Lost Mavic Mini In River !!!!! Need your suggestions

[Lone_Warrior_Innd]

Hi All, ,

Its my first post here while searching for Lost MM in River.

Last Sunday 23rd Feb, I was flying a drone in "P mode "over a river with a beautiful Bridge,
I flow almost for good 3-4 mins & capture really good shots. Battery level was 65%.

Suddenly in Mobile Screen, it Shows Aircraft Sensor Error & felt down in river, i try my best to move it up, but i was going down only. Finally submerge in water, due to depth of water, i thought now to jump & bring it up.

I reported this malfunction case to DJI, already have DJI care pack too,

Being new to this area, what expert suggestion whether it was really malfunction or i done some mistake. ( i am using it almost for good 3 months & flown around almost every weekend, never had any issue)


Following is AirDATA of that flight.

Also, i upload youtube video from my cache memory.
YouTube Video Link

Thank you in advance.

 

[new2mavic]

The drone did not behave well in maintaining the altitude right from the beginning. As can be seen from the graph below, there has been multiple occurrences of height drop with no throttle input, sometimes by as much as 4 meters. The last one occurred at low altitude so the drone joined the ducks in the water.

I am not sure why the drone behaved like that. It was flying over water which means the infrared altitude sensor at the belly may not be working well, the GPS signal was weak to none. May be some kind of bugs in the firmware made the altitude control messed up under such conditions ? There has been a report here about the new firmware version ( 400 ) having such problem. The poster claimed that the problem disappeared after downgrading the firmware to 300.



If you can posts the .DAT log file, may be experts here can shed more light onto the incident. Follow this thread to get the .DAT file from the phone : Flight log retrieval and analysis guide

 

[slup]

This isn't unfortunately nothing else then again an example of the "uncommanded descent" from a Mini ... If you have the possibility to attach the DAT log also we can take a look at the motor data. The DAT in question have FLY046.DAT in the end.

The flight were very unstable regarding pitch & roll even though the speeds were pretty low ... the angles maxed to all directions & it was actually hard to figure out which stick mode that was used on the RC.



In the very end of the flight it clearly descends uncommanded ... but then the OP himself put in the final push by command negative throttle by mistake I presume. The VPS sensors behaved ... AC being in sky were reported & no auto landing were started.



The flight wasn't, beside the cold bath for the Mini, thought through properly ... The navHealth were in parts of the flight so bad that I doubt that it found it's way back to HP if RTH had been commanded.

In the end of the flight just before passing under the bridge backwards a lot of error were thrown coming from the poor navHealth, both Position & Velocity errors.
The way back under the bridge is actually depicted like this ... a detour into the woods ...

 

[Deleted member 94047]

This isn't unfortunately nothing else then again an example of the "uncommanded descent" from a Mini

I am not seeing that slup. The issue with the AC not maintaining a stable altitude seems to me to be the result of a poor GPS reception. GPS health starts poor, improves intermittently during the period marked in grey below, and degrades to a level where the FC is no longer using it at around the 304s mark. The AC was using vision for navigation towards the end of the flight. You can see that the descents as well as the climbs are in response to throttle input from the pilot. Sorry for the poor graph, couldn't make it better.

 

[slup]

Pretty sure that it is... GPS shouldn't affect height. Furthermore the VPS were locked in the whole flight. Add to this the flight instability. But we will see once we get the DAT

 

[Deleted member 94047]

GPS shouldn't affect height

Are you certain about that?

 

[sar104]

Are you certain about that?

I've seen no evidence that GPS is used for altitude control - that appears to use just accelerometers and barometric pressure.

 

[slup]

The VPS & barometric based height tracks each other pretty consistent ... especially in the end when the dive occurs. That indicates that the height changes were sensed approx equal by 2 systems ...even though the height figures differ as the AC flies below the HP.

It's nothing that indicate that the VPS were fooled by the water as reporting being in Sky & no auto landing have been initiated.

Looking at the throttle commands relative the height differences ... looks to me like a really unstable & sluggish height keeping, a lot of height changes without throttle. In the end it goes down uncommanded until the very end were the OP makes a stick mistake by commanding descent.

