GPS vs. Smartphone Accuracy Test

GPS vs. Smartphone Accuracy Test

two years ago there was a discussion on the single tracks forums about whether smartphone GPS apps were as accurate as dedicated GPS devices from companies like Garmin so we decided to set up a test and see for ourselves to test for accuracy we would need a controlled course of known length and elevation a local quarter-mile running track offered the perfect place to conduct our test our first batch of GPS devices and smartphone apps included two iPhones and Android tablet a Windows Phone 3 GPS units two GPS watches and a GPS enabled a helmet camera we strapped all 10 devices onto a mountain bike and rode 10 laps around the track all but one of the devices reported distances within 3% of the actual length of the test course and all about one device over reported the distance the iPhone and Android tablet running Strava were the most accurate in this test followed by the Garmin Forerunner wristwatch and the Magellan cyclo 505 not surprisingly the Garmin VirB helmet camera came in near the back of the pack along with an older GPS device meant for hiking if you like to create maps based on GPS data collected during your adventures it's important that the GPS tracks generated by your device match up with the actual route you travel as you can see from the various maps we generated during our tests some devices like the Nokia Lumina smartphone produced tighter tracks than others like the iPhone or are hiking GPS in addition to testing distance accuracy we also wanted to know how well each device tracks elevation particularly elevation gain and loss during a ride since we test it on a flat running track we expected the most accurate devices to report close to zero overall elevation change once again the Strava app on an iPhone and the Android tablet reported the narrowest elevation change followed by the Garmin Edge 517 in our test the garmin phoenix 2 reported over a hundred feet of elevation change along the flat track despite featuring a barrel altimeter for improved elevation accuracy after our first round of testing a few readers suggested adding a cyclo computer to the mix to use as a baseline for our results we chose a simple cat I've lo7 which measures the number of revolutions the front wheel makes to calculate speed and distance we calibrated the cat I've lo7 by running a piece of tape along the circumference of the front wheel then measuring the tape this time we ran three one-mile tests to track not only accuracy but also consistency even though we used a fresh crop of GPS units and smartphones the result from this test were similar to those from our first test the smartphones running Strava and endomondo apps were the most accurate and all devices were within 3% of the actual course distance not only that all the dedicated GPS units were completely consistent from test to test our ciclo computer didn't fare as well in terms of accuracy but we chalked that up – calibration issues rather than a fundamental issue with the way the unit works all right the smartphones along with the Garmin Edge 520 reported the least amount of elevation change along our flat course keep in mind however that smartphone apps generally do not report raw GPS data instead they process GPS data in the cloud to help them improve accuracy as you can see from our test results there's no such thing as perfect accuracy when it comes to GPS devices or smartphones though some are slightly more accurate than others at least within a controlled environment so as always take your activity data with a grain of salt and realize that it will never be a hundred percent accurate no matter which device you use

10 thoughts on “GPS vs. Smartphone Accuracy Test”

  1. I broke it down into 3 parts, innately. Controlled course, elevation change & allied health (natural & relative) #lifeasiknewit

  2. Regarding the dedicated GPS receivers (such as the Garmin 60csx), this test is not entirely apples to apples. The Garmin, by default track points record as the heading changes. This method is intended for for navigation. Based on the tracks recorded by the other devices, it appears track points are recorded every second or two. All devices should have had identical recording interval settings (Garmin can go as high as 1 second). Poor comparison.

  3. I have done the same kind of analysis and data. Statistically, the iPad Pro was the most accurate even over GPSes.

  4. I find GPS on iPhone and Android phone, especially iPhone lack accurate and precise location. It misses by more than 100 meters at times. I would like to have a GPS that only misses less than 0.1 meter.

  5. Interesting i would like to see the results from walking around because you are going slower i assume it would be way more accurate

  6. I appreciate this test. I suggest to make another test based on chipset.

    For example:
    Iphone chipset
    Android chipset (mediatek, snapdragon, Kyro, etc.)
    Dedicated GPS device

  7. Did you guys make this a while ago? I could have sworn I saw this before but I'm glad I came across it again. Been meaning to share it.

Leave a Reply

Your email address will not be published. Required fields are marked *