Speed camera calibration

Nasqueron

[Deleted User] said:

TooManyPoints said:

it's been established that you must have the possibility of challenging the police evidence.

I don't think anybody is arguing about that. That possibility is provided when the defendant goes to court. What is important is that the defendant understands where the "burden of proof" lies. The police do not have to prove that their device is reliable. The defendant has to cast doubt on its reliability. That doubt does not have to be beyond any doubt at all; it has to be beyond "reasonable doubt". A court may find that it is reasonable to doubt that an unapproved and uncalibrated satnav, operated by somebody who was driving at the time, was more reliable than a Home Office approved device operated correctly.

They might find that, which is why I suggest two forms of measurement to help tip the balance.

Unfortunately it is fundamentally unfair, because the cameras are demonstrably fallible. What's more, GPS is extremely reliable, more so than the cameras for speed over a number of seconds. The receiver doesn't require calibration either.

The system is flawed and not in your favour. Hence my recommendation to do at least 10 MPH under the limit where there are cameras, or just all the time.

Speedometer cannot, by law, display a speed slower than you are going unless you modify the car substantially to screw up the car measurements. Using GPS you can verify this and it typically matches the road side alerts that tell you your speed. I have driven at 32mph on the speedo (mine is digital rather than a gauge) the GPS shows 30 and roadside detectors show 30. I find mine is pretty much a fixed 2mph over based on GPS - 52, 62, 72 all read 50/60/70 and I know from driving in areas like roadworks with a fixed speed, I routinely pass cars who are probably showing 50 but really driving 46-49. I drive at 62 (left lane of course) on the motorway so it doesn't really apply to me anyway, barring sections with a limit that is less than 60, as that gets me the best MPG without being dangerously slow

[Deleted User]

WellKnownSid said:

[Deleted User] said:

ontheroad1970 said:

maxmycardagain said:

My dashcam saved me, the ticket gave time, date and speed, the dashcam showed i was not yet in the 30, thier camera was detecting into the 60 before the 30 even started and i was doing 36(in the 60) then 28 in the 30....

That's where a dash cam CAN help, but you need the actual money shot to prove it, and you aren't likely to get that until you have turned down the out of court settlements.  Remember that the cameras have a range of 1000m and could get you before you have seen them since they are using very long lenses.

There is no way those cameras can be accurate at 1000m. Lasers diverge too much, and the simple radars in fixed ones diverge and can't discriminate at that range.

Additionally, tiny vibrations at the camera as vehicles go past are multiplied with distance, and at 1km would be randomly targeting things in that general direction, not a specific car.

We don’t trust ToF optics a mile down the road yet we put our trust in a weak radio signal from a random spacecraft 12,000 miles away? 😄

Yes, because we understand the laws of physics. The whole point of a broadcast signal is wide reception, and returns are irrelevant. Plus, is not one satellite, you need at least 4 for a position and speed fix.

A modern GNSS receiver can use multiple satellite constellations, so you have 2-3 separate measurements.

GNSS systems are used by commercial aircraft. They are certified to a very high level. While consumer receivers are not so certified, decades of data shows them to be reliable and comparable in practice.

The systems are military grade, designed for high reliability. The EU and Chinese ones are also resistant to spoofing, so there really is no reasonable doubt as to the accuracy of the data.

[Deleted User]

Herzlos said:

[Deleted User] said:

Unfortunately it is fundamentally unfair, because the cameras are demonstrably fallible. What's more, GPS is extremely reliable, more so than the cameras for speed over a number of seconds. The receiver doesn't require calibration either.

If the cameras were demonstrably fallible, you can presumably demonstrate the failure?

Lazers are incredibly accurate at any length as they are largely unaffected by external conditions, can have very precise wavelengths and timings. It's in the millisecond range so very little scope for inaccuracy. 

GPS isn't that accurate because it's triangulating between satellites, doesn't have any real awareness of the local geography, and can easily be skewed by obstacles. Have you never seen your car on a GPS jump onto an adjacent road, a field, change direction or anything? It's happened with my car and bicycle and seems more prominent on the bicycle because the route is logged and the trace line can be pretty erratic.

GPS is accurate to about 4-5m which is fine for navigation but gives a huge margin of error when trying to figure out an instantaneous speed.

Lasers are accurate to sub-mm level which means the instantaneous speed is accurate well beyond any meaningful debate. Was the car going at 46.877mph or 46.878mph? Who cares.

Lasers with a S.

You are confusing GPS positional accuracy at a given moment in time, with distance traveled over time. Positional accuracy is down to some metres, but the relative error between measurements is consistent, and thus the speed is typically better than 0.1kph for consumer grade hardware.

Newer multi constellation receivers have even better positional accuracy, sub 1m.

ElefantEd

The error isn't necessarily the same though, is it, at both endpoints of the measurement? It could be out by 5m to the North (say) at the start, and 5m out to the South at the end, total error 10m. If you are doing the measurement over 1km, that's a 1% error - higher over a shorter distance. And any speed calculations based purely on the time taken to travel a certain distance do not take into account the route taken eg curves, changing lanes etc. At the extreme end of this, you could go round in a circle at 100mph and arrive back at your starting point, your velocity as calculated by your GPS = zero!

