California teen offers GPS challenge to speeding rap
A California teenager is contesting a speeding ticket which claims he was doing 62mph in a 45mph zone, since a GPS system fitted to his Toyota Celica appears to show he was actually within the limit. Shaun Malone, 17, was caught on 4 July by a Petaluma police officer using a radar gun, AP reports. The lad had in the past …
An inventor caught by kent police for apparently doing over 40 in a 30 zone presented evidence from a GPS product he was testing which showed his true speed to be 29.7mph.
Kent police have now suspended use of speed guns pending investigations.
All given GPS locations have a level of inaccuracy due to various issues - most recievers will show this somewhere.
Since this inaccuracy can change by a wide margin very quickly this makes snap distance calculations a problem. Since speed is worked out using distance and time this will affect any speed readings.
I have seen this on a handheld unit which would alter the speed by upto several mph whilst a) standing still and b) moving at a steady pace.
GPS is a location system, and should not be relied upon for accurate speeds at a given point - only an average over some time.
Radar (and laser based) speed guns have issues with aiming and the electronics picking up the fastest thing in their entire field of view, which in the past has been shown can even be a fan in the officer’s car when improperly aimed.
GPS has the issue that if any one of the signals to the minimum of four satellites required for valid positioning drops out, the GPS does a combination of using its internal clock against the remaining signals plus the last valid velocity and direction readings to create a “pseudo present reading” to extrapolate the position ( and from that, speed & direction) until it reacquires that minimum of four satellite signals again to truly calculate valid position information.
The comment from the officer about having a satillite signal possibly causing a delay demonstrates his lack of understanding of how a GPS works.
The 'delay' is actually what the GPS uses to measure it positioning - effectiving triangulating itself based on the known position of the satellites and the delay in signal. As such, the position and time recorded are accurate to within 3-30 meters and 1-5 nanoseconds respectively. However the resolution of the signal is once a second, so speed is calculated over one second intervals.
In a nutshell, the GPS is dead accurate.
Clearly whats more important in this issue is intent to speed, as the officer indicated he may have been infringing for 3 seconds, but clearly there was no intent as he knew there was a logger attached to the vehicle. the logger clearly shows that this person did not intend to speed on his journey and clearly did not speed through-out his journey.
what type of vehicle was it? it has been proven that flat surfaces reflect speeding devices and can infact inclued closing speeds of vehicles infront (and even bridges of windscreens) or behind, This has happened in the UK where (because of our two photo requirement - fixed time calibration) it was proved that a driver was doing 15mpg in a 30mph zone but the vertical rear window reflected the closing vehicles speed also, so the camera fired.
For allegedly doing 36 in a 30 zone by a laser wielding copper. Thing is, i fought it because I had my Tom Tom running at the time and both warns me of the camera position (they were lasering right next to a fixed camera...) and displays my speed in RED if I exceed the limit PLUS I'd been done in that exact spot 2 years before and was thus quite careful.
Needless to say the magistrate wasn't having any of it and I got fined an amazing £232 + 3 points - presumably for not submitting to the unjust tyranny of the fixed penalty.
This whole speeding ticket situation is on the verge of collapse with in-car GPS speed monitoring.
ANPR is next, of course...
I read about that story too. If enough speeding tickets get reversed in this country I can see the speed guns being a thing of the past, but what can we look forward too? More gatso's?
It is a well know fact that GPS is not a reliable and accurate way of measuring speed. Especially when it's cloudy.
Can't believe the courts are even considering this...
If you have physical access to the gps device, then its completely possible to edit the data via a pc link. From the police's point of view it's probably not robust enough for a defense, unless there is an external, controlled logger somewhere the driver cant get at the data - an external company for example.
The older radar gubs have to be calibrated every 6 months. A friend of mine, who actually manufactured these, said this was the exception, rather than the norm.
The laser guns are more acurate, unless hand help. movement of a mm or 2 at the gun, at an object 2 miles away (which many UK forces do), means by the time it hits the vehicle,it is moving several metres. Not so acurate now !
Didn't he get off because the plod in question had retired so it didn;t even reach court.
If you are ever caught speeding, please do take it to court.
It took my case 9 months to be heard in court, after several letters to Thames Valley Police highlighting their mistake and repeated requests from Thames Valley Police to own up and pay £60 plus 3 points on my licence.
