Sense is counting too many floors

[Deepvalley]

My new sence dosnt counting the right floors there are too much what can i do. sg Kurt

Moderator Edit: Clarified subject

[Guy_]

@NatomCZ  -  interesting, but like the watch doesn't count steps when pushing a trolley it may not count floors if you aren't walking upstairs normally or are holding the bannisters.

[NatomCZ]

@Guy_  I have no suspicion in this regard - I don't hold the railing when walking/running upstairs 🙂

[NatomCZ]

@Guy_ - oh, only now I've discovered the link in the signature of your posts and realized that you're the Guy guy behind http://guytec.com/Fitbit/ !

Great job! I hope you can make some sense out of the difference in floor counting in your app and in Fitbit's algorithm. I'm completely puzzled by today's results. In my office, I walk/run up 4 consecutive floors - I've done that 3 times today. I cannot understand how SimpleFloors did not count a single floor on 2 of those occasions. It seems to be registering very slight changes in barometric pressure even when sitting at the desk and not moving...

[Guy_]

@NatomCZ  - thanks for useful information, well spotted!

Are you left handed by chance?

The app won't register floor changes if you aren't moving. (The Fitbit app will)

The watch holds steps before releasing them to the apps once it has decided they are valid steps.

If the delay is too long it is possible to go up stairs before the watch givrs an indication you are moving.

In case you have set watch to dominant hand try non dominant.

The other thing is the Fitbit count vs real count is quite close, so your pressure sensitivity is pretty close.

My sense did 104 today while my Versa 1 did 7 which is about correct.

Thanks for your input.

[NatomCZ]

@Guy_  I'm right-handed, wearing the Sense on left hand, using app setting "non-dominant hand".

That's interesting to know that your Versa 1 exhibits the issue as well. I used to have Versa 2 before and the floor counts used to be precise. I've thought that the issue was limited to the newest generation of Fitbit gear (Versa 3 and Sense). So, if it affects Versa 1 as well, there seems to be hope that Fitbit made some bad changes in their floor count algorithm and there's a better chance they would rectify this to fix the precision.

It's just so frustrating they don't communicate about the issue and we have no info what's the status 😠

[Guy_]

@NatomCZ , the Versa 1 is fine, it's only the Sense and Versa 3 watches affected.

Thanks for the feedback, looks like SimpleFloors will need a sensitivity option.

[Guy_]

@NatomCZ  - As SimpleFloors doesn't count floors when stationary or when walking/running try going upstairs more slowly, normally 14 steps in about 8/9 seconds might get better results.

[NatomCZ]

@Guy_  oh, I misread your remark related to Sense and Versa 1 - maybe some wishful thinking on my side 😕  Now I see it's quite clear what you said - in you case, 104 floors reported on Sense and just 7 floors on Versa 1, which was about correct. I somehow misread it - not sure why. It was clear.

OK, I'll try going slower upstairs. I'm trying checking again today with SimpleFloors. So far, I went 4 floors upstairs, SimpleFloors registered 0, FitBit 7. But I went quite fast (chasing a waiting delivery man 😉). Will try if slowing down makes any difference.

[NatomCZ]

Today's evening score...

SimpleFloors: 4

Fitbit: 21

Floors actually climbed today: 10

Walking up slower seems to increase the chance that SimpleFloors registers the floor correctly. But it's not always the case. And it's just for testing purposes, of course... none wants to adjust their pace to the detection abilities of tracking algorithms.

 

[Guy_]

@NatomCZ  - thank you for your feedback & time testing.

It would seem your pressure sensor is slightly less errant than mine. Yours reads only double.

My Sense was 80 today, to 11 real floors which is nearly 8 times.

Will look at adding sensitivity to align it to different watches behaviour.

The main point to note is that if a simple app with only rudimentary access to the data, and after the event, is able to get a reasonable reading, a modification to the firmware should be able get an even more accurate result.

The whole reason is to show how the pressure sensor works (and that it is working, because some people get no floors) and assure people their watch is not actually faulty.

[Cougar713]

I'm having the same problem. I just went on a half hour walk in Houston. It's flat as a pancake here, no wind tonight, and my Sense says I went up 11 floors on my half hour walk.

[Downisup]

They sense is designed to replicate floors when walking.  The intent is to show the equivalent of floors through everyday movement.  Hope this helps.  

[MichaelNYC]

@Downisup are you working for Fitbit? I'm trying to understand what is the meaning of what you wrote? Are you saying Sense floor counting is working properly?

They sense is designed to replicate floors when walking.  The intent is to show the equivalent of floors through everyday movement.  Hope this helps.  

---

[NyeeNyee]

I have never heard this as an explanation given by anyone at Fitbit to any of the 1000’s of us that have this chronic problem. Nowhere have I have seen it suggested that normal walking will show up as floor count activity. 

[Guy_]

@NyeeNyee  what @Downisup  was probably trying to say was your watch can count floors when walking uphill.

Unfortunately as you point out apart from counting floors when going upstairs, it will also count floors randomly, when stationary, when walking, when sitting, when going downhill even.

It should only count floors when going upstairs, or increased elevations, the equivalent of a flight of stairs.

[LokeAa]

@Guy_  I am living at sea level - 299 m according to Sense 😂

As I have said before, I have more or less given up on Sense. The clock face fell of again today, will be interesting to see if floor count will be more accurate on a new device. 

Can say a lot about Fitbit and customer service, but I'll give them credit for being very efficient when it comes to replacements. They will read my e-mail monday, send me a case number, which I return with a picture of mybroken sense and handwritten case no. They ship it the same day from The Netherlands with DHL Express, and arrives at my doorstep tuesday. 

[Downisup]

Hi Michael, I am saying that it is designed to use activities to replicate floors.  For ex: If you do yard work or go for a long walk, Fitbit will use that data not only to calculate your active minutes but also suggest the floors you might have climbed if you were walking up stairs.  Yes, this is by design.  Not all users have access to climbing so it provides this information as an additional resource.  

[down_and_out]

That explains walking up stairs before I get out of bed. 

If what you suggest is by design then it has to be the most awful design in a tracking device since they starting making trackers. 

Lying still w/o movement should not generate stairs while at the same time not counting any steps. 

What your design concept implies is that Fitbit is double counting steps and exercise and activity and stairs.  So when one rides a bike you get a stair count or when you do yoga you get a stair count or when you do stairs it counts extra stairs to show how many equivalent stairs the stairs you did are equal to.

Sadly, as interesting a concept of a design flaw you propose, the Sense and stair counting would still fail.

Why you may ask?  Because it does not count phantom stairs all the time and there is no rational equivalency between the phantom stairs and the activity taking place.  I walk 70 steps roughly first thing in the morning.  I can register no stairs, 1, 2, 3 or 4 stairs making the morning trek.  Stairs reflect a certain level of effort which 70 steps on flat floors comes no where near to being similar in effort to a set of stairs. Or 4 stairs or no stairs or whatever the Sense decides to count based on the position of the moon and the stars.

And why would "Not all users have access to climbing... "(I assume stairs) impact a design decision?   There are specific exercises you can track including using a stair climber, which is significantly different using stairs, and others that not everyone has access to  Such as swimming.

Especially when alternatives exist within the Sense to accomplish what you suggest such as Active Zone Minutes, Heart Rate Zone.etc.

[NyeeNyee]

Just curious, has Fitbit told you it is “by design,” or is this your hypothesis? 

[NatomCZ]

@Downisup Can you link some source where this behavior is documented by Fitbit?