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.

[Rollerskier]

It could be various things but the old Charge HR model calculated distance much more accurate than the new Charge 2 model.

Charge HR you can only start on the App. and it plos the route even if you not wear your tracker.

Then you only loose the heart rate related to your Run.

I have tested model Charge HR with iphone 6, iphone 6+ and iphone 7 and it worked ok on distance with all of them.

I have tested Charge 2 with iphone 7 and iphone 6+ and I got more inaccurancies.

However, I get much better results when wear the iphone in a small sport bag on my back(better and more steady signals) than just putting it in the pocket of my jacket.  With Charge HR it did not matter.

I am not saying the route on map is better or worse with Charge 2 than Charge HR.  There will always be inaccurancies in GPS-signals.  However, the Charge HR plot is more rounded path and the Charge 2 is zick zack more on the micro scale.  This has somthing with how you wash and smooth the raw data prior to plotting it on the map. What you plot is what you calculate as distance.

I will show you the track with Charge 2 and charge HR recorded on the same road and 2 iphones.

Charge 2.  I walked on the side walk next to the road and with a good accuracy of the GPS it should have plotted on the side of the road.  As you can see it looks like I walked in the gardens , dich, in the middle of the road ie.  But I did not..I walked strait ahead.  Also notice the micro pattern..no steady direction.  It adds up to 4,09 km

Charge HR.  Same section of the road.  Here it looks like I also have been walking all over the place as well.  But I not have the micro changes in all directions as with charge 2 above.  So when you calculate the distance it adds up to only 3,24 km

The GPS data must be treated different in the App programs when the Charge 2 is set up as a device compared the when Charge HR is set up as a device. 

[Rollerskier]

Charge 2

Here is the another walk on the same roads.

This is Charge 2 but with an upgrade of firmware to 22.48.14 and iOS 10.2 and Fitbit 2.31

As you can see it is more smother but not as smoth as Charge HR.  Distance is now 3.52 km aboy 0,3 km too long.

[Rollerskier]

[Glennbest]

I think the focus on the gps signal in these posts is not right.  The distance tracked via mobile run Fitbit app is accurate so Fitbit have got that working ok.  Because the tracker uses Bluetooth to obtain the gps data from the phone I believe the problem is related to this transfer link.  Gps data is very complex and I don't think the Bluetooth link is handling it correctly, introducing an error into the data resulting in slightly 'wobbly' gps tracks which increases the recorded distance. Because of the lack of transparency Fitbit are demonstrating we are all left to theorise on our own.

[bobby2478]

To your discrepancy between charge hr and charge 2, charge hr doesn't have connected GPS so you log all exercise directly from the mobile app on your phone. 

Charge 2 has connected GPS, and since it doesn't have its own GPS it interfaces with your phone and uses the phones GPS. 

In either case the maps should look the same, the fact they aren't is likely the culprit and likely has to deal with the charge 2s connection to your phone that causes the map and distance to get thrown off. 

[Rollerskier]

Fitbit is not responding anything other than they replied the are aware of the problem...without saying  anything about how everything is set up... I have been adviced to send in my 1. replacement tracker and they will send me a 2. replacement tracker.  It probably will not change anything and it will take about 2 weeks to do this.

[bobby2478]

Glennbest, if you are confident the map and distance is correct when logging activity directly from the mobile app on your phone, and multiple sessions end up with consistent results, then I would agree with you that the problem has to deal with the connected GPS interface between the charge 2 and the mobile app. 

In those 2 scenarios, logging via app vs logging via charge 2, the only variable is the fact GPS data gets sent to the charge 2 via Bluetooth link.  Either the GPS Bluetooth link isn't steady (meaning it isn't a constant connection and drops off regularly), or to your point the GPS data must be getting corrupted since in either case the source of GPS is the phone and therefore should be consistent. 

[Bibster]

Bobby2478,

You may well be right, but Bluetooth connections are notorious for dropping in and out, so it would seem foolish of FitBit to design a system like this.

FitBit has allowed Strava to sync with the FitBit app. A constant good Bluetooth connection isn't required to successfully record GPS data on the Strava app (phone). A Bluetooth sync takes place once a run or ride is complete, and FitBit then matches the heart rate data (tracker) to the same time period as the GPS data.

If FitBit can talk to Strava to make this work, why can't it do the same wholly within its own App?

[bobby2478]

The positive in this whole situation is at least Fitbit officially recognizes the issue with accuracy with connected GPS is still not resolved and they are working on it. 

I'm sure the first firmware update with a fix for this focused solely on using GPS for distance when connected GPS is enabled,  instead of calculating distance based on stride length and number of steps.  I'm sure they didn't realize there was also a problem with the accuracy of the GPS data itself when using Connected GPS.

Hopefully with all the people willing to test things and /or share findings with fitbit they'll actually get it fixed in an upcoming firmware update. 

