GPS Speed reading versus car speedometer

PixelTrawler

Hi

Was out for a spin today with a friend who had his GPS with him. We mounted it on my windscreen and he was showing me how it can display basic details like car speed, trip distance etc.

The thing is it was reading a lower speed then the car was saying - roughly 6-8 kph slower.

The guy that was with me was saying that car speedometers are set higher then the actual speed. To avoid speeding..

You would imagine if the GPS can tell where we are so precisely it can work out speeds pretty well too, so is it true my cars being lying to me all this time...

Black_Knight

I think all cars have the speedo set a bit higher then they actually are. My GPS says im doing about 91Km/hr when my speedo says 100km/hr. I suppose its a bit comforting to know if just accidentally rise over 100 on my speedo that possibly I wont be caught for speeding.
Not supporting using a GPS as a speedometer because that certainly wont hold up in court.

And I do think the GPS is more precise than your Speedo. Its accurate to a few meters so im guessing its pretty accurate in judging speed

Rovi

Yes indeed.
There's a piece of EU legislation that's been linked to in previous posts on this subject, that (essentially) states that a speedometer may not under read true speed, but may over read up to... erm... 10%? (open to correction on that bit)

Most vehicles seem to read between 5-10% fast; ie. a true road speed of 100Km/h will show 105-110Km/h on the speedometer.

This also keeps the manufacturers from getting sued by irate motorists who could otherwise got nabbed going quicker than the speedo indicated.


edited to add:
Here we go- LINKY
Go to point 2.3.7.

As lifted from a post by Unkel in this thread.

unkel

Yup that's it

As an example: 100km/h real speed can be indicated on a speedometer as anything between 100km/h and 114km/h

The latter being 100km/h + 10% of 100km/h + 4km/h

Lex Luthor

The GPS is always going to be more accurate than a speedo as all manufacturers set the tolerance differently.

Not sure about using the GPS not standing up in court. Why wouldn't it. Who is to say your speedo is more accurate especially when you have so much differences between different cars.

At 120kmh on the GPS, the Toyota speedo reads 130 and the Alfa reads 123.

I go by the GPS speed if I have it in the car

Seth Brundle

Speedo reading is taken at the gearbox or differential - it is never 100% accurate as there are always changes to tyre diameter due to wear, suspension component wear, etc.

IanOCol

I worked for Garmin for two years and be warned: GPS readings are not more accurate than the speedo. Depending on the model/manufacturer you're using the difference can be up to 16kmh. Furthermore it depends on the area you're in. Inner city readings are less accurate, same goes for areas with loads of high trees etc.

Stay with speedo reading to be on the safe side

Tenzor07

Correct me if im wrong, but is'nt there a 30second delay on GPS readings due to information bouncing back and fourth from the reciever in the car and satelite? Which would obviously not give an accurate reading...

I have an electronic speedo hooked up to the ECU directly so gives an accurate speed reading, and i notice the difference between that and the main speedo dial.

Stephen

30 seconds? No way. Maybe up to 1 second.

cabrwab

More like 30ms. 30 Seconds would be useless you'd be passed your destinaton and your sat nav would say "you have reached your destination"!!
There is a wee bit of delay, but still stick with the speedo!

milltown

Tenzor07 wrote: »

I have an electronic speedo hooked up to the ECU directly so gives an accurate speed reading, and i notice the difference between that and the main speedo dial.

Huh?
What ECU? Where is it getting it's info from?

I checked my speedo vs. a friend's GPS a few months ago and the speedo was never more than 3kph out, which surprised both of us. There is a delay on the GPS (nothing like 30 seconds though) so it's not really suitable for measuring acceleration but at any kind of steady speed it's as accurate as makes no difference.

Sundy

Tenzor07 wrote: »

Correct me if im wrong, but is'nt there a 30second delay on GPS readings due to information bouncing back and fourth from the reciever in the car and satelite? Which would obviously not give an accurate reading...

Theres no bouncing anywhere, GPS signals only go in one direction.

Alun

milltown wrote: »

Huh?
What ECU? Where is it getting it's info from?

