PDA

View Full Version : Which is correct..odometer or GPS



stevo68
15th July 2008, 12:34 PM
G'day All,

Been bugging me this has. What would be the correct guage to go by between km's travelled on odometer versus km's travelled on GPS...if that makes sense. Reason I ask is that when I fill up, I set odometer and GPS back to 0 kms. Within a couple of km's the GPS is showing higher kms travelled compared to the odometer. At next fill up the gap is around 10-12kms. So which is correct?

Regards

Stevo

Utemad
15th July 2008, 12:53 PM
The GPS will be correct everytime short of any long gaps in signal reception.

If the odometer is showing less kms travelled than the GPS then how accurate is the speedo?
If the odometer and speedo are connected then your speedo will be under reading. Which means you might get a speeding ticket.

Usually it is the other way around. Your odometer would read more than the GPS as that is how they are set up at the factory.

Do you have oversize tyres on your car?

weeds
15th July 2008, 01:16 PM
yep normally its the other way around........well it is with the defender running standard tyres

rovercare
15th July 2008, 01:26 PM
. At next fill up the gap is around 10-12kms.

Regards

Stevo

<2% error, do you really give a toss?

GPS wins every time

stevo68
15th July 2008, 02:08 PM
The GPS will be correct everytime short of any long gaps in signal reception.

If the odometer is showing less kms travelled than the GPS then how accurate is the speedo?
If the odometer and speedo are connected then your speedo will be under reading. Which means you might get a speeding ticket.

Usually it is the other way around. Your odometer would read more than the GPS as that is how they are set up at the factory.

Do you have oversize tyres on your car? Thanks for that, the speedo and km/hr on the GPS seem pretty accurate, have checked that on numerous occasions. Have just filled up today and going to see what the difference is this time. As for tyres, they are stock standard at the moment.


Quote:
Originally Posted by stevo68 https://www.aulro.com/afvb/images/imported/2016/08/768.jpg (http://www.aulro.com/afvb/gps/59844-correct-odometer-gps.html#post778996)
. At next fill up the gap is around 10-12kms.

Regards

Stevo

<2% error, do you really give a toss?

GPS wins every time Its not so much as giving a toss, it is more out of interest. I have little to no knowledge on these things :angel: hence posting the question. I would have thought that a difference of 10-12kms would have been unusual?

Regards

Stevo

JDNSW
15th July 2008, 04:06 PM
There are two questions involved - the accuracy of the odometer, and the accuracy of the GPS.

1. Odometer. The odometer should be consistent, that is, the same percentage error unless either the driving style involves a lot of wheelspin, or the drive member is slipping (which is quite possible on most Landrovers!). As stated above, most cars are set so the speedometer, and hence the odometer, read higher than the correct figure, because the legal requirement allows over but not underreading. Reports on this forum suggest a typical Landrover figure is 8%. Non standard tyre size will change this, oversize tyres reducing the over reading and potentially even going to underreading. Tyre pressure will have almost no effect - particularly with a radial tyre, pressure does not change the circumference of the tread.

2. GPS. It needs to be remembered that a GPS measures position, and the distance travelled (as well as the speed) is a derived function. In particular, the distance travelled, if it follows the jitter* in the position, can easily result in a longer GPS distance. Now the software certainly applies some smoothing to eliminate the problem, but if excessive smoothing is applied, it can round off real corners. I have no feel for how effective the actual algorithm used is in practice, but I find it easy to envision it over reading. Some software may have options as to how much smoothing is applied, and if this is available in your case, try experimenting with it.

*jitter is the apparent change in position due to errors in determination of position. How large it is depends mainly on the geometry of the satellites used and any poor reception, for example due to a partly shadowed antenna.

It might also be useful to compare GPS and odometer distances with the map distances (although I can think of a number of places where these are wrong - for example a couple of years ago I checked the map distance from Yass to Cowra - three maps, three different figures by up to over 10km; they can't all be right!)

John

p38arover
15th July 2008, 04:40 PM
I have to wonder about traversing hilly country.

Let's assume it's a dead straight but hilly road.

An odometer will measure the distance travelled by the wheels and thus the car as it goes up and down hills.

I assume the GPS measure distance from point A to point B but will it measue the distance travelled going up and down.

An extreme example would be driving up and down a road that looks like sine wave in cross-section.
https://www.aulro.com/afvb/images/imported/2008/07/756.jpg

Where's RMP when you need him? :D

Utemad
15th July 2008, 05:20 PM
A GPS position is not only latitude and longitude but altitude as well. So it knows if you are going up or down a hill. Some units are better than others for altitude (longs and lats too).

A GPS track is just a series of points. You can change your GPS to create points by time or distance. You can set the time or the distance.
In older units where track points are limited you would often get dodgy tracks as to get a longer track you sacrificed detail.
Newer units have bigger memories so you can use more points for a given track.
Depends on what you are doing as to what is the best choice.

