Want to develop a watch application which runs in the background during the night time. The application will wait for acceleratometer events. When movement happens,
I need to figure out the intensity and length of the movement
Store this in somewhere (sqlite?)
Along with heart rate at the time of the movement
Then I will need to offload this data into an app running on the phone (Android). Where I will be able to see history and charts and options and so on.
There are lots of options to choose from, Native, Web, .Net and so on. I think I also read hybrid.
I read that if you want background applications, you should be writing a Native Application in C/C++.
This page: https://developer.samsung.com/galaxy-watch-develop gives 7 different options to develop with. I am not a great C/C++ developer, but can get by. I am not great with .Net C# either, but the development environment is easier. I am developing on Windows.
Given the above, which application model should I choose?
I am primarily concerned with choosing the correct option for the watch for battery efficiency and not conflicting (for example) with the sleep tracker application.
To handle sensors like accelerometer, heart rate (humanactivity api) and store it in sqlite, a file or even local storage you can easily use both Native C or Web API.
I would worry about using accelerometer and heart sensors in the same time for whole night. It might drain battery heavy.
I appreciate the response.
Since I use sleep tracking on the watch, I assume the heart rate gear is already active for the night.
I was hoping that while tracking the accelerometer, at the time of a movement, I could just fetch the heart rate at the time.
Really, I was expecting that while sleep tracking is active, all of this stuff is essentially available to any app running. Since sleep tracking is using the accelerometer and heart rate to decide if what stage of sleep you are in.
Good to know there are more options than simply storing the data in a sqlite database.
I hope it’s ok to recover this thread to check about accelerometer data in background during night time. I’m interested to apply my own analysis ( post or in real time ) on that data on accelerometer even gyroscope. May be apply some machine learning analysis to recognise some gestures in case it’s necessary.
So my 2 questions here:
Any available docu/tutorial to get on that specific data? I can work on Native C and Web Api.
For such gestures recognition… Probably data should have a framerate of at least 25 fps… What FPS I could achieve? I suppose that might depends on specific device right? But in general, any SW restriction?