How many times have you heard at the end of a race someone complaining that the GPS was showing them a different distance from the official race distance? Maybe it just happened to you. It was always something more or less, and each time the error was attributed to those who had measured the route or to the GPSs themselves. Because if they all start from the same point and arrive at the same point, how is it possible that there are certain differences, sometimes even substantial differences?
Finally, it may also happen that people attribute these differences to different GPS manufacturers, and then endless discussions arise about which is more reliable and which is not.
The truth, as usual, lies somewhat in all these arguments but also elsewhere.
Do GPSs lie?
Let’s start with the first defendant: are the race routes really as long as the organizations say they are? Since these are official races, the answer is yes: distances are certified and measured with accurate instruments. The tracks, however, run on wide road sections, and the organizers draw an ideal line that actually measures what is claimed, but measured according to certain trajectories. What does it mean? It means that if you run outside this track you may happen to make more or less meters of road. In fact, if all competitors proceeded along the track, one would have to run in single file: not very practical, is it?
Then it’s the GPS! Yes and no. First: the GPS signal can be disturbed by trees and buildings and thus provide inaccurate readings. Assuming, however, that they work perfectly and are not disturbed, why are there still discrepancies between the “physical” data and the data measured by GPS? Essentially for two reasons.
- GPS are accurate up to a point. This is not out of malice on the part of the manufacturers but because the demands of miniaturization and low power consumption impose some sacrifices, such as detection accuracy. There are much more accurate GPSs on the market, and they are the ones geographers and seismologists use to measure earth crust displacements, but they have problems: they must be installed on fixed locations (which a runner, by definition, is not), they cost $5 to $8,000, they are the size of a book, and they weigh as much. The good news is that with the passage of time and progress we will have increasingly sophisticated and accurate systems on our wrist. Additional important note: the top models from different manufacturers (like triathlon models) already have more accurate GPS but also cost more. But in the near future they will also be available at lower prices to the average consumer with fewer needs.
- A wrist-mounted GPS detects your location by discrete, that is, non-continuous points. Imagine that I detect it now and in a fraction of a second when you are already somewhere else on the map. The track he will get is made up of points that often do not even stand on the same curve or line but discard a little to the right and a little to the left. To provide the most reliable reading of the distance then, an algorithm takes care of unifying this data so that the union of the segments representing your run stand on a single track. Which inevitably is not exactly what you have traveled but an approximation of it. An average, let’s say. What happens when you turn a corner? It may happen that the algorithm does not consider your run to the point where you actually turned the corner but “cuts it off,” thus providing you with a shorter distance. A few tens of centimeters, it is understood, which, however, when added up in the end can give deviations of meters. That is where the error may come from, not so much as the dissimilarity of the reading.
Is there a solution?
At present it is not possible to obtain centimeter-accurate readings but certainly very valid approximations.
To correct for the “necessary” error of the discrete point interpolation algorithm, some models also consider readings from the internal gyroscope, which more accurately detects changes in direction and skidding. In short, the more “dirty” the data provided by the GPS, the more useful it is to introduce the gyroscope correction that can confirm or not that you were really running like a drunk in that stretch. Were you really doing that? Come on!
As mentioned above, however, not all models also use gyroscopes, or at least not the low-end ones.
The cheapest and most effective solution, however, is not to take it on, knowing that the mistake is there, it is more or less small, and that you are also running for other reasons, like having fun.
(Via PodiumRunner)
Monitorare il cuore: è meglio la fascia o il sensore ottico?