A step backwards in traffic signal automation
 

The city I live in uses pole mounted microwave vehicle detectors on practically every signaled intersection, as harsh winters would destroy the in ground variety. Even on a bike they work like a charm, just move left into the driving lane as you approach the intersection and the light changes.

However over the past year I've been encountering more and more intersections where I never get detected. A traffic engineer got back to me to say that they started using "true-presence" sensors that can eliminate false positives caused by the pole bouncing in the wind. It would seem cyclists falls within the threshold of a false positive and are duly ignored.

It creates a bit of a dilemma, if I know the intersection I'm approaching will never go green should I even bother waiting at the red at all. Some of these detectors are at intersections with dedicated left-turn lanes so there isn't even a pedestrian button to press. The problem will only get worse as old model detectors are swapped out.

Since the engineer is talking to you (a rarity in itself) try to get him to consider seeking a balance between false positives and bicycle detection. It's possible that some fine tuning can work acceptably for all interests..

The answer might involve tuning and aiming to a standard "sweet spot" which bicycles can move into to be read. If that works, painting a marker on the pavement would be the next step.

If dialog and experimantation doesn't yield a solution, then it's a matter of speaking to the city council (or whoever's in charge) and seeing if they'll give bicycle needs some priority and build it into the plan.

My city does a pretty good job of adjusting for bicycles when I notify them of lights that won't trip. The few lights that still won't recognize me, I just go when the traffic allows.

Perhaps a tin foil helmet would do the job:)

It just might, need to google "how to increase radar signature".

edit: I gotta try and find one of these: http://www.westmarine.com/WestAdviso...adar-Reflector a 3" radius disc gives a 1 square meter radar signature, still kind of big.

Doesn't help with your microwave sensors, but some of the newest traffic signals around here use cameras, which have worked reliably for me on my bike.

For a radar reflector on the front of a bike, you need the apex cut off a cube (corner plus three sides) facing forward. This would be like 1/8 of the westmarine reflector which is intended to be omnidirectional. Corner reflector - Wikipedia, the free encyclopedia The material needs to be conductive. Effectiveness depends on its size compared to the wavelength of the radar. For K and X bands, 5 cm along an edge should be adequate.

"True-presence" doesn't sound like a very meaningful term for an engineer to be using. Are they microwave based too or something different entirely? Either way its a failing on the behalf of the engineers/planners involved - bicycles are vehicles and they all need to catered for. Point out that failing to do so could encourage cyclists to run red lights - thats a safety issue. For the same reason I don't think suggestions as to how to make you own personal bike pick up the signal are very helpful - I mean they are helpful to you, but they don't help other riders much.

Locally, I don't get detected, either.

Sounds like a term one might use when trying to respond to (or blow off) a lay inquiry.

I got search hits for that term in various patent applications, DOT documents and product literature, so it is probably an industry term to differentiate different types of sensors.

Does the foil go inside or outside the helmet?

Gave building my own reflector a try just for kicks. I figure unidirectional will be sufficient so I wouldn't need a sphere. Used a compass to draw a 3" radius circle on a piece of non-corrugated cardboard. Then carefully drew lines dividing the circle in 4 and cut out one quadrant. Prefolded it inwards, then applied foil HVAC tape.

I'll take it for a test ride tomorrow.

http://bikeforums.net/attachment.php...hmentid=397753

First test inconclusive but promising. I did a tour of various problem intersections, but other vehicles were present at most of them even waiting a cycle or two to see if it would clear. But there was one intersection I've never been able to trigger, that gave me an arrow when no cars were present. I'm going to have to get out very early in the morning to make multiple approaches with and without the reflector to see if it has any effect.

I noticed the sensors are in two different types of configurations. Pole mounted on the opposite side of the intersection pointing toward the traffic lane. And pole mounted on the same side of the intersection pointed downward at the lane. A single direction reflector might not work on that second type.