Thankfully they saw the light and admitted fault in court. I was only entitled to expenses I had receipts for, which was the £3 car parking ticket I had to pay to attend court that day: I did not engage a solicitor even though I received plenty of legal advice (for free!) from the guys at www.safespeed.org.
I kept the £3 cheque from TVP and will frame it one day!
Details of my case (and many others) are on www.safespeed.org 's forum.
I'd have to agree with Tim Epstein above.
Was recently given a ride in a friend's new BMW M3, which came fitted with a GPS thingummajig - which was telling us the names of the roads we were turning into as we were still turning, which implies that the delay ain't all that big...
Are we to assume that Paris Hilton didn't have a similar get-out-of-jail-free card BTW?
..is it take in to effect changing altitude? For example, Consider two roads parallel to each other but one of them has an incline. If two cars travelled at the same speed along both roads would their GPS readings be the same?
Just a thought!
Speedometer.
Who needs new tech when drivers just need to stick to the limits?
I think that doing "15mpg in a 30mph zone" should indeed be a crime - what were they driving, a huge lorry?
You should always contest speeding tickets - that way they will hopefully make sure of their facts before even thinking of prosecuting. If they can't be sure who was driving or the equipment is not in calibration or is misbehaving or they haven't been properly trained - they should not prosecute.
@Tim Epstein:
If speed is calculated by dividing distance and time, it can never be more accurate than the accuracy of either. The time is accurate to a few nanoseconds, and therefore clearly not the problem.
The distance is the problem. If a location by GPS can be accurate to with 3-30 meters, the distance between two points is accurate to 6-60m. So in a worst case scenario, to determine the speed with a 1% accuracy, the best thing your TomTom can do is give you the average speed over the last 6 km! Even at an accuracy of 3 m, you still need 600 m to get that 1% accuracy.
Conclusion: GPS is not 'dead accurate' and clearly proves nothing in this case.
GPS can be wildly inaccurate. Problems with satellite pickup, moving between satellites all sorts of things can throw it out. It takes a lot of careful analysis to actually validate a GPS report. In the sport of Sailboat racing "GPS peaks" are notorious. In one on line competition recently a potential winner was demonstrated to have GPS results so far out that some of the plots were on dry land half a mile away from the lake he was sailing on!
All well and good until you do get busted because of a dodgy speed camera/gun and have no way of proving you were within the legal limit.
However accurate the technology if it is maintained and used by people there is the potential for error and that should be a factor that is considered during prosecution.
However, it is not perceived to be in the immediate interests of the police, courts, government or suppliers of speed-gun/camera devices to admit the potential for error or to consider it during prosecution. The fact that the pursuit of justice ~should~ be the prime interest of all these groups falls by the wayside in deference to expediency, reduction of costs and (in the case of the treasury and manufacturers) income.
The point is that the driver _was_ sticking to the limit, but the speed gun said he was speeding. With just the speedometer you have no proof that you weren't speeding, so you get the fine and the points despite being innocent.
If you read the article closely it states that "it sends a signal every 30 seconds that records his whereabouts and travel speed", I assume the position is from gps and the speed from the speedometer.
So minor inaccuracies in the position won't affect the accuracy of the recorded speed.
"You should always contest speeding tickets - that way they will hopefully make sure of their facts before even thinking of prosecuting"
Well, when I got my 3 points the copper showed me the recorded speed on the gun, and sadly, it was correct :o) D'oh
Surely if they allow a GPS unit (which I'm sure the manufacturers probably don't guarantee for accurate speed calculations) as a defence for speeding, they should also be allowed to use the same GPS unit to prosecute?
Hedley, you do know that by law all speedometers are innacurate, so the only way to know your true speed is GPS.
Also as has been proven in a UK Court, Radar guns are not 100% accurate, so perhaps your smugness should be reigned in a bit.
ps. there are two types of people in this world, people who have broken the speed limit, and liars. ;)
you will need a GPS with speed signal from the car since those without are NOT accurate enough...
Any way, why not just install a taco in that boy race box?
http://news.bbc.co.uk/1/hi/england/wear/7033353.stm
"Who needs new tech when drivers just need to stick to the limits?"
When drivers stick to the limits and the police prosecute them wrongfully, we need all the tech help we can get to curb the advance of a Police State.
