Ionic GPS Inaccuracies

[MP4]

Hi, 

I'm a new Fitbit user, bought an Ionic on Monday and I'm having some issues with the GPS. Went on a couple of runs (at a track) and the gps seemed to be completely out both on the distance travelled and pace per kilometre.

I'm a long time strava user so yesterday I went for a run with both the Fitbit Ionic going and the strava on my iPhone running at the same time. Here are the results:

Strava - 7.9km   @ 4:15/km

Fitbit - 7.2km     @ 4:34/km

The difference of 0.7km is quite big and the difference in pace is also worrying. I waited till both had connected to gps before starting the run and I run in London where gps signal should be good. 

Is my Fitbit Ionic GPS not working correctly? Any help/advice would be appreciated. 

Thanks

Matt

Moderator Edit: Updated Subject for Clarity

[SunsetRunner]

On 5km run the difference between Ionic and my mobile ( with more accurate GPS tracking ) is 750m. It's a lot for such a short distance. What I noticed, accuracy goes down the faster I run and the bendier the route is. Seems like the GPS sampling rate is too low so Ionic is forced to take "shortcuts". Possibly to make power consumption lower. I don't mind to have less accurate but power efficient GPS on board as not always I need super accurate sensor but when I need I would like to be able to use Connected GPS which could provide more accurate readings. Otherwise, having always my mobile phone with me, yet I'm forced to use an inaccurate GPS sensor. Isn't possible to add an option of higher sampling rate at a cost of battery life somewhere in the settings?

[Ericws]

My Ionic has been spot-on perfect in GPS.  So glad I did't need to spend $200+ more on a device when this one works so well.

[SunsetRunner]

@Ericws I'm happy for you but how does this fact solve my problem that on 5k route Ionic shows over 700m less? If you know any way of improving the GPS readings the I'm all ears. Analyzing data I can clearly see that Ionic likes to take shortcuts on the bends and when I speed up ( other devices do that too, but that's why there is such thing like GPS sampling rate which on the phone can be programmatically changed for example if we want to save battery lifetime ). Not mentioning that it takes quite long to connect to GPS. If you don't experience that, it's fine. I'm glad that for somebody everything works. In fact, I believe users who experience such problems are not the majority. But it doesn't mean those problems don't exist.

[Ericws]

I don't know.  Is it always 700 m less and do you have an accurate measurement to compare it to?  

Like I said, I am comparing mine to other routes that I have measured out and run for several years, and Ionic measures them right on the money.  A few 5k's I have done have also measured correctly.  I don't run the tangents perfectly, so I might be at 3.2+ miles, but that is because of my running, not the Ionic.

[lckysquirrel]

I contacted Fitbit and found that they actually don't use the GPS for
distance tracking. You need to calibrate your stride length to get an
accurate result.

I will say, they must have done some change on their end to fix the
swimming as well unless somehow the run stride affects the swimming count
as well. Either way, easy fix is to reverse engineer the correct distance
to the stride length i.e. take the actual distance and divide by the
reported distance. Then multiply that number (should be 1.XX) by the
current stride length in the app and use that number instead. Simple easy
fix.

Again, Fitbit does NOT use the GPS to calculate distance.


Ricardo T. Sanchez | rtsanchez@gmail.com | 202-596-2234

[SunsetRunner]

@lckysquirrel if this is true then what is the point of having GPS in the watch? If I run for 10k and I see that watch shows 9k then I know 1k left. I use it for real-time tracking during run or cycling. Sometimes I get myself going off the planned route but still I want to run particular distance. I would like it to be clarifies by Fitbit because if it's true that Ionic doesn't use GPS to tell the distance, then it makes it even more useless.

