Speed Camera Calculation
By Kevin Ash
Pictures: Sussex Police, copyright free
Under UK law, any claim that a rider or driver is breaking a speed limit must be corroborated by an independent source, so in Britain (but not all foreign countries) all automatic speed devices must make two separate measurements of a vehicle’s speed.
The radar element of a GATSO camera is switched on all the time, and this measures the speed of every vehicle passing. If that speed exceeds a predetermined limit (a few percent above the posted speed limit), the second element of the device is triggered and a camera operates. This takes not one picture but two, at an accurately set time interval which will be something between 0.5 and 0.7 seconds. The smaller time interval is the most commonly used on conventional cameras.
Click on image for both pictures full sizeThe images are used of course for identifying the alleged miscreant from the number plate, but because there’s an accurately timed interval between successive ones they also show how fast the bike was travelling, using the white lines painted on the road to measure the distance moved in this time.
In the Peter Barker case it’s clear that the bike has passed four of these lines (in fact a little less) in the 0.5 second time interval – this is marked in the top right of the second image. In the top right of the first image the radar-measured speed of 38mph is indicated, far enough above the posted 30mph limit to trigger the double camera flash.
The spacing of the road markings can vary but usually it’s five feet, as it is in this case.
The bike therefore has travelled 20 feet in 0.5 seconds, a speed of 40 feet in one second, 144,000 feet per hour. Divide this by 5,280, the number of feet in a mile, and you get 27.2mph, pretty much what you’d expect from someone passing a speed camera in a 30mph limit.
Click on image for both pictures full sizeThe radar-detected speed should have been a close match for that, within legally specified tolerances that certainly don’t allow for the radar to read 40 per cent more than the camera-measured speed, as it does here. Clearly there’s a fault and certainly no basis for a prosecution.
Many forward facing cameras do without those white lines on the road. These are Truvelo cameras which use radar in the same way as Gatsos, but back this up with three piezo-electric strips close together on the road within the camera’s range. The pressure of a tyre crossing these produces an electric signal, and the time between successive signals gives a second measure of the bike’s speed. Only one picture is taken, an infrared one used for identification only which needs no flash (which could be blinding).
What about hand-held speed guns? Amazingly, despite requiring very close, mathematically defined corroboration from two independent automatic devices in speed cameras, a hand-held gun needs no second device to back it up as the law says the opinion of the policeman holding the gun is sufficient evidence! As if he’d disagree...
Fri, 07/08/2009 - 10:49
#2
Me again on this - Kev in you article you state the distance between the markings as being five feet. Can you tell me in general how many of the markings are there, how wide they are and at what point do they use for reference? Is it the markings themselves, the distance between them, the distance from the back of one to the front of the next or ?? It seems like there could be room for tolerance build up there.
Fri, 07/08/2009 - 16:28
#3
To answer your second mail first, the markings are five feet apart front edge to front edge - the measurements are taken from the same edge on one to the next. I don't know the number of markings, I'd guess they might need more where they think the limit might be broken by larger amounts. There's certainly room for inaccuracy in judging where exactly a bike is in relation to the markings, but this method should show if there's a major discrepancy, as it did in the Barker case. The scandal here is more that the police went ahead with their prosecution attempt anyway, and worse, then tried to cover it up.
The potential for radar inaccuracies is quite broad, although mostly things like temperature variations won't make a huge difference, less than 1mph, so if it says you're speeding it might be wrong by a small amount but basically you'll still have been speeding. The bigger inaccuracies comes with bikes in particular as these are small targets with multiple surfaces which can cause problems reflecting back the radar signal. I suspect that happened here with the bag hanging off Barker's bars. That's why there's a cross check with the camera, and when this difference is great the results should be rejected.
In fact GATSO and other speed test equipment is tested and recalibrated regularly, I think every six months, and if a certificate is out of date then a prosecution becomes invalid.
Mon, 10/08/2009 - 11:37
#4
How do you find out if the certificate is out of date & whos responsibility is it to service & calibrate speed cameras
Mon, 10/08/2009 - 11:48
#5
I understand that the equipment shall (or at least should) be calibrated on a regular basis but to what standard and by whom? In industrial metrology there are a myriad of comprehensive standards including those published by ISO which form a traceable base for establishing specification. These standards are independantly formulated and provide a rock solid base of reference. There are, howver, new technologies used widely in some very new manufacturing measuring equipment where there is no standard yet written and they are actually basically at the whim of the manufacturer as to how they determine the equipment accuracy. That is an outline of the situation and not for debate here but is it a base that also applies to this technology?
