set the detectors to a far lower sensitivity than actually needed, simply to make sure to avoid false detection.

It no longer has to be this way.

There is a simple tool that permits a single person to easily adjust a traffic signal sensor to detect bicycles, yet in most cases not be overly sensitive and give false detection.  It has existed for some years, but is little known.  It was developed by an engineer in California, Dr. Bob Shanteau.  I know of it from Dr. Steve Goodridge, an electrical engineer and cyclist who has been working to promote the use of this device in North Carolina.  He has made a number of the tools, and has succeeded in getting them into use in Raleigh, Cary, Durham, and Charlotte.

On Wednesday, February 19, 2014, I put this device to the test with an actual signal, working with two NCDOT employees: JP Couch, Division 9 Traffic Engineer, and Marty Headen, Traffic Signal Supervisor.  We found that the device works as intended.  I am writing to urge its adoption by other NCDOT divisions, and by corresponding agencies elsewhere.

The device is nothing but a discarded bicycle wheel rim, mounted on a piece of ¾” (one-by) lumber, using any non-metallic fastening system.  The wood holds the rim upright, at a distance above the road similar to a narrow tire.  It thus simulates the presence of a bicycle at the intersection.  The device is placed atop the center wire of a typical quadrupole sensor layout, and the control box is adjusted to the least sensitivity that will permit the tool to trigger the signal.  The device is much easier to carry around than an entire bicycle (I actually brought it to the test on my bicycle), and because it holds itself upright, adjustment is an easy task for one person.

We tested the signal with the device, and with my bicycle.  We found that the device simulated the bicycle very well.  The same setting that allowed one to trigger the light also allowed the other to do so.

When the signal was adjusted to the least sensitivity that would detect the bicycle or the tool, we observed no evidence of false triggering or other anomalous behavior.

The tool therefore offers a reliable, simple, and dirt cheap way to eliminate a very common safety problem.  Neither Dr. Shanteau nor Dr. Goodridge nor I claim any intellectual property rights to it.  We want people in charge of adjusting signal sensors to use it.

In California, it is actually law now that traffic signal sensors be configured and adjusted to detect bicycles.  With this simple tool, there is no reason that similar standards of detector performance cannot be achieved nationwide.

For more information, please see:

http://ncactive.org/sites/default/files/Background%20Paper%20on%20Bicycle%20Detection.pdf

http://ncactive.org/sites/default/files/NCATA%20Position%20Paper%20on%20Bicycle%20Detection.pdf