About accuracy, I took coordinates recorded by the exercise and put it on the map. Same I did with coordinates recorded by Polar Beat and S-Health ( both use mobile GPS and require to switch off power saving for better accuracy ) and guess which route doesn't match. Ionic accuracy goes down whenever there is a bend. I use actual data for comparison. I am replying from the phone now but when I find some time later I can make a visual breakdown to show where inaccuracies appear like I did yesterday with HR inaccuracy in another thread. I don't like to guess or believe in things and prefer to work with raw data and raw data can be unforgiving.

[Ericws]

It is not true.  Ionic uses GPS.  The other Fitbit’s such as Versa? No.  But it is untrue to say that there is not a built-in GPS in Ionic.  If there were not, how could it generate the maps of your runs/rides?  C’mon now. 

I ran almost 3 miles yesterday.  I did not have my phone with me. Only my Ionic.  It provided a map of my workout route.  Guess what?  The Ionic’s GPS worked.

In fairness, and I will defer to Fitbit- the Ionic May use GLONASS which is an alternative to GPS, but still is a very viable satellite-based system.  I think it has some plusses and minuses compared to GPS. But to say this is based on stride only seems silly to me Based on my experience.  

I am not a wilderness hiker.  So I am not too far off of the grid generally.  I have used Ionic and had no issues with this.

Moderator Edit: Merged User's Posts

[SunsetRunner]

All I said isn't based on the wild guess but on numbers. People lie, numbers not. I used coordinates from one of my 8km runs from 2nd August. This run for Ionic ended at 7.77km. In fact, it was almost straight run. The accuracy of GPS reads from Ionic indicate that probably I was involved in triathlon as significant part of my run is on the river 🙂 Yeah, I ended up all wet but I don't remember swimming at all 🙂 I'm going to share KML files recorded by my mobile phone and recorded by my Ionic. I processed TCX from Fitbit and GPX from Polar Beat. At the bottom, I will link the KML files and distance analysis.

There are a few things that actually surprised me as well:

- Ionic distance, according to the map is 8.03km ( wow! it's actually correct despite swimming was involved 😛 )

- Polar distance is... wait for it... wait for it... 8.03km!

- Ionic and Polar Beat show the same distance! - surprise! ( even for me )

- Yes, Ionic isn't accurate as the path doesn't even stick to the ground...

- ...but it isn't inaccurate so much that the distance goes far off. In fact. the error margin is surprisingly low.

- The difference is 5m which for long distance runs is absolutely nothing.

This sheds new light on the Running activity with GPS on. I think it proves what @lckysquirrel said about using stride rather than live GPS to show current distance. This finding makes Ionic useless for me when what counts is actual distance. I remember one of my 10k runs to Windsor where S-Health was reporting finishing the run but Ionic still was behind so I added some distance just to hit 10k on Ionic. I'm curious then, how the pace is computed on Ionic. It's always less than my real pace indicated by mobile apps. I very often want to keep correct pace when running for endurance. Is that another thing Ionic is useful? If it takes into account my stride ( which may change a lot during a single run ) not the GPS data, then it's another useless, bogus number. I know not everyone runs for the result but I do and after what I find I'm getting more disappointed.

Above you can see the inaccuracy of Ionic ( part of the path is on the water ). Still, it is acceptable, as final mileage is very correct. To be honest, if not that I see the Ionic path going through water I would have difficulty to say whether Polar ( connected GPS ) or Ionic is less accurate.

Detail paths can be viewed:

Polar H10: https://www.sportdistancecalculator.com/index.php?id=21626

Fitbit Ionic: https://www.sportdistancecalculator.com/index.php?id=21627

There are links to download KML and GPX files for each path.

Concluding, there are few questions in my mind that probably will remain unanswered.

- Why Ionic, despite having GPS on board, has been so crippled in terms of measuring distance?

- Is 'pace' not using GPS either?

- What about distance when cycling? There is no stride in that case so...

- ...does Ionic magically starts using GPS to compute distance for Cycling?

I know there are always people defending the device at all cost but I also know there are people believing that the Earth is flat. Data is data, hard to argue.

Edit: removed direct nickname, as this post is not really directed to anybody in person, but to the Fitbit a the company.