.
It is all good and well for the authorities to say that they have the cameras calibrated every six months but what is that calibration, who is determining how the test should be done and is it truly applicable or relevant? I am not saying that there is anything amiss but it most certainly could be possible and is not unprecedented.
Mon, 10/08/2009 - 21:08
#6
I haven't checked and I can't at the mo as I'm abroad - something called a 'holiday' apparently - but I believe the calibration process is carried out by independent labs approved by the equipment manufacturer. There are rigorous standards in place that they have to meet and performance criteria which the equipment must meet - it's all very controlled and scientific. There doesn't seem to be any argument with these or the process that I've heard, and I'm sure there would have been by now with so many people being prosecuted by radar and with so many lawyers trying to get them off again.
The police must provide evidence of the calibration being in date on request.
Tue, 11/08/2009 - 08:07
#7
I am sure that there has been many a legal examination and also that records are meticulously kept. I am also pretty sure that within the guidelines as there is then things will have been challenged and all found to be ok. I guess what I am driving at though is who is setting the standards? If the equipment manufacturers are the ones approving the labs then that, in my opinion, is questionable. I think it should be a truly independant authority who set everything up and they who accredit the manufacturer and the independant laboratories.
.
This is of course all conjecture and getting back to the original point what is more worrying is the attempt by the police to proceed with the prosecution in question. I may try to find out the answers to my thoughts however, just to satisfy my own curiosity.
Fri, 11/09/2009 - 07:39
#8
Are the cameras in police cars accurate in comparison? I have heard that it is down to the operator to choose the buttons on the camera. the distance covered is worked out, then the speed is worked out. how can this be so accurate?
Vinny
Mon, 28/09/2009 - 16:28
#9
The ones you're thinking of here are VASCAR speed devices. With these the operator has to push a button when the car or bike passes a certain point (often those white squares you see painted on the road surface) then has to press it again when it passes a second point. These might be a quarter of a mile apart or so, and have to be quite a distance to compensate for the obvious operator error or pushing the button too soon or too late. The problem here is, if a policeman is determined to get you, for whatever reason, he can simply delay pushing the first time and push the second time too soon, and there's no proof that he even attempted to push the buttons as you crossed the marker points.
There are so many questions and points that can be related to this issue that we could talk about it for hours on end. One thing though that I always wonder about and as it is an area I have a lot of background in then I feel I can question with some knowledge, is that of the certification and control of the equipment used by the enforcement agencies. I would very much like to know who checks, adjusts and then certifies all of the speed measuring equipment and how often they do it. I would also like to know under what conditions any calibration is performed and if these conditions are replicated at the time of use. By way of parallel explanation, it is very commonplace in industry for precision metrology equipment to be specified, manufactured and certified at an ambient temperature of 20 derees celsius, with a specified allowed variant over a set time period - say plus or minus one degree over a twenty four hour period. There are also conditions for humidity etc. In reality a large proportion of measuring equipment is never used in those actual conditions and thus they will not meet the manufacturers accuracy figures. So I do question if something similar could happen with radar guns, gatso cameras and so on.
.
In my apprenticeship days, many moons ago, I worked for one of Britain's leading radar manufacturers and the guy who was the engineering head of the place got nabbed with a radar gun. He took it to court and basically blew them out of the water technically saying that what they were using was inherently unreliable and he won. I accept since the early 80's technology may well have moved on but has every facet been truly examined? The answer to that, judging by the evidence given here is that it simply has not been.
.
In any manufactured system there is a tolerance on all components and that is a simple fact of life. No one can machine a perfectly round circle, a perfectly flat face, straight line etc. There are acceptable limits built into everything and the size of those limits determines the precision and as items are used things can and do drift out of tolerance. Thus for anyone to say that their system is infallible, just goes to show that they are not qualified to actually be talking about it.
.
I am sorry that this is a long post but it is a serious issue and has to be pursued all the way and we all need to make as much noise as we can about it.