Use magnets to change traffic lights?

Change red traffic lights to green with magnets??

http://www.metacafe.com/watch/828731...raffic_lights/

HT

If it does work (no idea, I don't know enough about that kind of thing to even guess) I think it would probably be illegal... isn't tampering with traffic control signals illegal in most (all?) places?

I was more disturbed by the "windows dirty trick" below that one... I'm sure it wouldn't take much to write a virus that could replace/edit that file, then have it set all sorts of commonly used sites to wherever the virus writer wants you to. Probably porn, it's almost always porn for some reason (probably because virus writers are lonely 30 year old men living in their parent's basements and wouldn't know what to do with a live female even if they COULD attract one).

Actually it wouldn't be illegal. You are not tampering with the light at all. Basically, a car pulls up, a sensor senses the car, then the light turns green. A motorcylcle pulls up, but the motorcycle doesn't have enough metal for the sensor to detect it. Put neodyne magnets under the motorcycle, now the sensor behaves the way it is suppose to and turns the light green for the motorcycle.
You are not causing the traffic light to behave any differently from the way it is designed to, changing to green when a vehicle pulls up, so it shouldn't be illegal.
Now whether or not it actually works, I couldn't tell you. I might mention it to my roommate who has this problem.

Originally Posted by TomToad

Actually it wouldn't be illegal. You are not tampering with the light at all. Basically, a car pulls up, a sensor senses the car, then the light turns green. A motorcylcle pulls up, but the motorcycle doesn't have enough metal for the sensor to detect it. Put neodyne magnets under the motorcycle, now the sensor behaves the way it is suppose to and turns the light green for the motorcycle.
You are not causing the traffic light to behave any differently from the way it is designed to, changing to green when a vehicle pulls up, so it shouldn't be illegal.
Now whether or not it actually works, I couldn't tell you. I might mention it to my roommate who has this problem.

That makes sense. When I worked shifts and used to cycle across town through silent deserted streets at 3am, there were a number of red lights that just would not change to green for a mere cyclist. Once I figured out it had something to do with the strips in the road in front of the light, I actually used to do bunny-hops on them.

I feel a bit stupid now.

ETA - If this system worked, I think it would contribute to road safety, since my strategy in the end was just to check thoroughly and then slowly jump the red light when it was the middle of the night and I could plainly see/hear nothing coming for miles around. Not a good habit to get into, and definitely not allowed, but my only other option was to dismount and be a pedestrian until I was past the traffic light.

[UEL]

Years ago, I owned an Austin Mini. Oftentimes it would not set off the magnet detectors on the road for the lights.

To ensure that I had the ability to pass, I used to keep a metal shovel in my car. It served two purposes: first, I could dig myself out of a snowbank, and second, I could swipe it under my car when the damned magnet would not read the vehicle.

Oh, I wasn't really paying enough attention probably, having only gotten maybe 30 minutes of sort of sleep, and didn't catch all the details.

I thought it was being used to trigger a green when it otherwise would NOT be triggered.

Blame my neighbors... they're the fools who insist on letting their cats run around loose outside.

I had the same issue on my motorcycle. I think most of the lights here are timed during the day, and on the sensors late at night. There were several lights that simply would not change if I was the only vehicle there....usually turn lights.

I just can't see how a magnet that small would be enough to convert the motorcycle into the equivalent of a full sized car (magnetic crosssection).

Note that back in the 70's, when I drove a small car, some of the sensors were not sensitive enough to trigger on my car.

Some discussions on intersection detectors:
Stuck Signal Lights, How It Works, The Basics of Loop Vehicle Detection (warning: PDF)

[hambubba]

There are products for motorcycles.

Megaira had to put one on hers.

[Troberg]

Will not work, at least not better than any non-magnetic metal lump.

Traffic light sensors are inductive loops in the road surface (sometimes you can see the cuts in the asphalt where it is). An alternating current is sent through the loops, and if metal passes over it, the magnetic field from the loop causes eddy currents in the metal, which will alter the electrical characteristics of the loop. These altered characteristics can be detected. It makes no difference if the metal is magnetic or not, just that it is an electrical conductor. In a way, it's just a big metal detector.

These loops are sensitive enough to pick up a bicycle (assuming it's over the loop), unless it's one of those hightech, all composite things.

Most traffic counters made for permanent installation work in the same way. The company I work for makes and sells the leading traffic counter in Sweden, so I have some insight into the matter. If you go to Sweden and drive a car, I will know about it! Muhaha!

[Eddylizard]

It's not unknown to have the sensitivity set too low. One company I worked for had automatic warehouse doors operated by an induction loop which is the same principal. One of the doors was very messed up. Drive a forklift over it no problem, a powered pallet truck - 50/50 chance at best, and then only if you drove it over the sweet spot. Hand pallet truck - forget it.

