Following https://github.com/ArduPilot/ardupilot/pull/5260
In this PR, the crash check is base on velocity that came from IMU+GPS fusion, so it is working if you got a normal behaviour on GPS. I work in an area where GPS is heavily disturbed (without triggering an ekf or gps failsafe) so sometimes it doesn’t work as speed calculated is greater than the min limit I have fixed. That lead me to a couple of question :
1)Should the min velocity limit be a parameter ?
2)Should the min velocity be calculated based on throttle : I assume 5% throttle is slow enough ?
3)Should we use another logic based on only IMU data to detect that the rover isn’t moving ? In that, case I will need some assistance. I try with things like ahrs.get_accel_ef_blended() but without finding a good law.