Connected GPS Pace and Distance is Inaccurate

[DanielMena9]

Went on a run today with my Charge 2 on one wrist and Charge HR on my other wrist. Also had RunKeeper to measure my distance. I started my run with about the same number of steps on both FIT devices. 

After the run, my Charge HR had almost 2K more steps than my Charge 2. In addition, my GPS map on FitBit matched my Runkeeper map which is about a 3.5 mile run BUT FitBit pegged the same route as 1.73 miles >.> 

Why doesn't this device work as advertised? I really like the design and feel of Charge 2 but if it can't do its job accurately, all is for not and I will be shorting FIT as it appears to be a widespread issue (I have maps and photos if someone from FIT wants the data).

Moderator Edit: Edited thread title.

[Glennbest]

Thanks Colin,

The problem occurs when using connected gps to track exercise, which is meant to calculate distance via gps signal. I don't expect the tracker to accurately calculate distance when using stride length.

Can you confirm you get inaccurate distance using the blaze with connected gps? (That is starting excercise tracking using the blaze).

[Glennbest]

Hi Rich,

Bobby is calculating percentage and multiplication correctly.

[Glennbest]

Hi Kenny,

They have been aware of the issue for a very long time, I think from early last year when the blaze was released but either the don't know how to fix it or the fix requires a product recall. Those of us who care about the issue aren't in their demographic, I suggest asking for a refund and buying a garmin.

[Rich_Laue]

I didnt say he wasn't,  i do see it is well within the accuracy that fitbit states.

I also have not been able to confirm yhat the GPS is used in the Walk mode, it is used in the Run mode.

I have been able to confirm that while in the walk mode with GPS connected,  i can deliberately physically  change my the distance i take with each step and the repprted distance does change in tedponse.

8 can also walk my normal stride,but change the stride  setting in app and the reported distance will change in response, even though the GPS is showing connected 

[hawkeye1952]

I do not think they are going to do a recall. The reason is that. The moment they do that. Then their stock will take another hit. It already taking a hit because of poor sales. Whit that said. They better get off dead center and get it fix. I use a lot of different programs. They all are with in .01 of each other. Fitbit been off as much as .10 miles not good.

[Colinm39]

@GlennbestThe example I posted earlier is using Walking Exercise with the Blaze and overstating distance by 12%. The Surges are using their own GPS.  I have many more from walking around our neighbourhood and the tracking path is consistent.. No jumping out and over houses. Around my area where I live I have line of sight of a Telstra Mobile tower. My Sony Android XP V7.0 hasn't missed a beat with any syncing or connection.  No forests etc.. The GPS signal has not faulted on MayMyRun or just phone connected GPS.  I will go back and check my records

When I had the earlier Sony which had Android V4 and V5 was hopeless and erratic on connection with many errors. I have had the new Android pair with all of the Fitbit devices with no issues.  So the last few months I have eliminated the phone as an issue.  I don't run.

My real issue is the loss of distance for domestics.  Just using the Fitbit for everyday use. The Fitbit One is spot on with the Auto distance of 71.4 cm for walking. I'm experimenting with stride change but the Blaze has a mind of its own, unless I'm having a planned walk at 5+ kmh and the distance is OK.  I have a closed case on this problem. With the Engineers...  The calculation is distance/steps = calculated stride.  Ignoring any missed steps...

In the example below I only had the Fitbit One on for the later part of the day.

[Kenny54]

Hi Glennbest,

Yes I was thinking strongly about talking to the shop where I purchased the Charge 2 with a view to probably having to negiotate a return and to sweenten the pot paying out further money to go the Garmin route. Looking at their forum to see if the "Forerunner 35" is a more reliable product, any thoughts?

Kenny

[hawkeye1952]

I thin k I just found the problem I do not think they can fix it.


I believe the problem is that the fidit connect GPS connects to the A-gps side of your phone. My phone has both. If this is true. Then that would explain why we all have such large difference in our readings. You lose connect with your network then you lose the connect gps.





. A GPS device communicates with 4 or more satellites to determine its exact location coordinates (latitude and longitude) anywhere on Earth. It works in any weather as long as the device has a clear line of sight to the satellites.



A-GPS stands for Assisted Global Positioning System. While it works on the same principles as a GPS (explained below), the difference here is that it gets the information from the satellites by using network resources e.g. mobile network, also called assistant servers.

George Howell Jr

[Glennbest]

Hi Kenny,

The distance error occurs when using Connected gps (as the title of this forum suggests) so any device that uses its own gps signal will be immune from the problem.

Something like garmin forerunner 35 or vivosmart hr+.

I downloaded gps data from two excercises from Fitbit once. The first was recorded via the app and second via charge 2. I used the data to calculate distance between gps points each second and compared that to the distance calculated by Fitbit each second. I found the distance I calculated for each second for the first excercise (using the app) was very close to that calculated by Fitbit. I found the distance I calculated for each second for the second excercise (using charge 2) was about 0.3m less per second compared to the distance calculated by Fitbit. So the charge 2 is adding 0.3m per second to the distance. That's 18m per minute. When I multiplied this by the duration of the excercise I got the same distance as the discrepancy I regularly experience.

Can I get everyone to try this;

Multiply your excercise time (in seconds) by 0.3m and let me know if the distance equals the discrepancy between a connected gps and mobile run (app) excercise.

Thanks in advance.

[Colinm39]

@GlennbestMy information is about the Blaze and I'm posting here because I'm trying to find out as much as I can about Fitbit's Connected GPS on the Charge HR 2

In the measured walk adjacent to our athletic track, (the very last image) I get over reporting of 130m to 180m of the 1208m walking track. That is about 11-15%.  6-12m increase/ minute depending on my speed of walking. normally 5.2+kmh (walking for exercise pace....)  Your will notice I had trouble with one of the Surge (direct GPS)  tests and started with Quick Start.

