We want to operate a drone autonomously and must satisfy very strict regulations, especially in terms of Geofencing. Therefore, I’m looking for some more clarification on the exact workings of the Geofence feature. Are the following statements correct:
- Whenever there is a Geofence breach, the FC will trigger the fence action. I can rely on that this applies always in any situation (“higher powers”, like wind pushing the drone around or bad GPS omitted, of course). Ardupilot, for example, has exceptions for RTL mode, which is allowed to breach the fence to return back home.
- When the fence is breached, the fence action is triggered once and then latched. Provided your fence action was “Hold”, you can now freely move around outside the fence afterwards without it triggering again. When you move back into the fly area, the fence action is unlatched and will trigger again on the next fence breach.
- When connecting with a companion computer, px4 doesn’t report the geofence violation to it, except in a human readable status text ( I get “MAVLink: critical: Geofence: approaching or outside geofence”, for example). There is no machine readable way to know of the breach, except you modify px4 itself.