MA Lost connection with remote control and landed in the ocean.

[Billk907]

.....

 

[Billk907]

Problem? Mavic Air lost connection with the remote control, It hovered in place until the battery was depleted and then landed in the ocean and sunk.
After several minutes of no control, I restarted the remote control. It appeared that I reconnected with the drone but it was too late. The drone had started an emergency landing and would not accept my commands. I saw the drone hovering above me and watched it land in the bay. Once the connection was lost I had no control over the drone.

Was unit in a crash? Yes

What have you tried so far?: downloaded teh DJI flight record and the DAT files

What device are you using? Samsung Galaxy Tab S2

What firmware are you running ( aircraft, remote controller)? newest

What Go app version are you using? DJI Go 4v4.3.20

Any modification? its all a stock unit

Did you change anything or install any apps? No

Do you have a video or pictures of the problem? No I don't sorry

The DAT and the DJI flight record are attached.

I would like to understand why this happened and how can I avoid this in the future.

I will appreciate any insight provided.

Sincerely,

Bill K

 

[sar104]

There are several things I don't quite understand here. The aircraft was in RTH when the downlink was lost at 460.1 seconds. At the rate it was returning it should have been at the home point 35 seconds later, but when it reconnected 596 seconds later it was still 22 meters away and autolanding due to low battery.



Once the aircraft had reconnected you stated that you had no control, but the log shows that you didn't attempt any stick inputs at all after the reconnect except for some brief (and strange) down throttle when it was autolanding but not yet descending, which it responded to as expected.

 

[Meta4]

I would like to understand why this happened and how can I avoid this in the future.

Here's what that flight data looks like:

DJI Flight Log Viewer | Phantom Help

Upload and view flight logs from your DJI drones

www.phantomhelp.com

Once the aircraft had reconnected you stated that you had no control, but the log shows that you didn't attempt any stick inputs at all after the reconnect except for some brief (and strange) down throttle when it was autolanding but not yet descending, which it responded to as expected.

There's something mysterious in the flight data but the left stick down is quite logical.
Signal was lost for 10 minutes and when the connection was re=established at 17:37.3 the Mavic was 22 metres from home, autodescending and had already descended 60 metres.
Normally in RTH the drone would fly to a point directly above home before descending.
But a minute after signal was regained, the drone was still autolanding but stuck 111 metres up.
At that point the OP tried left stick down, which brought the drone down a little but stalled again at 104 metres.
He tried left stick down again at 18:45.6 and brought the drone down much lower.
It continued descending when the left stick was released at 19:05.7.
It's unfortunate that he didn't use the right stick as well while descending.

The mystery is why the descent started at 22 metres out and what stalled the descent for so long ???
The usual cause of something like that would be a VPS issue but no VPS data is showing.

 

[RayOZ]

10 mins seems like a long time not doing anything to re-establish connection.
My guess:
It was data transfer between the RC and your tablet that stopped, but you assume it was your drone and your RC. Somehow, you’ve managed to cancel RTH (was the remote beeping the whole 10mins?). So, your drone was just hovering almost above you, but being so high, you probably didn’t hear or see it.
Eventually, you restarted your RC, which re-establish data transfer to your tablet, which you then see it descending.
I am not familiar with the MA’s. Can you get the dat file from the RC?

 

[Deleted member 94047]

I learned about a new flight mode - "Detour Mode".

To OP; sorry about your mishap but thank you for sharing this flight data. Hopefully, the wizards here will be able to figure out what went wrong and the rest of us can learn from it. One question though, why did you have APAS enabled while flying that high? What possible obstacle could there be at that height?

 

[BigAl07]

Sounds to me like the RTH was set to 'Hover' on loss of signal - REALLY useful when flying from a boat & over water! The analysts on here will be able to tell you for sure when they look at your data files.

That's a common thing when operating from a mobile platform (boat, car, motorcycle etc) because the Home Point is Dynamic and might be a LONG Way away from where you're at when the problem happens.

 

[BigAl07]

**** Duplicate threads Merged and cleaned up a bit ****

 

[Meta4]

Sounds to me like the RTH was set to 'Hover' on loss of signal

Not the case as the drone returned almost all the way back.

 

[Deleted member 94047]

Not the case as the drone returned almost all the way back.

Assuming AC was configured to hover on remote signal lost, and that up link stayed active for at least 32 seconds after down link was lost, here is a scenario for you:

1. At 7m 10.2s, pilot manually triggers RTH. AC starts traveling to the home point, which was 470m away.
3. At 7m 40.7s, down link is lost; AC is 254m away from home point at this time traveleling at a speed of 7.2m/s.
4. AC travels the rest of the distance to 22m away from the home point until up link is lost 32 seconds later.
5. AC hovers in place at 22m away from home point; and waits till remote signal is reacquired.
6. AC reacquires remote signal at 17m 37.3s. But by this time, it had already entered into forced landing.

Just a scenario.

 

[Meta4]

4. AC travels the rest of the distance to 22m away from the home point until up link is lost 32 seconds later.
5. AC hovers in place at 22m away from home point; and waits till remote signal is reacquired.

Why would a drone in RTH stop 22 metres from home and wait?
What made it start to descent?
Why would the descent stall at 111 metres?

 

[Deleted member 94047]

Why would a drone in RTH stop 22 metres from home and wait?

Was this a low battery RTH? Unlikely since battery was at 62%. Was it a fail safe RTH? If so, why would it get to 22m of the home point and stop? There would have to be user interaction to cancel a fail safe RTH, there seems to have been none. What does that leave, a pilot initiated smart RTH. And if RTH was manually triggered, what is the expected behavior of an AC that is configured to hover in place on remote signal lost when this happens midway its journey home? I don't know the answer to this; if you do please do share.

What made it start to descent?
Why would the descent stall at 111 metres?

I don't know; but if up link had stayed active for at least 32 seconds after down link was lost; pilot command could easily account for that descent. The Air is supposed to have a descent speed of 3m/s. It only needed to do about 2m/s to descend from 171m to 111 in those 32 seconds.

 

[Meta4]

And if RTH was manually triggered, what is the expected behavior of an AC that is configured to hover in place on remote signal lost when this happens midway its journey home

The flight data shows that the Failsafe Acton was set to RTH.

 

[Deleted member 94047]

The flight data shows that the Failsafe Acton was set to RTH.

In that case, my scenario is absolutely rubbish. Apologies for the side tracking.

 

[sar104]

It's a bit remarkable that the aircraft was even flying at the end. The cell voltages collapsed to under 2 V:



There were numerous battery communication errors in the event stream and I'd suspect that those values were simply wrong if they didn't agree well with the reported motor voltages:

 

[Paul2660]

If it was in "autoland" mode, could not the pilot have countered that up slight upward pressure on the left stick? At 22 meters, around 66 feet out, he or she could see it, and should have been able to manually fly home, or at least try. Sounds like there was still just enough battery to make it back as the drone hovered for a period of time, in view of pilot.

To me this is something all pilots to test, fly in a safe spot, get to 15%, and hover in a safe spot, let the battery drop to 10% and let the drone start to auto land. Then manually defeat is several times to know that it can be done and you still can land it manually. You can't turn it off, but simple upwards stick pressure defeats the landing and you can fly the drone normally. As soon as let the upwards stick pressure go, the drone will start to autoland again.

When a auto landing starts, there is a lot going on, all kinds of sounds from the controller etc. so getting used to this in a safe spot is important.

Paul C