It's no fun the first time you approach an automatic door standing on a pallet truck in the expectation it will open, then crash straight into it because it hasn't. Or fall off because you suddenly had to engage the emergency brakes (which are bloody sharp on those things.)

[niner]

Originally Posted by Doug4.7

I just can't see how a magnet that small would be enough to convert the motorcycle into the equivalent of a full sized car (magnetic crosssection).

Look at it this way - it's moving from a passive magnetic device (the metal car body) to an smaller active magnetic device (the magnet). It's kind of like a light sensor that sends out a beam of light and looks for the reflection - "activating" the light sensor by using a flashlight instead is going to require a much smaller beam back.

Henry

[niner]

Originally Posted by Troberg

Will not work, at least not better than any non-magnetic metal lump.

Traffic light sensors are inductive loops in the road surface (sometimes you can see the cuts in the asphalt where it is). An alternating current is sent through the loops, and if metal passes over it, the magnetic field from the loop causes eddy currents in the metal, which will alter the electrical characteristics of the loop. These altered characteristics can be detected. It makes no difference if the metal is magnetic or not, just that it is an electrical conductor. In a way, it's just a big metal detector.

Would passing a magnet over the edge of the inductor loop not produce more eddy currents than just a non-magnetic piece of metal? Similar to how a generator works? That is, a magnet passively sitting there won't do it, but the magnet entering or leaving the field triggers the change.

I'll admit, I pretty much forgot most of the fine details from my engineering days.

Henry

[Floater]

Originally Posted by Troberg

These loops are sensitive enough to pick up a bicycle (assuming it's over the loop), unless it's one of those hightech, all composite things.

I beg to differ. At least they (some) weren't 25 years ago when I had problems with one certain signal that didn't want to let me turn left at a T junction. After having been standing there for 3 or 4 cycles of changing lights (if you wanted to go straight across the junction, that is) I usually decided there was some malfunction (or rather bad design) and turned left anyway.

A FOAF has a similar problem. She's a bus driver and there is one particular place on one of her routes where there is bus only exit into another street. Time after time some eedjit in a car tries to go that way unknowing that it takes a bus to trigger the signal, a car is too small, and there they sit waiting for it to turn green (only this is a public transport signal that doesn't use colours, they have icons instead).

[Troberg]

It's not unknown to have the sensitivity set too low. One company I worked for had automatic warehouse doors operated by an induction loop which is the same principal. One of the doors was very messed up. Drive a forklift over it no problem, a powered pallet truck - 50/50 chance at best, and then only if you drove it over the sweet spot. Hand pallet truck - forget it.

Likely explanation: reenforced concrete floor. The rebars in the concrete mess up the detection, and sensitivity has to be lowered. We have had to do that on occasion when there are metal pipes or other metal in the road.

Look at it this way - it's moving from a passive magnetic device (the metal car body) to an smaller active magnetic device (the magnet).

It doesn't matter. You could put a one ton magnet or a one ton copper chunk, and they would register the same. The copper might even register better, as it's a better conductor. The biggest differenct would be the need to duck for flying pieces of metal if you used a one ton magnet.

Once again, the sensor "feels" conductive material, not magnetic fields.

Would passing a magnet over the edge of the inductor loop not produce more eddy currents than just a non-magnetic piece of metal? Similar to how a generator works? That is, a magnet passively sitting there won't do it, but the magnet entering or leaving the field triggers the change.

Movement has nothing to do with it either. Don't compare it with a generator, that's a mostly different scenario.

The alternating electrical field creates tiny eddy currents, in other words small electrical vortexes in the surface of the metal. The metal does not have to move, the "movement" is supplied by the alternating current. The eddy currents oppose the field created by the loop, thus giving it a higher inductance. A magnetic material would not produce stronger eddies.

Too bad I can't post the schematics of a traffic counter sensor, but I suspect that my boss would strongly disagree with such actions.

I beg to differ. At least they (some) weren't 25 years ago when I had problems with one certain signal that didn't want to let me turn left at a T junction. After having been standing there for 3 or 4 cycles of changing lights (if you wanted to go straight across the junction, that is) I usually decided there was some malfunction (or rather bad design) and turned left anyway.

A FOAF has a similar problem. She's a bus driver and there is one particular place on one of her routes where there is bus only exit into another street. Time after time some eedjit in a car tries to go that way unknowing that it takes a bus to trigger the signal, a car is too small, and there they sit waiting for it to turn green (only this is a public transport signal that doesn't use colours, they have icons instead).