[Ericws]

But I am having a different experience than you are. Just because you doubt your results doesn’t mean mine is not working correctly, too.  What I find ridiculous here is that every complainer thinks that the issue they have is in every one.  It is not.

[SunsetRunner]

@Ericws ok, I get it. Data won't convince you and I'm not even thinking of convincing you to anything. That wasn't my point at all. Maybe I shouldn't have even use your nickname. My bad. And please, don't call me "complainer", as for a "complainer" I did put lots of effort into trying to figure out what the problem is that I encounter and why. Seems like I busted my own theory of inaccurate GPS but then I still see different numbers on the watch during run and different when I use the actual GPS data ( from the watch ). And the distance in my activity log is shorter than what comes up after using GPS readings. Yes, I get it, it works for you - I don't argue with that. So please do understand too that it simply doesn't work for me. I'm not trying to spoil your experience and if there are happy users - great for them, great for Fitbit. But the fact it works for you doesn't solve any problem. You may argue with numbers and I'm fine with that. I'm not a complainer but disappointed user, who digs deeper on his own to find out why things don't work as expected for him. If it works for others - that's fine. But not "others" paid for my Ionic but I and I think I deserve some answers from people who made a product. How can I train using "pace" or "distance" as one of the goals when Ionic doesn't show me anything close to the truth? Maybe I shouldn't be digging and just blindly believe but I can't. It's not in my nature. There must be an evidence for everything. I know Fitbit won't answer any of my questions as every answer is a quick copy-paste of what's allowed to say. That's all from me.

[Ericws]

I can dig up my data to show that my Ionic’s GPS (GLONASS) is accurate.

if the allegation that Ionic is not using A global satellite deal in its watch is true, then they are open to false advertising because they publicize this watch as a GPS watch.  

I have no reason to doubt them. And I do not assume that because my experience has been good, everyone else is wrong, like I suggest to others that if your experience has not been good, don’t assume the whole Ionic deal is a dud- because they work.

[SunsetRunner]

@Ericws I think we speak different languages. I admitted that GPS in Ionic is accurate. Very accurate comparing to my mobile GPS ( the difference of 5m is nothing really ). I never said Ionic doesn't use GPS because it isn't true. Ionic uses GPS to track exercise. Otherwise, I would not have any data out of it. What I say is that not necessary it uses GPS for tracking distance real-time. What display shows doesn't match data collected by Ionic's GPS ( which IS accurate as I shown above ). This could be simplification as connecting to GPS takes time and if you start running before the connection is established, the distance will be already non-zero. I assumed that ONLY when GPS is not connected yet or occasionally lost the using stride for updating distance is a fallback. I didn't think it could be just using stride for the whole time of the run when GPS is in fact connected. You put words in my mouth which I didn't say.

"I can dig up my data to show that my Ionic’s GPS (GLONASS) is accurate." - so do it, unless it's just talking. I'm really curious because it could be just my Ionic. I have lots of other issues and I take into consideration the option that it's simply broken. The more data the better as I still try to figure out if there's anything I'm doing wrong.

"if your experience has not been good, don’t assume the whole Ionic deal is a dud- because they work." - where did I assume that? I said something completely opposite. English isn't my first language but it cannot be that bad to lead to such great misunderstandings or is it? Whatever. This conversation has no point.

[Ericws]

Sorry to be difficult- I think I have been responding more to the suggestion that Ionic does not use GPS than to your statements.  

I think we concur that Ionic does use GPS.  When I am just walking around and the Ionic is taking count of my step, then sure, stride etc., comes into play for measured distance traveled in a day.  But when I turn on a running workout, and activate the GPS, it is the GPS working to measure.

I don't know why yours is not quite as accurate as mine.  I wish Fitbit would help you address it.  Sorry to have been argumentative with you.

