Android Question Measuring distances (but not with GPS)

[AlpVir]

There are devices that allow you to measure distances of the order of centimeters and meters also using a smartphone.
See for example www.moasure.com.
One of the advantages would be to easily detect the plan of a building or the inside of a cave.
They certainly use accelerometers and other high-precision sensors but, we know it well, these sensors are also present in all smartphones. Why not use them?
These devices implement an inertial navigation system and it appears to be applied in the world of Android smartphones.
The question is: how to implement such a system in a smartphone with B4A, even at the cost of not high precision ?
Searching the internet for "inertial navigation android" I didn't find any great suggestions. Indeed I read that the implementation of this navigation system is not recommended as it would be very inaccurate.

 

[Erel]

See for example www.moasure.com.

It uses a specialized hardware. I don't think that it is based on the phone sensors.

You can theoretically read / aggregate the accelerometers and find the speed and distance. I guess that it will be too inaccurate to be useful. Search for an app that does it. If you don't find one then it is most probably not feasible.

 

[AlpVir]

I really think you're right, but I don't despair of finding useful documentation on the matter.
At least one app promises to implement this inertial navigation ("InertialNav") and therefore it seems feasible.

 

[DonManfred]

At least one app promises

try the app. Does it really work?

 

[AlpVir]

The app I reported ("InertialNav") actually doesn't appear compatible with my devices (which have the usual accelerometers, magnetometers and gyroscopes).
Conversely, I downloaded "Inertial navigation & location without GPS (INFA)_1.0.0_Apkpure.apk" and from a first test it seems very disappointing if used in the "only inertial" mode (the app also plans to be used in mixed mode, i.e. GPS + inertial).
Forward and backward for about fifty meters the track, instead of overlapping, has an error of 4-5 metres.
The problem seems exciting (also because - as Erel said - it seems unsolvable).