Amphometer Based Speed Cameras

[Sydneykid]

Recently I was asked to conducted a few tests on amphometer based speed measuring. To do this I accessed a timing system which consists of two rubber hoses and a digital timing system commonly used in motorsport events.

For the first test, I have found that by driving the front wheels over the strips at an angle (as in changing lanes) there are marked differences in the readings. These very from very minor differences at small angles of attack to quite large differences at big angles of attack.

The second test I have carried out is to let the pressure in one tyre down to 5 psi and have the other tyre at 40 psi, a reasonable hot run tyre pressure. Once again I get marked differences in the readings. Not as dramatic as those found with angles of attack, but certainly enough to make any reading suspect to within 20%.

The third test was to place a 20 mm thick piece of wood (to simulate a bump) either before, after or in between the strips. The effects were quite unpredictable and very difficult to repeat as very minor changes in speed, steering angle, acceleration and braking displayed dramatic differences in readings.

The most radical differences I get are when I have one tyre contacting one strip at almost exactly the same time as I have the other wheel contacting the other strip ie; trip both almost simultaneously. Several times I have seen timing which would equate to over 100 kph at less than 10 kph indicated speed. This is easily achieved at what I would consider normal lane change angles.

I have also asked the RTA to provide expansion and contraction rates of asphalt over the typical temperature ranges you see in Sydney in a 12 month period. This has not yet arrived, but in a number of articles I have read they typically quote numbers from -2.1% (-3 degrees Celsius) to +3.6% (+62 degrees Celsius). Remembering that these are tarmac temperatures, not ambient air temperatures So if the amphometer was calibrated to read accurately at +3.6% for 60 KPH it would read 63.5 kph at -2.1%.

I am unsure as to whether the buried amphometers used for speed camera detection have some form of control system logic which may overcome a number of these types of issues but that aside, and based on the above testing, I would have to say they are quite capable of giving inaccurate readings, just as any other form of speed measurement.

I am posting this to elicit some feedback from others who may have experience is this issue as research into this subject seems to either be suppressed or non existent.

[Roy]

I wasnt aware the RTA still used this form of speed testing, i was of the understanding they rely on radar?

[Jay95R33]

SydneyKid - thanks for the information.

Very interesting.

J

[Sydneykid]

Hi guys, a lot of people don't know that fixed speed cameras are amphometer based. The assumption has always been radar or laser. These are both far more expensive solutions, particularly on multi lane roads with no overhead bridge work to mount the transmitters. So the revenue raisers have resorted to the cheapest method possible.

Us old guys remember mobile amphometers used before radar, They got the flick because of inaccuracy, plus the 18 wheelers used to hit them with the cabin wheels and then lock up the trailer wheels and make big black spaghetti out of the rubber strips. Damn funny to watch.

[Redbaboon]

Very, very interesting information - thanks heaps

[Roy]

So i imagine the kind still used today are located beneath the road surface?

I This would remove many of the inconsistancies arising from those that are mounted on the road surface where the tractive/braking force of the tyre would upset readings.

[Ant97GTR]

Sydneykid,

Do you know if this applies to Vic fixed speed cameras? The bastards are mounted everywhere these days but I'm pretty sure they are digital and use a form of laser to trap cars. If they had to have the rubber strips mounted under the road they would have to rip up part of the road surface to do it and I certainly don't remember them doing that when they installed all those ones along the western ring rd.....

[Sydneykid]

Hi guys, the old rubber strips used to blow in the wind and jump up and down when cars went over them, they were a joke.

The new stuff is basically is just 3 X 10mm wide saw blade cuts about 200 mm apart, 40 mm deep. They then squeeze the sensors into the cuts, run the wires back to the box and then seal it up with tar. Quick, cheap and nasty.

So far I have measured 3 sets and they vary in distance apart by 8% ie: + or - 16 mm. Even the manufacturer www.poltech.com.au quotes accuracy of + or - 2% when installed perfectly and these are far from installed perfectly.

I am trying to gather as much info as I can from all sources, because the total accuracy factor has not been tested, as far as I can see, in court.

[Guest RedLineGTR]

didnt know about these...good stuff..

[browny]

On all the speed cameras (I have seen anyway) in NSW you can see the three lines that are cut into the road quite easily at normal travelling speed. In some cases on crappy roads its more noticable because the patch of tar has been replaced where the sensors are.

[predator]

I believe the VIC cameras are fully digital or radar based or something like that, and do not rely on any form of strip in the road to activate them. They basically look like more modern versions of the old red light cameras we have around here. I've seen them go up, and there are no marks are cut in the road, and are put up in a matter of days.

Some 90 new fixed cameras will be installed in Victoria before the end of the year. Considering if what you say is true, there is some +- 3% accuracy, and there is a 3km/hr tolerance in VIC, that could make a large difference.

Revenue raising however finds no distriction on the conditions of the road at the time... basically, they couldn't care less as long as they collect their hundreds of millions of dollars each year (read my sig)

[Airzone]

Yeah, I always look out for the 3 lines on the road I've never had the balls to go over them and then gas it though. Anyway, that's really interesting to know. After some more research, perhaps you might be able to convince some sort of current affair show to run a report on revenue raising and the inaccuracy of the cameras Those people will whinge at anything

[Mick Trainer]

THe Victorian speed camera (fixed speed cameras) are amphometer based. They all have three lines cut into the roads as described. Drive down the Western Ring road and you will clearly see the lines.