In the walks around where I live the Telstra tower, with 4GX is  typically 300-500m in line of site on top of our water reservoir.  Depending on my walking path.

The examples are self explanatory but my GPS has always tracked diagrammatically except the one image below where the signal went crazy...

The car test were to check the plotting, not for distance. As you can see, they plotted and because there was no movement detected on the Blaze,  It's 3.1 km to where I park. So another factor.

[Kobalto]

Hi,

I've been experience the same issues regarding inaccurate measurements. So I tryied the following using TCX data downloaded: I put the information in an excel, and calculated the distance between each point. To do this, we need to convert the GPS coordinates (in degrees) to meters. One degree of latitude is close to 111,122 m, so you can transform the distance between two points by multiplying the difference of the latitudes by 111,122. When dealing with longitude, this is more complicated, as one degree is not always the same. In the equator, one degree its 111,122 m, but in other latitudes its different. For instance, I live in Santiago, Chile, where 1 degree of longitude equals 92,791 m.

So thanks to excel, it’s not difficult to compare the distance between two GPS coordinates and extend it over the whole track, using the latitude and longitude difference between one point and the next one, multiplying by the correct factor, and then using Pythagoras to calculate the distance between the two points. Using this method, I could calculate almost the same total distance for a track that I got with Runkeeper or even the Fitbit app (used without the Charge 2). But as expected, and as we all know, when comparing this method with the figures given by the Charge 2 (field “ns1:DistanceMeters2” in the TCX file), the difference in a track was inaccuarate in more than 10% (the Charge 2 always showing a distance larger than real).

But I also notice that the error was not always the same, and followed some patterns: In the same track, while going north or south, the error was small, but while going east or west, the error was bigger. So, I tried changinf the factors, and used the 111,122 factor instead of the 92,791 for longitude. This is obviously mistaken, but to my surprise it worked very consistently to what the Charge 2 was showing!!!! The error was reduced to 1% or less!

I already made the same calculations for three different runs, one 10k, an 18k and 20k, and I got the same results. It is like nobody thought the Charge 2 that longitude coordinates are not dealt in the same way as latitude coordinates, and that a correction should be made if you don’t run exactly over the equator.

Below there are graphs for the 10k and 20k runs. I don’t know if this was already discovered by the development guys of fitbit, but looks like this shouldn't be something difficult to correct.

[Glennbest]

Hi kobalto,

I'm near Melbourne OZ (so Southern Hemisphere like yourself) and I also find north south distances accurate but east west inaccurate. I also did similar comparison to yourself using excel and found an average 0.3m discrepancy per second.

If the problem is easy to fix they would have done it by now.

[TurtleSpeed]

I am also getting incorrect distance when using GPU Connect via my Charge 2. Different from others, though, FitBit GPU Connect is reporting more KMs. For example, yestersday, I ran for 3.5 KMs but FitBit is saying that I ran for 3.91 KMs. How strange!

[Kobalto]

Hi TurtleSpeed,

I believe I have the same issue, Charge 2 showing 10% or more distance than the real one, and also showing a pace that is faster than my real pace when using GPS Connected. Here its a graph with a 20k run (according to the Charge 2 it was 22.18 Km, but according to Runkeeper it was 20.04). The blue line is the % of error comparing what the charge 2 is showing as pace and a calculation of the pace based in the same data of the charge 2, but calculated manually as explained in my previous post. The orange line is simmilar but shows the error if I intentionally use a wrong calculation method, using longitude coordinates as it they were latitude, also as explained in my previous post. It's suprises me how the error is reduced.

Please notice that in the center, the error of the blue line is almost cero. As you can see in the map, it's the part where I ran north-south, while the east-west parts have errors of 15% or more.

Glennbest, it good to know that you also experience more accurancy in north-south direction, it would be good to know if someone from the northern hemisphere has simmilar findings.

It also shows a big discrepancy in the begining, but I think its likely that the GPS wasn't well fixed yet.

[Rollerskier]

Well, in my case it's also phone related.  I find iPhone 6+ fairly accurate.  If I start the run on the charge 2 itself or I start on the app it's comes out about the same.  I have been test walking in a track that's like square of about 3,18 km distance measured by my car driving 5 times and taking the average. Results Fitbit app and tracker varies in the range 3,15, 3,16 and  3,17 km.  Less than 1% deviation between min and max distance.

However, doing the same with iphone7 and you have a lot of wobbling and large deviations up to 4,5 km. Apple has proposed today to replace the GPS unit in my IPhone 7 today after analyzing system log files. I will test again when it's fixed or the phone is replaced in 2-3 weeks from now.

I am staying in Oslo, Norway in Europe.  60 degrees north. 

[jvcindian]

I am on my second charge 2 and still have same problem. It doesn't count steps correct!!

[Cotsa]

HI, I have the same issue when running, to you work out how to fix this?  Thanks

[Glennbest]

Hi Cotsa,

There is no fix. If you are experiencing this problem Fitbit offers no solution. Obtain a refund and move on is your best option.

Sorry can't be of more assistance.

[Glennbest]

Hi Jv,

Step counting is a bit of a fudge anyway but that's not what this topic is about. This topic discusses the issue of the charge 2 calculating distance incorrectly (that is, when you start a gps excercise using the button on the side of the charge 2).

Thanks

[jvcindian]

Yes, step counting, distance, Miles they are all incorrect!!!

Jvc

Sent from my iPad