Fitbit, if your developers need assistance testing different theories to identify the problem please utilize this forum as there are plenty of people who I'm sure would be willing to test different scenarios provided the developers could get the underlying data somehow. 

[rafsanchez]

I agree with your thinking that there's something wrong with the Bluetooth link between the GPS device (smartphone) and the Charge 2.

In my scenario I use a Charge 2 when cycling to and from work and my verified 5.4 mile commute was variously reported by the Charge 2 as anywhere between 2.5 and 6.5 miles - crazy - but I think the fact that I kept the Charge 2 in my back pocket was a contributory factor and now it is on my handlebars and I get much more consistent results:

[rafsanchez]

I agree with your thinking that there's something wrong with the Bluetooth link between the GPS device (smartphone) and the Charge 2.

In my scenario I use a Charge 2 when cycling to and from work and my verified 5.4 mile commute was variously reported by the Charge 2 as anywhere between 2.5 and 6.5 miles - crazy - but I think the fact that I kept my smartphone in my back pocket was a contributory factor and now it is on my handlebars and I get much more consistent results:

However, even though I now don't get the completely weird results of 2.5 miles for a 5.4 mile commute (verified in both Google Maps and Strava), the FitBit is still regularly off by at least half a mile.

Below you can see a comparison of the results from FitBit (left) and Strava (right) for the same commute (I recorded the commute with both the Strava iOS app and also the FitBit Charge 2 using the button on the FitBit):

[rafsanchez]

I agree with your thinking that there's something wrong with the Bluetooth link between the GPS device (smartphone) and the Charge 2.

In my scenario I use a Charge 2 when cycling to and from work and my verified 5.4 mile commute was variously reported by the Charge 2 as anywhere between 2.5 and 6.5 miles - crazy - but I think the fact that I kept my smartphone in my back pocket was a contributory factor and now it is on my handlebars and I get much more consistent results:

However, even though I now don't get the completely weird results of 2.5 miles for a 5.4 mile commute (verified in both Google Maps and Strava), the FitBit is still regularly off by at least half a mile.

Below you can see a comparison of the results from FitBit (left) and Strava (right) for the same commute (I recorded the commute with both the Strava iOS app and also the FitBit Charge 2 using the button on the FitBit):

As you can see, FitBit is still recording the distance short!

[Glennbest]

Hi rafsanchez,

I had a serious online chat with Fitbit today. Although I'm reasonably sure Fitbit are taking this issue seriously the main complaint I still have is the complete lack of transparency about the issue. This forum topic is still listed as solved when it clearly isn't and emails are mostly met with the response 'we are working on it and appreciate your patience', I would really appreciate an update from Fitbit as to how they are progressing with the issue and where they are at with it, that would go a long way to convincing me to continue to assist them.

I just wonder if there is an issue with time dilation? There is a delay (albeit very small) between the phone receiving the gps signal and the the tracker receiving the data from the phone. If the clocks on the phone and tracker aren't in absolute sync, then this would introduce a small error between the phone and the tracker which would translate into a location discrepancy, because calculating a location via gps signal requires knowing the time it took the signal to get from the satellite to the gps device. It's strange that distances in the east-west direction are less accurate than north-south.

[bobby2478]

Definitely agree that the lack of transparency is frustrating. If they are aware of the problem then they should at least communicate exactly what they've found the problem to be.  If they've identified root cause then we can be confident they'll fix it. 

Just admitting you recognize there is an issue isn't the same as communicating that through weeks of testing and research you can pinpoint the exact root cause and are actively working on a solution. 

[Rich_Laue]

I have a Samsung S5 and find when using the connected GOS my route gets wobbly if my memory is to full, and find i have to clear it out or restart the phone before my walk/run.

@Rollerskier i find your post above a little confusing. When ysing the app to recotd tge run, tye heart rate and stes are still being recorded by the tracker snd both should show  up in the log. You stated the loss ofheart rate diring run.

You keep refferring to the Charge HR"s recording of the map, but have already said the Charge cant record a map and this is done by the app. Your comparing the app recording to tye Connected GPS recording .

As for Fitbit responding, what you received was from the crew that handles user problems when a user contacts Fitbit. These people are not the developers. As for sending it back, maybe after esculating your case tyey want to reseach your device to see what tye problem might be.

I 

[Rollerskier]

Hi,

The Charge HR model is only recording data and you can look at them on the small tracker screen and sync ithem back to the App.  I dont think the GPS data are transfered from the phone to the tracker and then synced back to the App later on this model. To start a RUN or WALK you do it from the App. You can look at the App during the RUN too. There is no bottom on the Charge HR tracker to start anything.

The Charge 2 model can start a RUN on the tracker itself or you can start it from the App too.

When you start it from the tracker the GPS data is synced to the tracker and all the information is synced back to the App first after completing the RUN.  The map can only been seen in the App.

