I would ask to FW developers what’s the degree of robustness of geofence. I would try to understand if and how we can be 100% sure that, since we have defined a geofence area not to be trespassed, the UAS won’t go outside of that volume of airspace. Either in case of manual force from transmitter inputs, as well as flyaways and other contingency situations.
Of course it will depends by hardware quality, but assuming that the hardware (AP, GNSS receivers, magnetometer etc.) will be reasonably the best on market, can we certify a certain degree of robustness/guarantee that the system won’t overpass the volume?
I would say test it your self to be sure.
You can minimise the risk of a fly away by reducing vibration if vibes are an issue. Also making sure the tuning is as good as it can be - not good enough by eye, but in logs! This way you can be more sure of the aircrafts behaviour and performance.
I’ve been using the Fence feature for years and it’s worked, even when set with a smaller radius than the mission distance from home. The quad would hit the fence and take the desired action, leaving us wondering why it didnt complete the mission. This was back before I had telemtry to GCS or yaapu telemetry on the RC radio.
Thanks Shawn for the quick reply. We can of course test, except for a flyaway simulation, which I don’t know hot to simulate.
Thinking about the worst scenarios: magnetometers errors, misalignments or failures, interferences, GNSS multipath or solar storms, IMU errors… all of these are extremely rare, but may happen. Also, a human error like set auto mode with an old mission which is far from operations area.
For all of these cases, geofence should be effective.
Most of those bad things would cause FS_EKF_ACTION which is Land by default.
Flyaway is usually caused by excessive vibrations, and that should only persist for a few second before the vibration failsafe kicks in and causes Land.
If your RC went crazy and was sending the aircraft off into space, you could always just turn it off - having tested your radio failsafe before initial test flights
Thanks again Shawn. In this case, shouldn’t the geofence limit the volume and stop the drone?
Yes true the Fence would limit the range, but if the RC is misbehaving near obstacles (rare but possible I suppose) and your RTL switch didn’t work, the RC failsafe should.
I’ve not had that happen but we are discussing many scenarios.