Data freaks: careful with that GPS cyclometer!

An ugly secret exposed about GPS measurement: it’s not always accurate. Although this article is about GPS watches for runners, it certainly applies to cyclists. The article can be found here: http://www.nytimes.com/2011/12/20/health/nutrition/gps-watches-may-not-track-runs-accurately.html?src=recg

For years Roger and I have been using handheld Garmin GPS devices for cycling, and recently I’ve been using an uber-expensive Garmin 800. We’ve always had these devices in addition to a regular cyclometer and we have noted the discrepancy in mileage between them. We have been careful to accurately measure the rollout distance at a specific tire pressure on a weighted bike in order to calibrate our cyclometers. One possible confounding error is the reliability of the wireless signal from the transmitter to the computer head. But in general we’re fairly confident in the accuracy of the cyclometers. Looking in detail at the path created by a series of GPS track points one can see that a track around any curve is simply a set of line segments. So, to us it has been rather amazing how close the GPS-measured distance usually is to that of the cyclometer. Typically our experience over a 50 mile route might be a discrepancy of one- to five-tenths of a mile. That’s pretty damn good. On the other hand, we’ve had some whoppers too. We did a ride in Austria this past summer and the difference was about three miles out of about 60, or roughly 5%! There didn’t seem to be any obvious reason why we would have such a difference. One day on our tour we were cycling near Hallstadt, which sits on the edge of a lake underneath a huge granite cliff. Our GPS signal rather abruptly severely degraded (as did that of other cyclists) and we had a track which had us cycling in weird zigzags over water! Surprisingly, the measured distance for the day wasn’t too far off between the Garmin and the Shimano cyclometer.
My recommendation is that you have a choice, use a GPS device which measures distance using a wheel sensor rather than relying on the GPS signal alone. For example, the Garmin 800 and 500 allow either. But the new Garmin 200 measures distance only by the GPS signal. Of course, unless you’re a data fanatic it really doesn’t matter, does it? And, if you have serious aspirations for ‘improving’ your cycling (i.e. going faster), then distance isn’t an important training datum anyway. Of course, for bragging rights it’s EFI!

Advertisements