Utemad
15th July 2008, 05:27 PM
This is interesting
Testing the Accuracy of Autonomous GPS in Ground Speed Measurement (http://adsabs.harvard.edu/abs/2005JApSc...5.1518A)


A hand-held Garmin eTrex Venture Global Positioning System (GPS) receiver operated in autonomous mode was used to investigate the accuracy of autonomous GPS in measuring ground speed. The accuracy of GPS-derived ground speed was tested by collecting and storing in a laptop computer, every 10 sec, ground speed data from the receiver at eight different ground speeds of a vehicle. The targeted ground speeds involved were 5, 10, 15, 20, 25, 30, 40 and 50 km h-1. The vehicle's speedometer was used as an initial indicator of the ground speed. A pulse transmitter was utilized to detect the frequency, for every 10 sec, of a magnet fixed on one of the vehicle's wheels and passing by the transmitter every full rotation of the wheel. Data from the pulse transmitter was acquired and stored in a data logger. Given that the wheel's circumference was known, frequencies were used to calculate the ground speeds which were used as a reference. By the use of matching times of GPS and data logger data records, GPS-derived ground speeds were compared against the reference speed measurements. For nearly 600 data points comprising the data set for the eight test ground speeds, it was found that the error in GPS-derived ground speeds was, on the average, 1.27 km h-1 (less than 7%.) However, high fluctuations in vehicle's ground speed when, for example, turning greatly impacted this error. An error of -80.16% was produced due to a drop of vehicle's ground speed from 18.65 to 11.19 km h-1. Results of steady state ground speed analysis revealed that the average error was less than 1 km h-1 (less than 5.3%), except for the 15 km h-1 data set where the average error reached 1.72 km h-1 (9.92%.

stevo68
15th July 2008, 05:39 PM
Awesome information folks, have learnt something today :D. I went for a drive this arvo and on the way back, thought I would check to see how things were "travelling" ;). This time odometer was 1km more than GPS....then I realised that it had searched for about a k or 2 due to heavy cloud covering.

Regards

Stevo

vnx205
15th July 2008, 05:43 PM
Results of steady state ground speed analysis revealed that the average error was less than 1 km h-1 (less than 5.3%), except for the 15 km h-1 data set where the average error reached 1.72 km h-1 (9.92%

So does that mean that the best result for the GPS is about twice the discrepancy that Stevo68 noticed between the odometer and the GPS?

I realise this test was speed and Stevo68 was measuring distance, but it does make you wonder about the GPS.

I was a bit surprised that Stevo68's difference was as small as 2% and even more surprised at the fact that the odometer gave the smaller figure.

I would have expected, as JDNSW suggested, to learn that it was about 8% and in the other direction. I have standard tyres on my Defender and I think the speed is out by 10% at 100kmh and the distance about 6% out.

Utemad
15th July 2008, 05:52 PM
Going by their graphs and conclusion it appears that the error in GPS speed is around 1.2km/h (they say 1.27km/h) regardless of vehicle speed.
So 1.2km/h at 25km/h and 1.2km/h at 50km/h. So not the same sort of error as a speedo.

101RRS
15th July 2008, 06:03 PM
then I realised that it had searched for about a k or 2 due to heavy cloud covering.


Then you must have an old GPS because this is not an issue for modern GPS systems.

Garry

RonMcGr
15th July 2008, 06:21 PM
GPS of course.
Satellite, accurate as a slide rule!
I have never seen an accurate speedo or odometer in my life!
Mind you, our 1925 Citroen Cloverleaf was not too bad, it went so slow you could measure the road :D

stevo68
15th July 2008, 06:40 PM
Quote:
Originally Posted by stevo68 https://www.aulro.com/afvb/images/imported/2016/08/768.jpg (http://www.aulro.com/afvb/gps/59844-correct-odometer-gps-post779163.html#post779163)
then I realised that it had searched for about a k or 2 due to heavy cloud covering.

Then you must have an old GPS because this is not an issue for modern GPS systems.

Garry Nope its a modern Garmin, though more likely a km of travelling. When it storming up here in the hinterland our Austar also gets affected,

Regards

Stevo

waynep
15th July 2008, 07:04 PM
I'd swear in court by a spot speed measurement from GPS, ( which uses a rubidium clock ) but not a distance measurement. The reason being you can never be 100% certain your GPS was locked on a fix for the whole time.
So I do it this way - measure the difference in percentage ( plus or minus) between spot GPS speed and car speedo - use the car odometer for distance measurements but factor same percentage difference into the odometer reading.

victa125
15th July 2008, 08:26 PM
Seems to me Stevo that both have variables, Tyre wear, pressures, wheel speed, etc, would affect the odometer which if ever had been calibrated properly, less than the actual speedo. The hilly thing reception etc means more calculations for the GPS. Should have no effect on odo, but what happens if we chuck heaps of corners in, the odometer can only work on a mean average of actual wheel distance traveled. Without any slippage that is. should both be exactly the same I would be more concerned.