Gatsos over-read your speed
Court case could prove thousands of speeders innocent
Gatso-based evidence could be responsible for convicting thousands of innocent motorists, according to research by David Edgar, a retired electronics engineer and former professional inventor.
The Gatso speed cameras he has investigated over-estimated motorists' speed by up to 25 per cent because the time between the two flash photography images, which constitutes legal evidence on which prosecution is based, was much longer than specified.
Edgar's 35-year unblemished driving record is under threat from prosecution by West Midlands police for allegedly driving 41 mph in a 30 mph zone. Wallsall-based Edgar has pleaded not guilty to the driving offence and will defend himself in the Birmingham Magistrates’ Court on Thursday 13 January 2005.
After receiving his summons, Edgar became suspicious of the accuracy of the Gatso-based evidence and developed a laser aligned, optically triggered digital timer that measures to an accuracy within 1/100th of a second the time lapse between the two flashes produced by a Gatso speed camera.
He became concerned about the accuracy of the Gatso speed camera when his requests for disclosure about how speed cameras are calibrated for accuracy were ignored by both the police and the Crown Prosecution Service. His requests for information sent to Gatsometer in the Netherlands and UK were also ignored, as was his formal application to Birmingham Magistrates’ Court to force disclosure.
Edgar said, “I really felt they had something to hide so I decided to investigate the critical timing accuracy."
Having tested the opto-digital timer on a number of Gatso speed camera sites including Newtown Birmingham, Walsall and Cannock area, Edgar soon discovered that well over 80 per cent of them were inaccurate. In particular, there were serious timing errors between the two flashes which are supposed to flash at exactly half a second (500 milliseconds) apart.
Edgar's tests show the timings are anything but accurate. Typically they are 630 milliseconds apart, which changes legal evidence about how far a vehicle has travelled. In particular, it creates the illusion that a vehicle has travelled much further between flashes than it actually did -- some 25 per cent further.
For instance a vehicle travelling at 35 mph would have travelled an extra 2.03 metres when the timing between the two flashes is 630 milliseconds, and that puts the vehicle in the next set of parallel line markings which are spaced two metres apart.
Since these serious inaccuracies clearly affect the reliability of the actual recorded speed of a vehicle, it suggests that photographic evidence cannot be relied on by the prosecution, as reasonable doubt exists concerning the accuracy of that evidence.
Having now investigated and researched the Gatso method of speed camera entrapment the police and CPS are relying on Mr Edgar said “ I have also discovered some other disturbing facts that affect the accuracy and reliability of the entire measuring system, these will be brought to the attention of the Birmingham Magistrates Court on Thursday 13 January 2005, come along it should be an interesting day in court.”
Paul Smith, founder of the Safe Speed road safety campaign commented, "If Mr Edgar's claims are proven then almost everyone convicted by a Gatso speed camera will be able to apply to have their case reopened. I have spoken to Mr Edgar and his extensive research seems unequivocal."
I suspect Mr Edgar has absolutely no chance defending himself on such a basis for the reason above.
However - good on him for giving it a go! Anything that deters the proliferation of these infernal machines is good IMO.
Don said:
Presumably this is why trigger speeds are set high for these things. After all - this would prove beyond reasonable doubt that the driver was speeding - its just not accurate by about how much.
But if he can bring into question the accuracy then surely the case should be thrown out. If GATSOs a regularly and variably inaccurate then it's possible that some are extremely inaccurate. It's impossible to know how much... (I hope
)Isn't the speed calculated using the doppler radar effect, in some way? And not from the photographs, which are taken as a backup to confirm the speed of the vehicle, which may possibly be slowing at the time it was 'caught'?
If the speed recorded directly the system is proven to be inaccurate (which it may be) AND the photographic evidence is also uncalibrated, that would be something of a defence for all those caught by scameras.
It would help if the figures quoted were to show the difference between travelling at a legal speed and being 'caught' incorrectly, rather than just the severity of how much over the limit the car was going.
With the percentage error being indicated, that must make a noticeable difference when trapping vehicles that are travelling along at NSL speeds, rather than just in 30mph zones.
Anyone care to do the maths?
everyone with marginal speeding fines (ie not those caught doing 60 in a 30 zone) could get their cases re-opened. Penalty over-ruled, fine refunded, court costs paid and compensation too.
It could cost the government a fortune.....
.... which is exactly why he probably wont win.
IMHO if he is proven right and wins there is a case for revisiting previous convictions. Even if the Government adopts a policy of non-refund of fines paid they could still reverse the points allocations which, again IMHO, is what people are most bothered about.
Not sure if this multiplies straight through, will think about that when I've got some time, but if it does:
30 turns into
30 * 1.26 = 37.8
And 60
60 * 1.26 = 75.6
Which is a significant error.
As has been pointed out though the main measurement is taken with radar which should be bang on and the picture is just a correlation (the picture is triggered by the radar). I'm surprised the camera isn't calibrated to guarantee a shorter rather than longer time period though as the radar trips are always on the high side.
Good luck to him with the case, but apart from proving the back-up is a bit wonky can't see this doing much. Shame, I'd love to see these bloody pests removed from the roads and replaced by the enforcement of proper driving standards!
interesting news item said:
He became concerned about the accuracy of the Gatso speed camera when his requests for disclosure about how speed cameras are calibrated for accuracy were ignored by both the police and the Crown Prosecution Service. His requests for information sent to Gatsometer in the Netherlands and UK were also ignored, as was his formal application to Birmingham Magistrates’ Court to force disclosure.