I'd agree with the raw figures, but the logic behind the calculation is flawed.
The reading does not suddenly jump straight from one extreme to the other (in my experience, probably only about 200 hours total of real/constant GPS use), and therefore the probability of it being that significantly out is minimal.
I've been playing with a home made satnav/speedo/etc using a GPS receiver on my motorbike, and I've never had it be far enough out to get a ticket when I didn't expect to. In fact, I think it's more accurate than my speedo most of the time.
The GPS receiver I have (overnight) varied by up to 500m, and that was within a walled building - if it were in the open I'd expect it to be a lot smaller range. That was also (IIRC) before the intentional inaccuracy was removed.
Final comment - the speed guns are known to be flawed and give flawed readings, the GPS is known to be normally pretty good. Sounds like "reasonable doubt" at the very least to me.
Speedometers over state the speed by up to 10%, so are possibly less accurate than GPS devices. But anyhow regardless of that, given the story about that inventor who PROVED that he was going CONSIDERABLY slower than he was accused of, there appears to be a chance that he was under the speed limit (either according to his speedometer or not).
If these types of camera's are not completely accurate then they should be banned. PERIOD. Everyone knows these are just money making scams by the police, so they probably don't really care about the accuracy unless the public begin to realise that it is a big CON or get ‘outed’ in public (like that inventor did).
Why haven't all forces reviewed their policy on using these camera's? Why... because their inaccuracy is making them money. When you get the ticket, it tells you that if you contest it they might decide to make the punishment worse. Surely that's blackmail?! But it works, and most people give in for fear of greater punishment.
Camera Safety Partnership - pah! Jobsworths the lot of them. They need there to be an issue so they can all have jobs. No-one is independent. Speed is a contributing factor to the amount of impact in an accident, but in most cases is NOT the CAUSE. If speed = crash, then why don't most plane's always crash when they take off. The most common cause of an accident is bad driving (or driving error). To solve this, get more Police actually out there driving around, picking up people with bald tires, no lights, or driving like an idiot.
This comes from a frustrated driver who is fed up with being taken as a mug by a bunch of money grabbing morons. I know very few people who haven't had points on their license, and those that haven't I don't know many who haven't come close. And fact, I know of no-one who hasn't sped at at least one point in their life - shock, horror... anyone who says they haven't is either a complete liar or is one of those drivers who should be pulled over for going 35 in 60 zones (clearly not recognising the national speed limit sign).
</rant>
My (admittedly very old) Garmin records the maximum speed I've reached is 199 mph. Given I have a Landrover, I'm trying to remember exactly when it was I drove off a cliff...
The GPS inaccuracy is mostly due to variance the system introduces unless you are military. So your calculation is totally off. Between two readings separated by one second, the signal will bump about around 10 meters. GPS devices know that so they do a lot of averaging before presenting the data.
My GPS in the car gives stable readings to the point that I trust it to have at least 0.5kmh accuracy.
The trouble is that my GPS and many other commercial car systems do not offer any logging that can be used in court as evidence.
There needs to be a better understanding on how position accuracy affect speed calculations. Most errors in position accuracy do NOT come into play with speed accuracy. Sounds totally illogical I know but it is true. This is because most position inaccuracies are caused by atmospheric conditions. Another by terrestrial interference. The ones caused by atmospheric conditions are a gentle drift and affect the position relatively slowly and affect all samples by similar amounts. This type of interference has virtually NO impact on speed accuracy.
If you are in a city or have other natural things around like hard mountains, you can suffer terrestrial interference. This tends to happen very quickly and cause a very rapid deviation in position. This one can cause a large spike in speed but tends to only last for a couple of refresh cycles.
Another issue with accuracy is actually satellite geometry. If the satellites the receiver is using are all close together in the sky the math gets much harder because the angles are so small. If the satellites are overhead as well as near the horizon the accuracy in position and speed will be much higher. If you find your speed jumping around more than usual check the DOP reading in your receiver. You will likely find it is high or jumping around itself. This would indicate a geometry issue with bad satellite position relative to where you were.
Newer GPS receivers have filtering for terrestrial interference. They will automatically remove any samples that are considered outliers. The slow drift is also correctable here in the states if the GPS receiver has WAAS (Wide Area Augmentation System) ability. This system enhances position accuracy as well as accuracy speed accuracy.