Test 1

Today I will first leave my Charge 2 tracker home ( so the tracker will not affect the results ) and go for the same walk starting everything from the App.  The Charge 2 will however appear in the Account setup but there will be no connection during the RUN.  Then I will look at the map and the distance calculated.

The difference compared to the first Charge HR model GPS test is now that Charge 2 is paired to my phone in the settings instead of Charge HR is paired to the phone.  And that Charge HR had an active link to the phone App with heart rate recordings and steps.  While now there will be no active link since the tracker stays in my home.

So only the phone and the App involved this time.

Test 2

Afterwards, I will put the Charge 2 back on my wrist and start the RUN from the App again and do the same walk as before.

Then there is active communication with the tracker during the RUN.  However, it might not be the case for GPS data...we dont know for sure how this workes.

The question is will these 2 test today come out with the same distance and will it be the same distance as the earlier HR test?  And I will check the smothness of the track too.  And then post the the results.

[Rollerskier]

Test 1  Start RUN on App ( left Fitbit Charge 2 home - like the case when the battery is flat )

Distance 4,41 km ( True distance 3,2 km )

Map below

Test 2 Start RUN on App (  Fitbit Charge 2 on my wrist )
Distance 3,87 km ( True distance 3,2 km )
Map below

Surprise.  It did not realy matter that I started the RUN in the App.

I did not matter when I left the tracker home either.  The only relationship with Fitbit Charge 2 is that it was configured up versus the App.

The distance in the map is far off the real distance.

What worked fine when Fitbit Charge HR was configered in the App does not work when Charge 2 is configered in the App for the same iphone 7?  Unfortunately I can not test more on the old tracker since I gave my Charge HR to somebody else because I belived that Charge 2 was a better tracker..... 

[bobby2478]

Rollerskier 

I'd suggest the following 3 tests, for all 3 bring your phone along. 

Test 1: wear charge hr, from the Fitbit mobile app manually start and stop recording exercise to track the path you take. 

Test 2: wear charge 2, from the Fitbit mobile app manually start and stop recording exercise along the exact same path.  Be sure to start and stop recording at exactly the same spots as test 1.

Test 3: wear charge 2, manually start and stop recording exercise from the charge 2 along the exact same path as 1 and 2. Be sure to start and stop recording at exactly the same spots as previous 2 tests. 

If you do these, I'd expect 1 and 2 to be nearly identical as they both are manually recorded in the Fitbit app and thus neither should use the tracker whatsoever for calculating distance, they should both use GPS on your phone. 

[Rollerskier]

Those are goods test that you mention..and I have basically done them in the past.

The walks have been excatly the same route.  

I started and stoped at the same streetsign and walked in the middle of the sidewalkes on the same side of the streets every time.  

I used the same zebracrossings every time.  

I chose to follow the streets the entire walk so I knew I could compare the drawn path on the map versus the strees to see deviations.

I used the same iphone7.  

The iphone was in the same  pocket on the same side as I wear the tracker so the distance between them were minimal every time.

Use RUN mode on both App and RUN on tracker display.

I have filled in the distances versus your tests below:

Real distance is about 3,2 km

Test 1:  Charge HR: 3,24 km  ( can only be start/stop on App - tracker on arm)   6. Nov 2016 (walk)

I can not test this tracker again since I gave it away....

The Fitbit App has been updated 2 times since then.

Test 2:  Charge 2: 3,87 km  ( start/stop on App - tracker on arm )  13. Jan 2017 (walk)

Test 3:  Charge 2: 3,52 km  ( start/stop on tracker )  11. Jan 2017 (walk)

Test 4:  Charge 2: 4,41 km ( start/stop on App - left tracker home) date:  13. Jan 2017 (walk)

Test 5:  Charge 2: 3,14 km ( start/stop on App - phone in car window - best GPS )date: 20. Dec 2016 (driving). 

The rout is not exactly the same since I drow in the lane..but cant differ much..

See what a nice path you get driving 30 km/h.  If I could run/walk that fast things will be ok.

Distance in meter between 2 GPS datapoints is 6 times longer than walking.

 

So whats the problem with Charge 2 vs Charge HR?

Has fitbit started to have higher datapoints samplingsrates than with Charge HR?

Has there been some algorithm of smoothing and washing GPS data taken out or not working?

Fitbit need to put some effort into solving this...the distance calculated and pace right now cant be used for any practical purpose during workout...

[bobby2478]

Main thing with any test is to limit the number of variables. 

Regardless of what tracker you have on your wrist, if you manually record exercise via Fitbit mobile app (not auto recognized), and you select the same activity (Run vs Walk), and you take the same path, then results should be nearly identical.  Are you saying that you get radically different results performing this test?

Meaning you go for 2 runs, once will charge hr, once with charge 2, both times you manually record run activity via Fitbit mobile app, taking exact same path, and results are radically different? 

If so, try going for 2 separate runs, each without a tracker on at all, and manually record the run activity both times.  If these 2 results are different then it's your phone GPS and nothing to do with either tracker.