Elevation gain and floors are completely inaccurate - any Fitbit product team comments here?

[AlmaEC]

Elevation gain on hikes or floor counts is completely overstated and inaccurate.  I noticed it broke in one of the previous software updates.  I have noticed others having similar issues on similar devices.

Is this community only a way for customers to vent? or is there actual Fitbit product teams monitoring and able to provide answers?  These are expensive devices and we expect expensive devices to perform more accurately.   Further more, we are paying a premium services for metrics, expect metrics to be accurate.   Anyone?

[Rich_Laue]

Hi @AlmaEC floor counting is accomplished by the changes in air pressure. Yes this does have its inherent problems. We also are aware of a Fitbit logic issue 

On hikes, altitude is determined by GPS and only available through the Web interface, will not be seen on through the fitbit app. 

Gps also has its inherent issues in that 4 satallites need to be acquired and heavy tree cover or clouds can affect the readings. 

[SunsetRunner]

@Rich_Laue , how can you explain, then, that yesterday night, just 20 minutes after midnight, when my Sense restarts all daily values from zero, including Floors, and when, after quietly watching a little TV, I was getting ready to go to bed (I live in a one floor only flat, no stairs), my Sense Floor counting showed already 6 floors?

And this is just one example. Since, at least, last May/June firmware update, these kind of errors happen everyday (I have stopped paying attention to that value at least a month ago).

Honestly, I can't find any reasonable explanation for the huge daily errors on Sense's Floors. As far as I lnow, air pressure variations alone can not explain these kind of huge errors happening in just 20 minutes.

NOTE 1: There was no storm or hurricane situation happening here, either yesterday or today.

NOTE 2: When I woke up this morning, Floor counting was already at 12. Amazing!

[SunsetRunner]

@SunsetRunner 

Miguel, one of the solutions to this ongoing issue...decided not to include these stats in the app and watch! LOL! It wasn't worth the time and effort for me to even consider or look at the floor count. And, you're right on your "NOTE 2", it is amazing! LOL 

[SunsetRunner]

You are absolutely right, @SunsetRunner . LOL

I wish the problems were only with Floors...

[SunsetRunner]

@AlmaEC , @Rich_Laue , @SunsetRunner ,

I just saw the weather forecast for the coming days. A low pressure is coming, with rain and bad weather starting tomorrow.

Suggestion: maybe use Floors as a Weather Forecast function instead? At least it couldn't be worse than floor counting. As for now:

• High Floor counting: air pressure is going down, bad weather is coming.
• Low Floor counting: air pressure is going up, sunny days are coming.

Just an idea. But I will start paying attention again to Floors.

[Rich_Laue]

@SunsetRunner please reread my post, the last sentence of the first paragraph. 

This issue has been going on since Sept of 2020. 

[SunsetRunner]

Thanks for your reply, @Rich_Laue .

Here are 3 graphs, taken from Fitbit app, with my monthly total steps. Sorry that they are in portuguese. "Andares" means Floors, "Ano" means "Year". I started using my Sense only in the beginning of January 2021.

From left to right:

Graph1: total month floors, according to Fitbit Sense. Yes, I usually don't go up many floors and, for health reasons, I also just do daily walks and had been slowly increasing my daily step countings, from 5000-6000 steps/day by January-February to 8000-9000 steps/day in July-August. That was a plus for Sense. It helped having more control on that. But I could have done it also with a US$40 tracker.

Graph 2: the same graph with Sense's April total Floor count.

Graph 3: the same graph with Sense's August total Floor count.

Monthly Sense Floors from January to August 2021Total Sense Floors in April 2021Total Sense Floors in August 2021

Looking at the graphs and knowing that on the second half of August I had health problems that made me have to reduce my walks and Floor climbing, it is quite obvious to me that:

1) The first OS/firmware update, in May/June (don't remember exactly when I applied), made the Floor counting much much worse. To the point of making it totally useless, which was not the case before. Before that update, it was not good, maybe it was even bad (I really don't know, as by then I payed attention to Floors only once and a while) but I still could a see a difference when I climbed more floors on a certain day.