Expect a GPS to report your speed accurately within about 0.3mph under good satellite geometry and low terrestrial interference conditions. Plenty accurate to beat a traffic radar.
Anne van der Bom is someone else who doesn't understand GPS. Accuracy may be +/- 3-30 meters, but it doesn't wobble by that amount between samples. If it did then someone standing still would be timed at between 0 and 67MPH. In practice the worst case drift you will see is +/- a couple of MPH, which is better accuracy than most car spedometers, and averaged over 5 or 10 samples you can improve on that.
To be honest Im not sure I like the idea of of using GPS (or any other continuously recording telemetry device for that matter) to verify speed. If everyone is happy to start using this information to disprove static and hand-held speed guns it is not that difficult a step for the powers that be to say "fair enough, we will stop using speed cameras (General public Huraah!).... we will just use the telemetry directly form you car (General public D'oh!)...". How many speeding tickets do you all think you could automatically generate on your daily round trip to work???
Jon
@Anne:
"If a location by GPS can be accurate to with 3-30 meters"
Yes, but that's an /absolute/ location accuracy - the differential accuracy is much much better, especially over consecutive time intervals.
I will leave it to a GPS boffin to say how much better ...
As most gps units do not rely on position to calculate speed the accuracy (even discounting dodgy weather conditions etc) is usually +/- 0.5mph
A quick search for 'gps speedometer doppler' will explain.
To achieve any degree of accuracy with a gps system, you need to apply DGPS, or 'Differential GPS' . This requires at least 2 gps recievers and also often employs terrestrial radio beacons too. A moving, low-end gps reciever is not capable of precision calculations.
Also a car travelling up an incline will always appear to travel faster than one on the flat that has the same landspeed. Conversly racing downhill will give a slower reading on a gps than is really the case. This is real hair splitting, I know, but just to prove that gps is a global *positioning* system, and the devices do not generally posses the computing power required to calculate speeds. This is why an OEM navigation system in a car also includes a hardwire connection to the vehicle's own speedo system. (makes it work in tunnels too). TomToms etc are just cheap 'compasses' with a map database.
Even my handheld eTrex (1998) shows an altitude of -8 m when I walk on the beach......
@Anne van der Bom
"If a location by GPS can be accurate to with 3-30 meters, the distance between two points is accurate to 6-60m'
Doesn't follow.
GPS may have an uncertainty of several metres in terms of absolute position, but if you move exactly 10m south it will tell you that you've moved exactly 10m south. It will still have the same uncertainty about your absolute position, but not your relative one. That's how differential GPS can be accurate to millimetres.
As for it's accuracy against RADAR, GPS measures distance travelled horizontally. If you're going up or down a hill you WILL be travelling faster along the road than a GPS shows, and a RADAR gun should accurately measure your road speed, which is what you get ticketed for. From what I remember there are a lot of hills around Petaluma...
30 km per hour is 8.3 meters per second (30,000m/3600 seconds)
GPS recievers get a signal every 6 seconds?? (I did not find a definitive source for this), so the reciever will go about 50 meters between signals. Assuming an error of 5 meters, there is a 10 % variance in speed calculated vs actual speed.
If GPS sends more often (say once a second), the calculated velocity is useless unless you average out the signals over time (5 meters of possible error in 8 meters of distance). If GPS sends a signal only once every 30 seconds, then a 5 meter error over 250 meters between signals means the speed calculated is accurate to 2% - you get the same if you average the speed over 30 seconds.
The problem is that the police are measuring instantaneous speed while the GPS is measuring average speed. So don't tell the judge this :-)
Sorry for not using Reg units, the conversion to such is left as an exercise for the bored reader :-)
@Anne van der Bom
You would be correct if your assumption about how speed was averaged in the receiver was correct. However it is not correct. In your assumption you imply that the errors are cumulative over a long period. They are not.
GPS does use time and distance to calculate speed but it is done over 30 or fewer position fixes. With the average sample rate of 4hz the time to accumulate 30 samples is only a few seconds and not a whole trip. Most modern receivers use a variable number of previous samples to average into the displayed speed. When speed is changing rapidly fewer samples are used to make the reading more responsive. When you are traveling at a steady speed more samples are averaged together to give a more stable reading.