[SunsetRunner]

@Ericws no worries. I also let myself to get a little heated. You see, this is what I would like to find out because things for me never add up. What you're saying is exactly what I've been expecting from Ionic - GPS on so it's used to measure distance ( and measure stride if one uses automatic stride calculation as there is such an option in the app ), GPS off - using stride, which can be as accurate as it can but it's better than nothing, a fallback. So far I've been thinking it's GPS inaccuracy but I admit, I had to eat my own words about that. Fitbit came up, at least in this one test really accurate! That's opposite to what I expected before analyzing the data. But then in my dashboard, the activity looks like this:

Everyone will agree that 7.77km isn't same as 8.03km. Nothing matches here. That's why I direct my question to Fitbit how is this possible, that the watch captured proper 8km run, the dashboard displays same path ( using same navigation points ) but numbers are incorrect. Where this 7.77km come from? Can it be a bug in an algorithm that computes a distance between two waypoints? One thing that comes to my head is trying to switch to miles and see what results I'm gonna get then. But really I'd like Fitbit to respond and give any explanation. It's not like I'm seeing things 🙂 And there are so many options starting from my Ionic malfunctioning to the bug in software in the code that computes distance - it could be really anything. @Ericws in your case, do you use miles or kilometres as your unit? I'm going to see if there is any difference for me after switching but you may give me a clue ( please, say miles 😛 ).

[bmw54]

---

@SunsetRunner wrote:

@Ericws I think we speak different languages. I admitted that GPS in Ionic is accurate. Very accurate comparing to my mobile GPS ( the difference of 5m is nothing really ). I never said Ionic doesn't use GPS because it isn't true. Ionic uses GPS to track exercise. Otherwise, I would not have any data out of it. What I say is that not necessary it uses GPS for tracking distance real-time. What display shows doesn't match data collected by Ionic's GPS ( which IS accurate as I shown above ). This could be simplification as connecting to GPS takes time and if you start running before the connection is established, the distance will be already non-zero. I assumed that ONLY when GPS is not connected yet or occasionally lost the using stride for updating distance is a fallback. I didn't think it could be just using stride for the whole time of the run when GPS is in fact connected. You put words in my mouth which I didn't say.

"I can dig up my data to show that my Ionic’s GPS (GLONASS) is accurate." - so do it, unless it's just talking. I'm really curious because it could be just my Ionic. I have lots of other issues and I take into consideration the option that it's simply broken. The more data the better as I still try to figure out if there's anything I'm doing wrong.

"if your experience has not been good, don’t assume the whole Ionic deal is a dud- because they work." - where did I assume that? I said something completely opposite. English isn't my first language but it cannot be that bad to lead to such great misunderstandings or is it? Whatever. This conversation has no point.

---

This topic has been hashed over through the course of months in the 'GPS Innacuracies' therald. Through experiments and conversations with Fitbit chat many of us came to the realization that the ionic uses GPS to create the maps of your run, but it uses clock time and stride length to calculate distance and pace.

For consistent joggers who have stride length stated accurately this is fine. For competitors who vary pace and stride strategically, it is not good. If you have a few hours to spare, check out the thread.

I'd like to say 'case closed' but in the past few months my ionic has become much more accurate on race day. I don't know why that's changed, but I'm treating it harshly (in surf and heat) so that it will die and I can justify replacing it with garmin. 

[SunsetRunner]

@bmw54 this all sounds nice as a theory but it is enough to turn off GPS for Run activity to see totally different results. Onscreen pace went incredibly up comparing to when GPS is on:

6:00 became 7:20

5:00 became 7:15

4:00 became 7:05

This is what I sort of expected from "fallback" which is using stride, time, steps and maybe acceleration to estimate parameters of run without knowing location. With GPS on the results are not accurage but they are closer to truth. Because of that I consider GPS being involved in computations. Also I use automatic stride which I presume is being adjusted based on GPS everytime I run and then the stride is used as a fallback when GPS is off.