I would be very surprised if a laser made a less accurate measurement than this, it only takes 7 microseconds for the pulse to travel 1km and back, dispersion is neglible (and in any case would be irrelevant as any reflected pulse that didn't come straight back wouldn't hit the receiver). And the calculation is done over a short distance travelled by the car. The claim that a laser beam reflects off random objects in the vicinity, giving false readings, is just nonsense. Apart from anything else presumably most of them would should a velocity of zero!

Herzlos

Over a short distance the error in GPS is likely to be the same, but if you were to take 2 location readings 1 second apart (likely the fastest your car GPS will refresh, some high end systems may be faster). At 50mph that's a distance of 22.4m, but with a ~5m accuracy you could show a distance travelled between 12.4m  (27mph) and 30.4m (72mph).

Over longer distances (in the hundreds of m) then the accuracy becomes less of an issue.

A laser can take a much faster reading with a much shorter distance travelled and a much narrower margin of error. It's using time-of-flight calculations to get an accurate indication of distance from the device, with 2 in quick succession being able to give you an accurate indication of the distance travelled between pulses. Even if they were 1ms apart it's still 1000x better resolution than GPS. They aren't even in the same ball park. 

eskbanker

Nasqueron said:
Speedometer cannot, by law, display a speed faster slower than you are going...

Fixed that for you!

ontheroad1970

[Deleted User] said:

Herzlos said:

[Deleted User] said:

Unfortunately it is fundamentally unfair, because the cameras are demonstrably fallible. What's more, GPS is extremely reliable, more so than the cameras for speed over a number of seconds. The receiver doesn't require calibration either.

If the cameras were demonstrably fallible, you can presumably demonstrate the failure?

Lazers are incredibly accurate at any length as they are largely unaffected by external conditions, can have very precise wavelengths and timings. It's in the millisecond range so very little scope for inaccuracy. 

GPS isn't that accurate because it's triangulating between satellites, doesn't have any real awareness of the local geography, and can easily be skewed by obstacles. Have you never seen your car on a GPS jump onto an adjacent road, a field, change direction or anything? It's happened with my car and bicycle and seems more prominent on the bicycle because the route is logged and the trace line can be pretty erratic.

GPS is accurate to about 4-5m which is fine for navigation but gives a huge margin of error when trying to figure out an instantaneous speed.

Lasers are accurate to sub-mm level which means the instantaneous speed is accurate well beyond any meaningful debate. Was the car going at 46.877mph or 46.878mph? Who cares.

Lasers with a S.

You are confusing GPS positional accuracy at a given moment in time, with distance traveled over time. Positional accuracy is down to some metres, but the relative error between measurements is consistent, and thus the speed is typically better than 0.1kph for consumer grade hardware.

Newer multi constellation receivers have even better positional accuracy, sub 1m.

What difference does spelling laser with a z make?  None.  

I'm surprised with your expertise on the accuracy of GPS you haven't been contracted by Mr Loophole.

wongataa

ElefantEd said:

The error isn't necessarily the same though, is it, at both endpoints of the measurement? It could be out by 5m to the North (say) at the start, and 5m out to the South at the end, total error 10m. If you are doing the measurement over 1km, that's a 1% error - higher over a shorter distance. And any speed calculations based purely on the time taken to travel a certain distance do not take into account the route taken eg curves, changing lanes etc. At the extreme end of this, you could go round in a circle at 100mph and arrive back at your starting point, your velocity as calculated by your GPS = zero!

GPS units tend to calculate speed using the Doppler shift of the GPS signals.  They don't work out the distance between locations and use the time taken between those locations to calculate speed.

ElefantEd

wongataa said:

ElefantEd said:

The error isn't necessarily the same though, is it, at both endpoints of the measurement? It could be out by 5m to the North (say) at the start, and 5m out to the South at the end, total error 10m. If you are doing the measurement over 1km, that's a 1% error - higher over a shorter distance. And any speed calculations based purely on the time taken to travel a certain distance do not take into account the route taken eg curves, changing lanes etc. At the extreme end of this, you could go round in a circle at 100mph and arrive back at your starting point, your velocity as calculated by your GPS = zero!

GPS units tend to calculate speed using the Doppler shift of the GPS signals.  They don't work out the distance between locations and use the time taken between those locations to calculate speed.

From what I've read GPS speed calculations can be done either way (presumably the cheaper ones use the easier time between two positions method). However even with the doppler effect version you are still relying on EM waves, which in the GHz range of the carrier wave potentially might suffer from diffraction and refraction effects; at any rate this measurement isn't going to be more accurate than the laser method. At best it will be just as accurate for all practical purposes.

Nasqueron

eskbanker said:

Nasqueron said:
Speedometer cannot, by law, display a speed faster slower than you are going...

Fixed that for you!

D'oh!
Edited