2) After June OS/firmware update, Sense's Floors countings got much much worse. I would say that for me, and according to these graphs, Sense Floor countings increased 30 fold, which is ridicullous and made this parameter totally useless.

3) After the last OS/firmware update, which was supposed to correct this and several other issues caused by the May/June update, apparently everything remains quite the same, concerning Sense Floors count. Maybe a little less worse, but still ridicullous and useless. I apllied this second update only by August 20. Here is the graph with the daily values of Floors from the past month ("mês" means month, "Média diária" means "Daily average"):

So, my point is this. Maybe this is an issue going on since Sept of 2020. But, at least as far as I'm concerned, since June 2021 and after two Fitbit OS/firmware updates, it is much much worse, to the point of having become useless. Fitbit's graphs speak for themselves.

[Rich_Laue]

So you agree with what other users and I have been saying, @SunsetRunner that firmware updates have a logic issue in respect to stairs. 

[SunsetRunner]

@Rich_Laue , it looks like we all agree there is an issue here. For me, describing it as just a logic issue seems too short for the huge and unacceptable issue we are talking about. But we are now talking semantics. Let's leave it here.

Thanks for your feedback.

[Purrstachio]

Here's an idea. How about having the feature count elevation only - forget the air pressure changes. Then again, that might give you floors credit for riding up in an elevator - that wouldn't be any good either. How about elevation plus steps without forward movement or air pressure changes?

I actually have the opposite issue with floor counting. I live in very flat Florida in a one story house and rarely have an opportunity to climb actual stairs. When I do (like at an office building where I'm going to the 3rd floor), it actually does pretty well, e.g. I get credit for 2 or 3 floors. I avoid using elevators whenever possible. When I do video "hikes" on my incline treadmill, it more often than not registers zero floors. There's lots of elevation, but little forward movement and probably no air pressure change. That's probably the same reason floors don't count on a Stairmaster machine. Since Covid showed up, more folks are exercising indoors on Nordic Track or similar machines like me. Either drop the floor count entirely or make it useful - how about that, Fitbit?

[Rich_Laue]

Well @Purrstachio how do you propose to count elevation? 

Currently there are only two ways. 

We can measure changes in air pressure 

We know that the atmosphere has pressure, and the higher we go, the less air and its pressure is above us. 

We can also look at our relation to where 4 known known objects are located, even when these objects are moving at 12 times the speed of sound. However this does not work indoors are under heavy cloud cover. 

If fitbit does not detect steps, it should not be counting floors. Pointing to a possible problem with logic. 

There is a third way, and that would be to put in a tracking system inside each building. 

The cheapest option is still the same option, airplanes, mountain climbers, sky divers, and skin divers, and submarines use, that is to measure the pressure above. 

Of course the incline treadmill or stepping machine, will not calculate your imaginary climb. The user physically is not gaining altitude. 

On hikes, or even walking up stairs. The user has a relative short time to gain 10 feet. Floor counting is strictly designed to count the sets of stairs the user climbs up, not the changes in elevation while hiking. 

However while hiking, gps may be used to track changes in altitude. 

You point has been answered, the Charge 5 does not have a barometer and can not count floors. This has made many users upset. 

[SunsetRunner]

Hi @Purrstachio 

When you say:

"I actually have the opposite issue with floor counting. I live in very flat Florida in a one story house and rarely have an opportunity to climb actual stairs. When I do (like at an office building where I'm going to the 3rd floor), it actually does pretty well, e.g. I get credit for 2 or 3 floors."

That's about what I had until the first OS/firmware update in May/June. I also live most of the time in a quite flat area. I'm glad it keeps working for you. Not for me, now 😕

[scottscannon]

Some the responses seem to be missing the point.  The Sense USED TO measure floors accurately, and now IT DOES NOT.  Air pressure, phases of the moon, harmony of the planets, or Biorhythms, may or may not have anything to do with why it's not working now, but IT USED TO WORK.  So, there should be some way for FitBit to go back and FIX WHAT THEY BROKE.  That's all any of us want.

(I haven't done exhaustive research on the subject, but I believe that other trackers on the market have the ability to measure floors climbed.  WHY CAN'T FITBIT FIX SENSE TO DO THIS, TOO?)