Sydney Kid I am very interested in your finding etc and would like to get hold of more info as it becomes available. Who has commissioned this work and for what purposes? WIll your finding be freely available?

Can you tell us if when the tyre pressure was high was the finding higher or lower? Was the differecen in the size ofthe tyre at these various pressures taken into account of calculating the speed? How was the offical speed measured?

Like I said I am very interested in this and look forward in reading more.

Mick.

[Guest INASNT]

Originally posted by predator666

I believe the VIC cameras are fully digital or radar based or something like that, and do not rely on any form of strip in the road to activate them. They basically look like more modern versions of the old red light cameras we have around here. I've seen them go up, and there are no marks are cut in the road, and are put up in a matter of days.  

Some 90 new fixed cameras will be installed in Victoria before the end of the year. Considering if what you say is true, there is some +- 3% accuracy, and there is a 3km/hr tolerance in VIC, that could make a large difference.  

Revenue raising however finds no distriction on the conditions of the road at the time... basically, they couldn't care less as long as they collect their hundreds of millions of dollars each year (read my sig)

I think those 90 new speed cameras should be rammed up steve bracks (Vic Premier whos has a speed camera fetish) ass 1 at a time. :uh-huh:

[Sydneykid]

Hi guys, some questions I can answer, some I can't.

Tyre pressure only seems to be relevant when it is different accros the front axle ie; one tyre is much lower than the other. This would also lead to an inaccurate reading when you had different diameter tyres, like say a space saver or if you had a different size, say 15" and a 17".

I have now also completed another test under heavy braking and with ABS active and found some random unbelievable results eg; we had a reading over 200 kph when stopping with the ABS activated at 30 kph.

These units made by Poltech are much cheaper than imported laser based units. Noting that radar is basically useless on multi lane roads because of its wide angle of spread. The laser based units also suffer from the requirement for regular servicing, this is costly. Plus they move (vibrate) when car & trucks go past, so they record errors and miss catching offenders. They also suffer from wind induced vibration due to the height required and the size of the structure, which is also costly. All this makes the Australian manufactured, amphometer based units very attractive.

My opinion would be that they will become even more popular in all states due to their cost effective revenue raising potential.

I would suggest that anyone who receives an infringement notice for a fixed speed camera offence less than 10% over the limit considers defending it it court. A copy of the data sheet from Poltech (available on their web site) indicates + or - 2%. Add to that the fact that most car manufacturers quote + or - 3% for speedo accuracy. In addition different tyres have different circumferences, even the same size tyre can have different measurements in different brands. The DOT standard allows + or - 4% for the same size tyre when new. If these were all cumulative you could easily be infringed at 65 kph when in fact you were doing 60 kph indicated on your speedo.

This of course does not include the sort of errors I am currently researching.

[predator]

THe Victorian speed camera (fixed speed cameras) are amphometer based. They all have three lines cut into the roads as described. Drive down the Western Ring road and you will clearly see the lines.

It depends on the place and the type.. Freeway cameras are different to the ones going up at all the intersections. As far as i can tell (and i will check) no extra lines are cut into the road for those.

[Mick Trainer]

Intersections already have the lines cut into the surface for the weights to sense traffic.

Mick.

[hippy]

i'm not geting it .. how does this detector work?? as opposed to laser or radar?

[Airzone]

If you have 2 weight sensors, one after the other and an object travels over the top of them, then you can measure the difference in time that the sensors were activated. And if you know how far the sensors are away from each other (30cm for example) you can determine the speed of the object. Speed cameras are a little more sophisticated than that, but it's generally how it works.

[Sydneykid]

Hi guys, KPH = distance over time. If the distance is constant then the time taken to travel the distance tells you the speed. Shorter time = higher speed.

The common system has 3 parallel sensors embedded in the road. The clock starts when one wheel crosses the first sensor and stops when the same wheel crosses the second sensor. Crossing the second sensor also starts another timer which stops when the same wheel crosses the third sensor. The unit then compares the two times, if they are widely different then the unit discards the reading. If they are close (don't know how close) then it converts the time to kph, if it is in excess of the posted speed (60 kph) then it takes a digital photo. In fact 2 photos, one to establish location and the other a close up of the number plate. It then sends these photos and the accompanying data to the infringement processor. They have a digital scanner that automatically recognises the number plate, looks up the owner and then sends you an infringement notice in the mail. Basically no human intervention.

On a multilane road they simply embed 3 sensors in each lane and send their signals separately back to the controller. It knows which set of (3) sensors is in each lane, records times (speeds) and takes photos appropriately.

The sensors have to be set up to record speeds for motorcycles so they only need one wheel to start the timer and one to stop it. If the vehicle is travelling straight and both front wheels hit at the same time then there is no inaccuracy other than problems as spelt out in previous posts.

The sensors for red light cameras have to laid out completely differently and triggered firstly by the red light and secondly by a wheel crossing them. This is quite different logic and sensor location in the road way. However there is the potential to use common photographic equipment and use the software programming to determine which sensor set to use at which time. Combination red light and speed camera, what a great way to earn twice as much revenue.

Hope that adds further to the knowledge base