Faking out traffic sensors

July 4, 2005 4:47 PM

Inductive traffic sensors: can I make them "see" my bike better if I stick some strong rare-earth magnets on the frame?

My bike (actually a recumbent trike) doesn't trip the traffic sensor and there's basically no car traffic at that point to trip it for me, although there's a lethal amount of cross traffic. If I put some really strong magnets on the underside of the frame, will my trike cause enough disturbance in the sensor's electromagetic field to signal my presence? The whole trike weighs about 20 kg and the frame is 4130 chromoly. Thanks!

posted by Quietgal to travel & transportation (9 comments total)

no. at least, afaik, they're inductive loops. so you need something that reacts against a change in magnetic field, rather than a source of magnetic fields. you might take some thin, enamelled copper wire and make a couple of turns around the frame (ie making a big loop - not around a tube, but around the whole frame). don't forget to join the ends after you've added a few loops (clean off the enamel insulation).

(i have no idea if that will help!)
posted by andrew cooke at 5:35 PM on July 4, 2005

This suggests that the wheels should work well enough to set off a signal, you just have to position your bike correctly on the sensor.
posted by borkencode at 5:48 PM on July 4, 2005

The trick is to keep moving. Try pedalling backwards (assuming this just spins the chain), or just keep rocking the bike from side to side if it's not one of those three-wheel recumbents.

Also, make sure you pull up in the middle of the lane. Here in Australia, some of the older sensors have little bikes painted over the middle of the sensor so riders know where the best place is to be noticed.
posted by krisjohn at 6:06 PM on July 4, 2005

(if bike wheels generally work, but your recumbent has small wheels, i think my suggestion makes sense, from a physics pov, because the important thing is the size (area) of the loop and the number of loops. so several loops of wire around the frame will probably be equivalent to a large wheel.)
posted by andrew cooke at 6:41 PM on July 4, 2005

What I do (with a regular bike) is to tilt the bike at about a 45 degree angle to the pavement. Seems to work.
posted by winston at 6:59 PM on July 4, 2005

When I get desperate, I lay my bike down on its side. Works every time. (Ridiculous, I know.)
posted by waldo at 8:24 PM on July 4, 2005

Many intersections around Seattle have a small "T" painted where you're supposed to put the front wheel of your bike to trip the sensor. Also, if the sensor isn't tripped by a bicycle, it's officially defective and they want to know so they can come out and fix it (adjust its sensitivity, or something). Local laws may vary.

andrew cooke: My understanding is that the detectors have generally changed away from being metal-detector-type coils, which are triggered by a change in the inductance or lossiness of the coil. Now they're magnetometer-type detectors, which sense the momentary change of the field when an object moves nearby. In which case, dangling a magnet under your bike would make your bike more detectable.
posted by hattifattener at 12:04 AM on July 5, 2005

interesting. googling suggests that they do exist, but i can't find anything saying they're now used by default. but if they're used because they work better with bikes (no idea if that's true) then inductance is still the kind you have to worry about.
posted by andrew cooke at 4:53 AM on July 5, 2005

Thanks, everyone! I tried positioning my trike carefully over the "sweet spot" and it worked! You guys are the best!
posted by Quietgal at 7:33 PM on July 5, 2005