This is the bit that really gets my goat. What happened to the basis of UK law? Evidence? The chap is up for something that he wishes to contest and wants the evidence that they have. But as we have seen so many times before, we are all guilty unless we can prove otherwise - something is seriously wrong here!!!!
What is also alarming is that the whole system is against you: manufacturers of the kit, the police, the CPS and the SCP..... collusion and conspiracy? Well taken on raw evidence, how could one not come to that conclusion?
Depressing and I really hope this chap wins...
mechsympathy said:
Presumably this is why trigger speeds are set high for these things. After all - this would prove beyond reasonable doubt that the driver was speeding - its just not accurate by about how much.
But if he can bring into question the accuracy then surely the case should be thrown out. If GATSOs a regularly and variably inaccurate then it's possible that some are extremely inaccurate. It's impossible to know how much... (I hope
If the period of time is longer than the prescribed 500milli/s then it does not prove that the driver was speeding. It proves that the camera has judged the car to be travelling faster than it actually was therefore no case if the camera relies on a set time between flashes. interesting to see if the camera takes into account the time between flashes as a variable and calculates accordingly or if it assumes .5secs.
I have my fingers crossed as i could do with some gov't approved tipp-ex on my licence.
I remember an old get out years ago for those captured by a hand held was to ask for the calibration evidence for the gun used. Plod often neglected to do the daily cal so loads of cases were thrown out.
In this case I wonder how the polis can define what the cal periods are before the readings become unreliable... maybe he has a slim chance of defending himself.
Also, I heard that a good get out of speeding cases (for married couples) is to plead that neither spouse can remember who was driving. It is not allowable for a spouse to testify against the other partner in such cases - some loophole of marriage law
- just fyi.fergusd said:
The time between pictures (flashes) is not relevant IF there is an accurate timestamp on both photographs . . .
...and IF that accurate timestamp is then used by the SCP office, rather than them assuming the interval is 500ms, on those occasions when they double-check the radar speed measurement against the photos before sending out the NIP.
I think this case depends on whether or not the timing of the two photographs is supposed to be precisely defined, because if the photos are simply triggered by a relatively inaccurate timer but are then timestamped with data from an independent and accurate timer, then as you say the interval is not relevant provided no assumptions are being made about the actual interval. But without more knowledge of exactly how the Gatso system works, it's not possible to say one way or the other - at the very least, this case may be useful for uncovering some of these details...
fergusd said:
Does anybody have some evidence from a Gatso that shows the data recoded along with the picture . . . ?
There's some pics shown at the bottom of www.devon-cornwall.police.uk/dcsc/enfotech/gatso.htm
Without an official description of what the text panel contents mean, there's nothing shown in those examples which looks like an obvious interval timestamp - there's a HH:MM:SS timestamp on the second line along with the date, but clearly that isn't accurate enough.
info From Link said:
The time interval between photographs, typically 0.5 to 0.7 seconds, is derived from a timing source that is standalone. Therefore the resulting speed estimate is independent of that measured by the radar.
It doesn't say that it has to be 500 milliseconds there, it could be anything from 500 to 700.
Grouch.
A Gatso has two supposedly independant methods of measuring speed. The two results should be in agreement for a successfull prosecution. If the photo system is proven to be considerably inaccurate and not matching the radar reported speed, then you have to question how any prosecutions were made with evidence from that Gatso.
If the photo system is measuring speeds too high, but it is agreeing with the results from the radar, then something is very, very fishy. A paranoid person could think of a very nasty conspiracy theory.
New film goes in the gatso.
Operator presses the test button, and the CPU prats about for a bit, tests the radar unit with a tuning fork, and tests the camera mechanism and flash unit.
Operator then performs 5 or 6 test runs above the trip threshold and these are recorded with a test code on the film.
gatso is then locked up and left to do it's 'safety' duties.
Gatso is revisited some time later (week or so) and 5 or 6 test runs are done again.
So, we have say 6 test runs, x number of speeders, then 6 test runs recorded on film.
Film removed and taken to be processed.
first and last six test tuns are examined. If there is any discrepancy between radar speed and distance travelled, the entire roll is binned. if there is any discrepancy between the speed recorded and the known speed of the test car, the entire roll is binned
The entire film is then examined to get vehicle details.
If there is a discrepancy on ANY photo between radar speed and distance travelled, then entire roll is binned.
In an event of these errors occuring, the Gatso unit is removeed from service and sent for repair/recal.
Sounds pretty fair to me, to be honest!
bryan35 said:
Right, the low down is this.
Operator then performs 5 or 6 test runs above the trip threshold and these are recorded with a test code on the film.
WHAT ?????????????????????????
You mean that the Scamera people deliberately SPEED in what is by definition an accident blackspot to check the calibration of their instruments
I'm kind of struggling to reconcile the "all speeding is evil" mantra with this practice
bryan35 said:
If there is a discrepancy on ANY photo between radar speed and distance travelled, then entire roll is binned.
Surely that can't work as the Gatso is being checked against itself,if the flash timings are out the gatso wont know its out so will calculate assuming 500ms the road markings will tie up because these are not distance travelled in 500ms but longer
Gassing Station | Speed, Plod & the Law | Top of Page | What's New | My Stuff