Some possible reasons:

1. Sensitivity. As I said earlier, sometimes it has to be lowered to compensate for metal in the road.
2. Location of loop. Usually, they are not that big, covering a 2 meter long strech of the lane or something like that. The thing is, if you are close to the side of the road with a bike, you may be outside it. If you are too close to the stop line, you may have driven past it, especially if it's placed further back to trigger for buses. You are only detected above the loop, to minimise the risk of crosstalk between adjacent lanes.
3. Real life vs lab conditions. Basically, the measuring is done in the crappiest conditions imaginable. Temperature, moisture, a road body full och wires, pipes, power lines, drains, reenforcements, nearby guardrails, movements of the road body (especially during spring), trying to get it to work on a bridge and so on. Sometimes things don't work according to plan, and sensitivity has to be lowered to avoid false readings. If the traffic light was set up indoors in a controlled environment, it would detect bicycles every time, even if most of the bike is composites. Out on the road, things are not so simple and while it works nicely in most places, sometimes one just have to settle for "good enough", and bicycles may be left out.

Look at a new traffic light installation. You can see the lines cut into the asphalt where the loops are placed. On smaller intersections, there is usually only one on each entrance, a few meters from the stopping line. On larger instersections, there may be another (or more) loop further out for queue detection, to allow the lights to adjust to the traffic. If you are not above one of those loops, the system will not see you.

You can probably read more on Peek Traffic's site, they dominate the European market in traffic lights.

http://www.peekglobal.com/

Don't buy anything from them, though, as they are our competitors in traffic counters. Besides, our counters are (and I say this honestly, not just as a representative of my company) much better.

Adding to Trobergs explanation, one problem with detecting bicycles (and to a slightly lesser extent motorcycles) is that much of the conductive material is vertical to the field being disturbed, and thus creates a much smaller change in the eddy current. Thats why one of the best suggestions is to stop directly over one of the cut lines that runs parallel to your direction of travel. That puts you directly over a leg of the inductive loop, and causes the most change in the circuit, increasing your chance of triggering the signal. And many sensor systems have the memory turned off, so they don't trigger for vehicles that pull up and then turn. The sensor has to stay triggered to work, and pulling too far forward (or not far enough) will cause the system to act as if no vehicle is there.

[FullMetal]

Originally Posted by Lady Neeva

I was more disturbed by the "windows dirty trick" below that one... I'm sure it wouldn't take much to write a virus that could replace/edit that file, then have it set all sorts of commonly used sites to wherever the virus writer wants you to. Probably porn, it's almost always porn for some reason (probably because virus writers are lonely 30 year old men living in their parent's basements and wouldn't know what to do with a live female even if they COULD attract one).

well, a lot of spyware does in fact edit the hosts file, and most antivirus and antispyware software does keep an eye on that file for changes. re-directing you to another website that is full of ads, pop-ups etc. or preventing your anti-virus/anti-spyware software from updating. It's vulnerable yes, but it's still a potentially necessary file. I know I've had to use it several times when a server wasn't listed in the DNS server here at work, so it wouldn't connect over a wan connection.

granted for most people it's unnecessary, but it's there because people need it.

[SchmooPie]

Originally Posted by Robigus

Thats why one of the best suggestions is to stop directly over one of the cut lines that runs parallel to your direction of travel. That puts you directly over a leg of the inductive loop, and causes the most change in the circuit, increasing your chance of triggering the signal.

So, on a scooter, I should stop over one of the wires instead of in the middle of the oval shape?

[Crius of CoH]

Well, Troberg has it pretty fully covered - the only thing I'll add in re: the US of A is that a lot of loop sensitivity depends on how well the damned thing was installed. Too often the installing contractor will bury the loop too deep, or not deep enough and then not seal it properly, and in either case will have nicks in the loop and the wire going back from the loop to the cabinet, all of which can mess up things, especially after a few years have passed and moisture has had it's way with the wiring (sorry, bit of a pet peeve here )

Another point: most loop cards are set to disregard a constant presence after so much time has elapsed. For example, if you park your car on a loop, for the first few traffic cycles, your car is putting a call in for your traffic phase, but after a while your car essentially becomes part of the loop, and no is longer sensed - until you leave and the loop card can reset itself. I still get the occasional call about a "light not working" from the police because a patrolman's been sitting on a loop for a while.

Final sort of point: sometimes, loops aren't where you think they are. Case in point in my town: a street that is used as a two-lane road in each direction, though really it's only one lane each direction. Because it's really only one lane, there's only one pair of loops (paired end-to-end) in each direction, fairly near the double yellow line. Those folk who insist on creating that second lane near the curb have no loop detection. Sometimes, when traffic's on the light side, people will pull up in the "second lane" and wait for the light to turn... and wait.. and wait.... and I get another call about a "broken light".

That is all.