I'd assume from exactly the same place as the normal speedo, except it doesn't add in the 'fudge factor'. On my Focus, you can set the diagnostic display to show speed also, and it is different to what is displayed on the normal speedo and almost identical to what is shown on the GPS display. In other words, the only sensible explanation is that the percentage over-read is actually deliberately added by the cluster firmware. The analogue speedo in most cars these days is implemented using accurate stepper motors, so there's no inherent inaccuracy there to account for the difference.

Lex Luthor

kbannon wrote: »

Speedo reading is taken at the gearbox or differential - it is never 100% accurate as there are always changes to tyre diameter due to wear, suspension component wear, etc.

as kb says, the speedo can be affected by a lot of factors. The GPS is always accurate no matter what has changed with the car

Lex Luthor

IanOCol wrote: »

I worked for Garmin for two years and be warned: GPS readings are not more accurate than the speedo.

Thats why I bought a TomTom

Masada

the margin of error with GPS is very low though, it constantly calculates how long it takes you to get from point to point rather than where exactly your located., the standard cars speedo can be out by a fair bit as standard and especially if you have after market wheels fitted to the car.,

Esel

Sundy wrote: »

Theres no bouncing anywhere, GPS signals only go in one direction.

Sure about that?

Cantab.

esel wrote: »

Sure about that?

You're right. Multi-path propagation. The main reason why GPS is not as accurate in urban areas with tall buidings and lots of flat surfaces.

I'd say on the open road, with a clear view of the sky and no clouds, the GPS speed measurement will give you pretty good speed measurements (steady-state speed only).

dr.bollocko

I also think factors like weather seem to cause problems with the cars natural speedo reading and it tends to over estimate speed by figures approacing 10km/h. At least that is in my experience

Biro

For the GPS to give a true reading, you need to be holding the car at an exact speed on a level, straight stretch of road, and the longer you hold it at that speed, the more accurate the GPS will be. It stands to reason really. If your GPS is 10 metres out, and it registers you at point A at a specific time, and you only give it 100 metres to guage your speed, then you could potentially be 20m out (taking into consideration that you were 10m from point A in one direction and 10m from point B in the other, very unlikely!), then that's a 20% potential for margin of error. But if you travel 1km from point A on a straight, level road, then by point B you've decreased the margin of error to 2% maximum. And that's worst case. It'd probably only read your position about 2 metres off at either end, making that maybe 0.4% inaccurate.

craichoe

esel wrote: »

Sure about that?

Yes, hes 100% correct, its completely passive and works off signals coming from satellites, it transmits NOTHING back, because quite simply it wouldn't have the power to do so.

Satellite TV works in the same fashion, any uplink is provided via Dialup for internet and if it is two way it gets a whole lot more expensive.

Alun

Biro wrote: »

For the GPS to give a true reading, you need to be holding the car at an exact speed on a level, straight stretch of road, and the longer you hold it at that speed, the more accurate the GPS will be. It stands to reason really. If your GPS is 10 metres out, and it registers you at point A at a specific time, and you only give it 100 metres to guage your speed, then you could potentially be 20m out (taking into consideration that you were 10m from point A in one direction and 10m from point B in the other, very unlikely!), then that's a 20% potential for margin of error. But if you travel 1km from point A on a straight, level road, then by point B you've decreased the margin of error to 2% maximum. And that's worst case. It'd probably only read your position about 2 metres off at either end, making that maybe 0.4% inaccurate.

True, however GPS receivers use Doppler shift techniques to calculate speed, not reckoning based on distance and time, so your observation, although correct, is not relevant. Reception situations in which a good quality positional fix cannot be given, can still result in high quality speed calculations.

Biro

Alun wrote: »

True, however GPS receivers use Doppler shift techniques to calculate speed, not reckoning based on distance and time, so your observation, although correct, is not relevant. Reception situations in which a good quality positional fix cannot be given, can still result in high quality speed calculations.

I see.

