Lost control, then 2 seconds later lost connection, on my Mavic Pro what happened ??

[Gryphon962]

It's a very unusual scenario for the aircraft to attempt to descend but lose no altitude but, in that situation, I don't think it needs full down throttle to conclude that it landed.

Wow. We can only wonder how many others have gone same way, by just flying into an updraft.

This is certainly something DJI needs to be told about: if the MP is shutting down its motors through presumption its landed when GPS velocity, IMU velocity, baro_alt, and lack of ground proximity sensing all say otherwise, its not airworthy. If the price to be paid is that it wont 'land in your hand' then its worth paying, IMHO.

Do we have 'channels' to DJI?

 

[sar104]

Wow. We can only wonder how many others have gone same way, by just flying into an updraft.

This is certainly something DJI needs to be told about: if the MP is shutting down its motors through presumption its landed when GPS velocity, IMU velocity, baro_alt, and lack of ground proximity sensing all say otherwise, its not airworthy. If the price to be paid is that it wont 'land in your hand' then its worth paying, IMHO.

Do we have 'channels' to DJI?

GPS, barometer and IMU tell it nothing about whether it has landed, because they don't know the ground elevation. The VPS system could confirm landing or otherwise, and I'm surprised that it isn't used for that. That said, I've never heard of any other aircraft being lost that way.

 

[Gryphon962]

GPS, barometer and IMU tell it nothing about whether it has landed, because they don't know the ground elevation. The VPS system could confirm landing or otherwise, and I'm surprised that it isn't used for that. That said, I've never heard of any other aircraft being lost that way.

What I meant by reference to the GPS, barometer and IMU sensors is that each has the potential to provide clear evidence that the MP is still moving and therefore cannot be on the ground. For example, if I gave you a log that had only data from those three sensors, you would have no difficulty deciding if the drone had NOT landed in all cases where it had not. I concede that you wouldn't be able to determine with certainty if it HAD landed, but that should be a pilot decision anyway.

 

[sar104]

What I meant by reference to the GPS, barometer and IMU sensors is that each has the potential to provide clear evidence that the MP is still moving and therefore cannot be on the ground. For example, if I gave you a log that had only data from those three sensors, you would have no difficulty deciding if the drone had NOT landed in all cases where it had not. I concede that you wouldn't be able to determine with certainty if it HAD landed, but that should be a pilot decision anyway.

And my point was that the GPS, barometer and IMU cannot provide that clear evidence if the aircraft is throttling down but not descending.

 

[Gryphon962]

And my point was that the GPS, barometer and IMU cannot provide that clear evidence if the aircraft is throttling down but not descending.

You are right.

I hadn't considered the case where you land on something that is moving (which was rare in my flying career, but it happens). In that case, you are landed but a non-zero velocity vector (from GPS or IMU) still exists. So yes, VPS is really all you have for confirmation.