My running is mostly pace oriented. I change my stride and pace quite often and it really matters to keep it right if I don't want to run out of energy and keep it going 😉 I am sort of "pace runner". That's why I am digging that hole as I was very happy after I bought Ionic that I could track my pace in real time and adjust my efforts immediately.

I have been thinking and mixing up GPS and stride could lead to some minor differences but thag would mean the GPS connection is lost often for longer period of times. I didn't analyze data from this angle but if suddenly my pace from 4:00 would become 7:00 so just 30 seconds of list GPS would result with 179 meters lost. Lose GPS several more times and it adds up. It is just theory like any other. I see GPS lost few times during run but never bothered with it as I assumed that it will get approximated with next waypoint sampled after getting connection back. Seeing how big difference is made when I don't use GPS in my pace almost explains what I am getting if I consider that connection gets occasionally unstable and data is substituted with "slower" fallback and not corrected at the end of run. But it is just a theory and I would need lot more data and controlled environment to see whether it works. Best it would be if Fitbit simply stated how the run parameters are being computed and what variables contribute. I wouldn't mind to see occasional falkback error if right after getting connection back error is corrected. Even when running with Strava or Endomondo occasionally I hear "GPS signal lost". But again. I'm just thinking and without Fitbit confirming or denying it doesn't bring lot to the table.

[Peach]

6km into a 10km walk my activity stopped recording.  I won’t bother contacting Fitbit support, when I know the advice I’ll be given:  Reset Fitbit, hope it doesn’t happen again.  I am so unhappy right now.

[SantiR]

@SunsetRunner I hope you're doing well! Thanks for sharing all of those details regarding your GPS runs.

I went ahead and created a case with customer support in your behalf. Please keep an eye on your email inbox for next steps. 

@bmw54 Thanks for sharing your experience here. I'm glad to read Ionic is now more accurate than before. I hope it will resist the harsh training you are performing.

@Ericws Thanks for sharing your positive experience with GPS runs on this thread. 

@lckysquirrel Welcome to the Fitbit Community! Thanks for contacting customer support and sharing your experience on this thread. If you keep experience any trouble with your watch, please feel free to continue to troubleshoot with them. 

@topo_treker It's great to see you in the Fitbit family! Thanks for sharing your experience with us and for reporting this to customer support. If you keep experiencing the same situation please go back to them and let them know that it is still happening. That way, they can check what other option they can provide you to get you back on track. 

Keep me posted! 

[SunsetRunner]

@SantiR Thanks. I went further in analyzing data and found more interesting things. I think there may be a bug in distance computation if GPS is taken into account. The TCX file contains "<DistanceMeters>" tag which is exactly what Ionic shows. Simple computation of the distance between waypoints indicates that it is clearly wrong. Most of the times the spherical law of cosines does the job ( this is what most of the tracking software uses including Google Maps ). Using coordinates captured by Ionic the distance is perfect when SLoC is used. But even using the less accurate method like equirectangular approximation the difference to SLoC is just 2m on the 1km split. Ionic reports ( and records in the TCX file ) 921m for the same set of coordinates. My point is that everyone so far was saying that GPS is not accurate while this seems not to be the case. GPS seems to be as accurate as any. Yes, there are minor errors but nothing preventing anybody from using it as a reliable feature. But other metrics don't add up and the more I analyze the data the more I think there is a bug in computation. If it happens that I am right, then the good news is that it's probably fixable by the software update. The worst case scenario would be that GPS readings are wrong and fortunately, this isn't the case.

[IainCostes]

I regularly run the same routes but recently the GPS log of the Ionic has become inaccurate. It regularly misses out whole sections of the route or shows me running on roads that I haven't been on. This seems to have happened since a recent update as my stats were always accurate up until 2 months or so ago. Now I cant rely on it and means I am unable to compare results for the same route on strata, as it thinks I've run a different route.

Also, uploads to Strava are not always reliable and sometimes don't happen at all. This too seems to be following a recent update.

Any one had similar issues?