Accuracy of GPS as a speedometer
Discussion
GPS is fine, not that accurate for my work, but for general uses, accuracies of 3 – 10 metres or so is okay if you’re lost at sea or in the mountains, it narrows it down a bit.
I note that a number of contributors appear to rely on their GPS speed, particularly for camera sites, knowing that it isn’t that accurate how accurate is it?
My car shows three speeds displayed – speedometer, cruise control and GPS
The speedo will show 83, cruise will be set at 80 and GPS shows 77 occasionally 78 or 76
I note that a number of contributors appear to rely on their GPS speed, particularly for camera sites, knowing that it isn’t that accurate how accurate is it?
My car shows three speeds displayed – speedometer, cruise control and GPS
The speedo will show 83, cruise will be set at 80 and GPS shows 77 occasionally 78 or 76
Thing is the position error seems to be stable over reasonable periods of time ... i.e. if you leave a GPS stationary, although the position it gives may be out by a few metres, it doesn't dance around from second to second. For example, my mate had a cheap and cheerful GPS attachment for a palm pilot a few years ago (after selective availability had been switched off) that could quite reliably show me crossing the road through our village. It might say that I had moved from the front garden to the pavement when in fact I had crossed the road, and other times it would show me magically walking through the post office ... point is it was quite accurately showing my movement, but the absolute position that was a bit off. That would suggest that speed measurements could potentially be rather more accurate than absolute positioning.
I had my Chim speedo checked on a rolling road - amazingly it was spot on at both 30 and 60 mph !. Then compared it to GPS at same speeds and GPS agreed 100% with the speed. Then I compared it to my other cars. Cant remember the exact figures but VW Golf speed over read by about 3MPH, Vauxhall Omeaga by 4mph - or thereabouts. Pretty much what I would expect speedos should err on the side of caution and over read a bit.
I always rely on the GPS reading when approaching scameras now.
>> Edited by bigdods on Saturday 7th January 20:22
I always rely on the GPS reading when approaching scameras now.
>> Edited by bigdods on Saturday 7th January 20:22
justinp1 said:
Car speedos all read slightly over.
Not so. By law a cars speedometer should be accurate to witin -0% to +10%, but it only takes someone to put a secondhand gearbox with a different final drive in it and you can have a speedo that under-reads. Also the popular aftermarket white dial kits require the speedo needle to be removed in many applications, and it's very easy to put it back on in the wrong place.
Most commerical GPS receivers are stated to be capable of an accuracy of 0.1mph or better under ideal conditions, but in the real world it would be safer to assume +-0.5mph due to multi-path reflections etc.
MR2Mike said:
justinp1 said:
Car speedos all read slightly over.
Not so. By law a cars speedometer should be accurate to witin -0% to +10%, but it only takes someone to put a secondhand gearbox with a different final drive in it and you can have a speedo that under-reads. Also the popular aftermarket white dial kits require the speedo needle to be removed in many applications, and it's very easy to put it back on in the wrong place.
Most commerical GPS receivers are stated to be capable of an accuracy of 0.1mph or better under ideal conditions, but in the real world it would be safer to assume +-0.5mph due to multi-path reflections etc.
Ok, just to confirm, I agree: if you change the gearbox of your car, or change tyre diametres, it will fool your speedo.
Car speedos *should be* reading slightly over in almost every normal scenario to cover for the fact that over thousands of miles your tyre will wear, and just change the exact circumference of your wheels and thus alter the actual speed reading.
Depends on a variety of things, such as whether your spped is constant, whether the road is very twisty or not. The GPS speed is calculated from the position coordinates as an average, not as an exact speed at that point in time. In general they should be pretty good, unless the Americans turn down the satellite accuracy from 10 metres to 100.
I always reading the Autocar road test for the McLaren F1.
They always have a table showing indicated speed versus actual speed in the acceleration tables, and invariably for every single car out there, the figures would differ. An indicated 60 would be an actual 58 and an indicated 120 would be an actual 110 - this is fairly consistent across the board.
But the McLaren F1? Indicated 60, actual 60; indicated 120, actual 120 etc.
Now I understand that it's no technical feat to get a speedo to read accurately (so much so that I've been told car manufacturers overread their speedos deliberately), but it was always one of those things which made it different from every other car out there.
And back on topic, I am of the unscientific opinion that my GPS-indicated speed is very accurate in a wide-open-area and at high speed (where it doesn't change very rapidly when accelerating). When my old Fiat Coupé went off the clock, GPS told me it was doing 168mph. That's evidence enough for bragging rights.
They always have a table showing indicated speed versus actual speed in the acceleration tables, and invariably for every single car out there, the figures would differ. An indicated 60 would be an actual 58 and an indicated 120 would be an actual 110 - this is fairly consistent across the board.
But the McLaren F1? Indicated 60, actual 60; indicated 120, actual 120 etc.
Now I understand that it's no technical feat to get a speedo to read accurately (so much so that I've been told car manufacturers overread their speedos deliberately), but it was always one of those things which made it different from every other car out there.
And back on topic, I am of the unscientific opinion that my GPS-indicated speed is very accurate in a wide-open-area and at high speed (where it doesn't change very rapidly when accelerating). When my old Fiat Coupé went off the clock, GPS told me it was doing 168mph. That's evidence enough for bragging rights.
-DeaDLocK- said:
I always reading the Autocar road test for the McLaren F1.
They always have a table showing indicated speed versus actual speed in the acceleration tables, and invariably for every single car out there, the figures would differ. An indicated 60 would be an actual 58 and an indicated 120 would be an actual 110 - this is fairly consistent across the board.
But the McLaren F1? Indicated 60, actual 60; indicated 120, actual 120 etc.
Now I understand that it's no technical feat to get a speedo to read accurately (so much so that I've been told car manufacturers overread their speedos deliberately), but it was always one of those things which made it different from every other car out there.
And back on topic, I am of the unscientific opinion that my GPS-indicated speed is very accurate in a wide-open-area and at high speed (where it doesn't change very rapidly when accelerating). When my old Fiat Coupé went off the clock, GPS told me it was doing 168mph. That's evidence enough for bragging rights.
The F1 is probably a notable exception to the rule as if their aim was to make 'perfect car' then the speedo should be perfect too!
The other factor is that for that car more than pretty much any other they had a good idea that the end user wouldnt be messing with the wheels or tyres that much.
Gassing Station | Speed, Plod & the Law | Top of Page | What's New | My Stuff