Well, I wasn't gonna get sucked into this thread, but...
The following is a compilation of what I wrote in
this thread.
The 'in pavement' sensors are a wire loop. The loop has a specific AC frequency running around the loop which is monitored by a computer. When a significant sized FERROUS object rolls over and into the loop it causes a significant change in the AC frequency. The computer identifies this change to mean a vehicle is present. When your ALUMINUM motorcycle rolls over the wire and into the loop the frequency change is substantially less. For your motorcycle to be detected the computer must be programmed to have significantly better sensitive to small frequency changes. There is reluctance to do this because the more sensitive setting can cause adjacent lane sensing errors. In electronics terms, a vehicle rolling into the in pavement wire loop acts as an inductive tuning slug that significantly alters the AC frequency running in the wire loop. Motorcycles and bicycles are a small tuning slugs which yield a small AC frequency change.
Over the last decade there have been improvements in loop design which will improve detection, will reduce adjacent lane falsing and be more accommodating to large trucks whose trailers can be 3'-5' above the wire loop. The new loop design is supposed to reliably detect bicycles without causing adjacent lane falsing. These intersection detection systems can easily start at $1 million and go up from there so there hasn't been a rush to upgrade existing systems.
When I was working in the city of Manchester, New Hamster they rebuilt several intersections around our building. After the work was complete there were 3 significant intersections where one or more lanes would not recognize my FJR. On a lark I looked up the city's Highway Department and sent them an email about the vehicle sensing. In a Stunning Turn of Events™ I got a letter back from the man in charge of the Highway Department in less than 2 hours. I provided detailed information about the intersections, lanes and issues. In Stunning Turn of Events II -- by noon the next day the city work crews had readjusted all the intersections. Nobody in the USA was more shocked than I was at noon the following day, when all the intersections worked. The Highway Department guy even wrote back to ask if things were working properly.
In an area that you travel regularly you are not helpless or an eternal victim to the situation; at least make an effort to let someone know. At worst nothing will happen, but sometimes it may actually help and improve things.
Magnets are the equivalent of adding the same mass of ferrous metal down low; the magnetic field does very close to zero. Magnets are to tripping lights as copper bracelets are to managing arthritis; if you believe, then it works.
The traffic light controllers have a pre-programmed time delay between when the light is triggered and when the actual light change takes place. Some times a left turning car from the opposite direction will have already tripped the light before you pull into the loop. These two factors can make it seem like putting the center stand down or a strong neodymium magnet has initiated the triggering. In any case, the size of the vehicle does not determine when the light changes, it is a delay set by the software.
If you approach an intersection which has a rectangular box cut in the pavement (Dipole loop) you will have the best chance of triggering the light if you pull up with both wheels aligned with either the right or left long side of the rectangle. If the intersection has a rectangular box with a third cut right up the center (Quadrupole loop) place both wheels directly on the center cut. There are no guarantees, if the sensing loop is broken or out of adjustment there is almost nothing that a bicycle, scooter or motorcycle can do to trip the light, no matter how clever the rider is or how well advertised the Light Tripping gizmo may be.
Edit: I see a lot got posted while I was pasting these bits together.