Black_Knight

Biro wrote: »

For the GPS to give a true reading, you need to be holding the car at an exact speed on a level, straight stretch of road, and the longer you hold it at that speed, the more accurate the GPS will be. It stands to reason really. If your GPS is 10 metres out, and it registers you at point A at a specific time, and you only give it 100 metres to guage your speed, then you could potentially be 20m out (taking into consideration that you were 10m from point A in one direction and 10m from point B in the other, very unlikely!), then that's a 20% potential for margin of error. But if you travel 1km from point A on a straight, level road, then by point B you've decreased the margin of error to 2% maximum. And that's worst case. It'd probably only read your position about 2 metres off at either end, making that maybe 0.4% inaccurate.

Thats what I wanted to say, but as you can see, its long and complex. I agree though. At a constant speed on a straight road, a GPS should be accurate.

Steve

It's possible that the in-car GPS nav system manufacturers are also building in fudge factors to over-read the speed (for the same reason, so they won't be sued).
I've used a Garmin 48 handheld marine GPS and it always reads 5 -10 kph under the cars speedo.

They make no secret about how the speed is calculated, it's distance/time calculated twice a second and averaged over a number of previous readings. So if you're on a straight road at a steady speed for at least 10 seconds then the GPS is going to be very accurate. If you're changing speed on a twisty road then it'll have a hard time catching up.

Steve

Biro wrote: »

For the GPS to give a true reading, you need to be holding the car at an exact speed on a level, straight stretch of road, and the longer you hold it at that speed, the more accurate the GPS will be. It stands to reason really. If your GPS is 10 metres out, and it registers you at point A at a specific time, and you only give it 100 metres to guage your speed, then you could potentially be 20m out (taking into consideration that you were 10m from point A in one direction and 10m from point B in the other, very unlikely!), then that's a 20% potential for margin of error. But if you travel 1km from point A on a straight, level road, then by point B you've decreased the margin of error to 2% maximum. And that's worst case. It'd probably only read your position about 2 metres off at either end, making that maybe 0.4% inaccurate.

Agree in principle with what you are saying, however, the 10m error you refer to is inaccurate.
Most GPS systems are accurate to less than 1m.
There is a built in positional error that (they say) can be anywhere up to 100m, this error is only ever known by the US military and is to stop undesirable powers using GPS for missile guidance. It also floats, i.e. it's never the same from one day to the next.
The vehicle nav systems get around this by assuming that you are not driving in a field and place the vehicle on the nearest road.
The point is, that the error at position A is the same as the error at position B and so the distance between them is accurate.
I have seen this while driving and recording a track: if I drive the same road an hour later, the blip is perfectly on the track, a week later and it could be the same or 10 or 20 or 70 metres off depending on what the military setting is.

ednwireland

stevec wrote: »

Agree in principle with what you are saying, however, the 10m error you refer to is inaccurate.
Most GPS systems are accurate to less than 1m.
There is a built in positional error that (they say) can be anywhere up to 100m, this error is only ever known by the US military and is to stop undesirable powers using GPS for missile guidance. It also floats, i.e. it's never the same from one day to the next.
s.

i thought that was got rid of the inbuilt error when civilian equipment became more commonplace certainly when i'm out walking and look at gps its not a 100m out, btw my car is consistently 10km/h faster than the gps at 100km/h but hardly any difference at 80km/h and 50 km/h (but only when i've got a 5 to 7 sat lock anything less its varies .

this on wikipedia (make of that what you will)
On May 2, 2000 "Selective Availability" was discontinued as a result of the 1996 executive order, allowing users to receive a non-degraded signal globally.

Steve

ednwireland wrote: »

On May 2, 2000 "Selective Availability" was discontinued as a result of the 1996 executive order, allowing users to receive a non-degraded signal globally.

I know they turned it off during desert storm because most of the GI's were using civillian GPSs sent to them by their moms.
I thought they eventually turned it back on though - mustn't have.

towel401

IanOCol wrote: »

I worked for Garmin for two years and be warned: GPS readings are not more accurate than the speedo. Depending on the model/manufacturer you're using the difference can be up to 16kmh. Furthermore it depends on the area you're in. Inner city readings are less accurate, same goes for areas with loads of high trees etc.

Stay with speedo reading to be on the safe side

how is the speedo on the ordinary bog standard yellow etrex? i tried this on an open road