Yes, a lorry in a traffic jam... im sure its on google somewhere.. (maybe13mph??)
Gatsos Specifically: At low speeds with flat backed vehicles passing traffic reflections can cause false readings, say you were doing 12mph stationary object reflections would be 24mph , and passing traffic at say 26 would be a reflection at 50mph. hence the camera is triggered.
However these are easy to appeal as the only evidence is the pair or photos with a calibrated timegap showing distance travelled, and hence prooving innocence.
The Police are supposed to issue tickets based on operator measurements but have been know to just issue tickets to all people photographed. (its cheaper!)
The 3-30 error in one point does not correlate to another 3-30 error added on between two points. They are orthogonal and the two (assuming identical error properties) add up to be the square root of 2 times the error of one. Or about 5-45m.
However, a laser speedo has errors in calibration and in that the measurement is a field average or max (usually the latter, since the device isn't all that useful to police if it's adding in the stationary trees).
The speedo in the car has errors too. Gearing isn't 100% accurate and the change in wheel size as pressure changes (the speedo is really measuring the speed of rotation) and of course the ammount of slippage is a large error margin there too.
So, the person says they weren't speeding. Well, they would.
The policeman says they were. Well, they would for very much the same reasons.
The laser says they did (but the police can change the logs) and it has errors.
The GPL says they didn't (but the owner can change the logs) and it has errors.
So we can't prove who's telling the truth.
But, since the policeman is not on trial, we aren't trying to prove him lying. So we must allow the putative speeder the benefit of the doubt.
@ Peter D'Hoye
"The GPS inaccuracy is mostly due to variance the system introduces unless you are military."
Em... May 1st 2000?
Ring any bells?
End of US military induced SI (Signal Interference).
F5 your info dude.
Simple, buy a bike. no licence plate on the front and a pic from the back cant incriminate you. ;)
@ Anne van der Bom
You probably only need 10% accuracy in a case like this (unless the police are really anal), which at 3m GPS resolution is only 30m. Still, at 30m resolution, you'd need 600m, which is still definitely greater than most speed traps.
Nonethless, the point remains that even if GPS speed measurement could achieve that level of accuracy, the devices would need to be calibrated, certified and tamper-proof in order for their evidence to be admissable in court. Consumer/commercial navigation/tracking systems are of no use for that present.
Question. Why GPS? Why don't commercial vehicle owners just attach a datalogger to the speedo?
The comments up till now assume that GPS speed is measured by dividing the difference between two positions by the time difference. This is not usually how it works, because position measurements can be wrong, by a large margin.
What GPS receivers actually do, is measure the doppler shift in the satellite signal as the receiver moves relative to the satellites. Usually this is accurate to 0.1 m/s or 0.36 km/h or 16.68E-9 Ssx in El Reg units.
The upshot of this is that speed measurement is much more accurate than the accuracy of positions would lead you to conclude (this also answers Dave Baraham's question on the altitude differences, although the altitude component is less accurate).
If this system is being used in his defence, they should take all the historical records, plot them onto a map with speed limits and also use the data to convict him if he broke the speed limit at any time in the past.
This should only be used in defence if you are going to be bound by all the data collected.
I've got a Garmin c320, and once I have a location fix the speed is pretty darn accurate. I've never had a problem with overcast days or snow fall. I can accelerate and de-accelerate and see my speed realtime, and it always matches the speedometer (Go straight to map mode to see your speed. You can't see it if you have put in a destination).
...Also...to the person says they've never received a ticket because they use their speedometer and stick to the limit....I guess you've been lucky enough never to have been hit with an inaccurate gun, or been hit in POP mode which they aren't supposed to use due to inaccuracies.
The GPS is not checking location every 30 seconds, it is continually monitoring its own location. The device UPLOADS data every 30 seconds, which INCLUDES the average speed across that 30 second interval. You guys are assuming that the average speed is calculated using a location every 30 seconds, which could lead to inaccurate readings. However, the average recorded by the device is accurate since it is a continual reading of the position in fractions of a second.
Random inaccuracies every 30 seconds could lead to a 60m range difference from point A to B. But if the readings are every 1/10th of a second, then you have 300 readings from point A to B, and thus those inaccuracies average out to nothing. Because it's being monitored continually, there's no chance that the kid can speed up between points A and B without the device monitoring it.
