I am trying to build a Fitbit style fitness tracker for my 4 month old puppy. I am using a TinyZero processor, Pulse Ox sensor, Accelerometer sensor, and SD card shield. My plan is to classify activity into 4 different categories. The first would be walking which would be determined by lots of change in X and Y and a heart rate between 80-120. the next is running which is classified the same way as walking but the heart rate is between 120 and 200. the next category is playing which is classified by moderate change in X, Y, and Z with a heart rate between 80 and 200. The last category is resting which is classified by low to no change in X and Y with a heart rate below 90.

I was wondering if someone knew what the units for x, y, and z captured by the Accelerometer Wireling are so that I can get an accurate view of what moderate and high change in the x and y values is actually like in the real world.

Thanks in advance

Hiya,

There's some example data of the accelerometer output available in the tutorial via a plotted graph: https://learn.tinycircuits.com/Wirelings/Accelerometer_Wireling_Tutorial/

That data will probably not be helpful since I was just waving the accelerometer around before I took that screenshot. The best way to know the change in accelerometer values is to test with the real situation/factors that you will be working with.

I will offer a few words of caution with the pulse oximeter - these sensors are very sensitive in general and this hardware in particular takes a few readings over multiple seconds before being able to print values to some degree of accuracy. For a human, this means holding a finger gently and very still over the red LED of the wireling sensor for a few seconds in order to get a reading. Most real world pulse ox's are embedded in some sort of pinching-finger-apparatus to keep the sensor still with even pressure over the course of a few seconds - at least the ones I've encountered. I'm not sure if the pulse oximeter will be usable with a dog without some sort of fixture (and possibly some hair shaving).

It sounds like a fun project and I would love to hear more about your execution and results! Pictures of the puppy would be an extremely welcome bonus. (Sorry if the pulse info is a bummer.) It reminded me of this neat turtle tracker project I saw a bit ago using our hardware: https://www.hackster.io/brian-k2/accelerometer-and-data-logger-for-small-animal-research-c877c6
The accelerometer code in that project might help!

Cheers,
Réna