Is Garmin/GPS distance measurement inaccurate?

[I like Skol]

I only ask because for years I have been using traditional cycle computers with a wheel sensor. These have been calibrated against repeated journeys using a number of methods...

1. Wheel circumference, measured by rolling the bike forward in a straight line for 5 revolutions. This is pretty accurate and gives a good starting point.
2. Ride a known distance multiple times and compare to the computer reading. This known distance is measured using multiple sources to get a best guess figure. I use various online route planners and mapping sources as well as comparing against distances suggested by a TomTom unit. If all the sources are slightly different I will use the average. If one source gives me a figure that is vastly different to the rest then I will use the modal average.

Once the wheel circumference is tweaked to calibrate the computer against the mapped data my computers are consistently accurate and give repeatable and comparable results over a regular route (my commute which never changes) to within 0.5%, trip to trip and bike to bike.

I have recently bought a Garmin Edge and now find my trip distances are consistently recorded around 2.5-3.0% lower than my other bikes. Today I went out with my son, me riding with GPS and him on my other road bike with IMO a calibrated traditional computer. We rode together along exactly the same route yet he has recorded 14.00 miles and my GPS is showing 13.63 miles

I don't believe the GPS, am I wrong?

 

[Supersuperleeds]

I'm not 100% sure, but I think it is to do with the fact that GPS doesn't taker into account the elevation effect on distance, whereas the wheel sensor does.

 

[Crackle]

I bought myself a speed/cadence sensor to make sure that the GPS is accurate and to compensate for climbing and poor satellite cover. I might get one for the mtn bike too as that is very prone to losing the signal and telling me I've stopped when in fact I'm still crawling up a woody slope just faster then the surrounding snails.

 

[dan_bo]

To confuse matters more, get strava on your phone and see what that says.

 

[Crackle]

Oh and before I owned an 810 I used to ride with a cateye and a Garmin FR305 and they pretty much agreed. Now I've got the 810 I've taken the Cateye off. Last summer I did 67 miles on another bike with an older cateye on and my tyre size must have been off in the cateye as it recorded 70miles. It doesn't need to be off by much to be wrong. Likewise the 810 doesn't seem to be as accurate as the fr305 in its tracking, which is why I bought the speed sensor but it does seem to get the same distance as the 305 even when the track doesn't seem to be as accurate. So in short, I don't know but 2 or 3% over that distance is quite a lot.

 

[I like Skol]

I'm not 100% sure, but I think it is to do with the fact that GPS doesn't taker into account the elevation effect on distance, whereas the wheel sensor does.

This is a good answer. I thought the GPS would do it's working 3 dimensionally but maybe not. I will have to google this.

 

[ColinJ]

I'm not 100% sure, but I think it is to do with the fact that GPS doesn't taker into account the elevation effect on distance, whereas the wheel sensor does.

For any sensible numbers, that effect is neglible.

For example, I do a local audax ride which is EXTREMELY hilly - over 2,500 m of climbing in about 102 kms of riding. The route is roughly 1/3 flattish. 1/3 uphill, and 1/3 downhill. If you enlist Pythagoras you will see that the uphill bits measuring 34 km on a map (or by your theory, the GPS) would actually be 34.091 km measured along the road surface. I don't think that we need to worry about a 182 m error in 102 kms! (Ok, we would if an official record were involved, otherwise not.)

I have done tens of thousands of kms/miles of rides using my old Garmin Etrex. I plot the routes on digital OS maps so I know how long they are before I even ride them. I find that the distance recorded by the GPS is so close to what I expect that I completely trust it now.

 

[jefmcg]

GPS records your position in space every x seconds (2 or 3 dimensions, depending on the device).

It calculates distance by "guessing" the route you took between the two points. Mostly simply using Pythagorus and ignoring elevation, assuming your journey is a series of straight lines between the points. More sophisticatedly using elevation and estimating a curve that is closer to your actual line.

So it's always going to be a little off. The more points it records, the more accurate it will be but the shorter your battery life.

 

[User6179]

If you check your recorded route on the map depending what device you use you may see it cuts corners especially when they are sharp turns at speed so whatever speed/ distance it records you have most likely went further at a slightly higher average speed .

 

[Sharky]

Would be interesting to try a few controlled experiments.
1) Try with your son's bike having both devices on it. This would allow for any difference in riding style.
2) Try riding each other's bikes over measured distances. Difference in rider weight might affect calibrations.
3) Ride a 10 mile TT on a recognised course and see how close to the 10 miles you register.

Good luck

 

[potsy]

Daft question but are you making sure the gps has 'locked on' before you set off?

Mine can take up to a 1/4 mile sometimes, and if you've already pressed start it will not count that distance.

 

[DaveReading]

I'm not 100% sure, but I think it is to do with the fact that GPS doesn't taker into account the elevation effect on distance, whereas the wheel sensor does.

Unless you're in the Alps, elevation has a negligible effect on distance.

Even if your entire journey was up a 1 in 4 incline, it would only add about 3% to the map distance.

 

[ColinJ]

If you check your recorded route on the map depending what device you use you may see it cuts corners especially when they are sharp turns at speed so whatever speed/ distance it records you have most likely went further at a slightly higher average speed .

That is the most likely source of significant error. I have just looked at a couple of my tracklogs and spotted this ...

You can see that generally the GPS is doing a good job of recording my progress along those lanes but at Scout Road it 'took a shortcut'. That is a fast descent with a steep wooded hillside to my left. It potentially would have lost sight of a satellite or two going down there and it clearly took a guess at where I was going and then corrected itself.

 

[si_c]

I'm not certain as to where the assumption of GPS not recording elevation comes from as calculations from GPS do provide an elevation component. The issue is in part that it can be inaccurate especially when changing rapidly, this comes more to the forefront in applications such as flight or skydiving. In such circumstances it is important to have barometric readings to provide elevation, which of course some (I'm not sure if yours is included) GPS devices also include.

As @potsy says though, the greater likelihood is that either you are setting off before the GPS has acquired a proper lock or possibly the recording interval is set too high and you are "missing" small sections of the ride. For example when recording on the Strava app I often find that roundabouts are recorded as a straight line and some corners are recorded as a straight line, for GPS applications this will result in an under recording of distance (See Reimann sum for how this works, although that evens out over time as a result of directionality not applied in this case).

 

[User6179]

That is the most likely source of significant error. I have just looked at a couple of my tracklogs and spotted this ...

View attachment 336104
You can see that generally the GPS is doing a good job of recording my progress along those lanes but at Scout Road it 'took a shortcut'. That is a fast descent with a steep wooded hillside to my left. It potentially would have lost sight of a satellite or two going down there and it clearly took a guess at where I was going and then corrected itself.

I get a few like that per route , Endomondo was quite bad , Strava is better on my phone (HTC) but I see it a lot on iphones .