I (we) await an answer FROM FITBIT.

[AlmaEC]

I agree @scottscannon ; I am suspicious about the last software update.  @Rich_Laue I have a good understanding on how the technology works.  While we don't expect 100% accuracy, technology is able to provide good approximations today and at the price level of the Sense, I expect better.  With regards to Elevation gain and GPS availability, the GPS details on the graphs for my hikes, are accurate.  The graphic depicts the starting elevation point, the highest elevation point, etc. etc,  The Elevation profile depicted on the hike is accurate but the Elevation gain calculation algorithm is completly wrong.  Any normal software development/testing team should be able to see that.  As a matter of fact, I have been using the elevation profile and manually calculating my hike elevation gain by manually substracting from highest point, the starting point elevation.  Example below shows Elevation gain of 3228 ft.  If you substract highest point 4918 - 3736 = 1182 (times two because it was an out and back, this is @ 2200 elevation gain, not 3228.  This functionality used to work before and now is broken and nobody at Fitbit seems to care to do anything about it.

[Purrstachio]

Ok, I see your point, @Rich_Laue. I actually have exchanged e-mails with someone who runs a large fan next to her treadmill while doing inclines. Somehow, she gets floors counted. That's a little too weird and I don't have room for a fan where my treadmill is. However, once my workout room is finished, there will be room - I might give it a try and see what it does.

Meanwhile, I'll either make do with it counting steps, calories burned, active zone minutes and distance - that'll do. If I find real flights of real stairs to climb, I'll do it, but that metric is just not going to be a measure of exercise, fitness, whatever. 
Thanks for the explanation.

[Ravsta]

This is obviously a reflection of broken code somewhere in the last update, as the elevation used to be reliable.  Whilst my wearable is not a Fitbit, mine registered the correct elevation gain on the same hike under the same conditions at the same time.  The variation was not air pressure.  I am assuming that the wearables are acquiring their GPS signals from the same satellites.

[SunsetRunner]

@Ravsta, just one point.

Sense counts Floors, or stairs, which are usually indoors, in places usually not reached by GPS signals. GPS satellite signals usually only work outside buildings. Inside they are blocked by walls. And, for the sake of precision, even outdoors, they usually are not very accurate on height, only on the two horizontal dimensions used for location on a map. I worked many years with GPS location and it's just a simple geometrical problem. Satellites are in orbits just about 20-30 km high and small errors on their signals often cause big errors along the vertical direction.

So, I understand that Fitbit uses air pressure to count floors. And they may add GPS info outdoors mainly to improve the overall accuracy, but here I am just guessing.

[Giampi71]

@scottscannon is right on point. The total inaccurate floors counting by Sensa/Versa 3 started following an update (we could actually tracked down what update it was), before it was not perfect but acceptable.

The fact that Fitbit after 6? 9? months is still not able/interested ot doesn't have resources/time to fix it is very concerning.

Let me analyze the possibilities

1) Fitbit is not interested: this is quite common in Fitbit, they create a bug and then they don't go back to fix despite the complains. I wonder if actually the moderators are communicating with the developers or if they are completely independent, just saying "fitbit is aware and it is working hard...." when actually they have no clue/possibily to speak with them

2) Fitbit doesn not have resources/time: all attention of Fitbit is towards news devices (this has been always a problem). However somebody might argue that at least one (or two) FW update has been produced since the origin of issue so why they didn't throw in this bug fix. This brings me to the third possibility

3) Fitbit is not capable to fix it: this is a long shot however if (and it is proven) the issue was caused by an update why Fitbit developers don't track back the source of the bug? All non amatorial developers should use a git method to organize/track their programming activities so this should be quite easy to achive. Doesn't Fitbit use GIT or similar methods? I would be worried about it.

As you can see all 3 options are possible. What's your pick?

[Rich_Laue]

Fitbit has never done anything with this floor count, they simply display for the user. 

Yes since a floor is about the same weight change as a 2 sq inch piece of paper, I would not be surprised. When the watch is on the Leeward side, it will be in a low pressure zone.