Me and my partner are doing an engineering project of trying to predict errors and failures of the copter.
We have a great amount of logs and we now want to implement a Neural Network to identify what features cause errors.
We have tried searching the git on how each sensor decides if there is an error, but it is overwhelmingly massive code and we thought maybe someone here would know.
So for example, Subsystem 11 (GPS) has some error code, and we ant to know, based on which features does the subsystem decides that it has an error.
If anyone has a link/knowledge about this it would be very helpful.