Keeping my fingers crossed for that the DAT is possible to retrieve ... & finds it way here ?

 

[new2mavic]

Are you certain about that?

It's easy to verify. Turn off sensors at the belly and fly a drone indoor ( so no GPS ), the drone will drift around but the altitude can be maintained reasonably well.

 

[sar104]

It's easy to verify. Turn off sensors at the belly and fly a drone indoor ( so no GPS ), the drone will drift around but the altitude can be maintained reasonably well.

To be clear - that would confirm that the aircraft can control altitude without GPS, not that it doesn't use GPS if available.

 

[Deleted member 94047]

I have seen entries like the below in DAT event stream logs

313.907 : 20530 [L-FDI]NS(0) GPS(0): fault on , height_drift

I have also seen similar entries followed by a message about height control failing. I understand this could mean anything but I assumed altitude data from GPS was being factored in altitude control.

 

[new2mavic]

I have seen entries like the below in DAT event stream logs

I have also seen similar entries followed by a message about height control failing. I understand this could mean anything but I assumed altitude data from GPS was being factored in altitude control.

I fly paragliders and I use baronmeter data for most of the time because the precision is much better than that of GPS. Barometers can detect altitude changes in cm compared with meters for GPS. The disadvantage of barometer is that the reading will drift over time so it need be corrected by using GPS as the reference.

 

[slup]

I have seen entries like the below in DAT event stream logs

I have also seen similar entries followed by a message about height control failing. I understand this could mean anything but I assumed altitude data from GPS was being factored in altitude control.

@Doppler ... I do my very best to chose wordings like "It's my belief" or I'm pretty sure" cause I'm far from fully educated ... if I ever will be

I also saw below entries in the log ... but if it's any difference between those 2 or not, effecting height keeping or not?

But still have a "belief" ... waiting for the DAT to either get confirmation or get smashed & some learning.

 

[Deleted member 94047]

@Doppler ... I do my very best to chose wordings like "It's my belief" or I'm pretty sure" cause I'm far from fully educated ... if I ever will be

I also saw below entries in the log ... but if it's any difference between those 2 or not, effecting height keeping or not?

But still have a "belief" ... waiting for the DAT to either get confirmation or get smashed & some learning.

View attachment 95177

I could be very wrong. All I have is just my assumptions. I too am a learner

 

[steve_lyt]

if it was a metal bridge I could have developed magnetic interference. the other possibility is that during the flight it lost too much height . the drone loses height during forward and backward flight in manual mode, unless height is added to compensate for the loss. look at the airdata it say you max height was about 11ft

 

[RodPad]

Have you guys seen this mini type problem "uncommanded descent" over land?
Every one I come across has been water, whether they go for swim or not.


@Lone_Warrior_Innd,
Where did you go?

I you have not synced your mobile device with DJI yet, do it later, we can explain why.

Rod

 

[slup]

Have you guys seen this mini type problem "uncommanded descent" over land?
Every one I come across has been water, whether they go for swim or not.
...

Yep ... several, actually think I've seen more cases over land ... or if it's 50-50

 

[steve_lyt]

if you use waypoints it tends to keep the height fairly consistent. if you are in anything but p mode in my case on my mavic 2 pro there are no safety sensors operating

 

[Bebopperoo]

I have

Have you guys seen this mini type problem "uncommanded descent" over land?
Every one I come across has been water, whether they go for swim or not.


@Lone_Warrior_Innd,
Where did you go?

I you have not synced your mobile device with DJI yet, do it later, we can explain why.

Rod

I have seen several posts warning pilots not to sync their go4 app with dji.
I noticed that syncing is used by dji to change the app, without your knowledge, specifically mine was changed to prevent me using my Spark with a false gps, even though i never updated the app from 4.16.
Please can you advise why you advise NOT to sync "till later" . I think this is more significant than we realise.

 

[slup]

...
Please can you advise why you advise NOT to sync "till later" ...

It's due to a theory that during sync the Fly app deletes the DAT log needed to see motor & raw sensor data, important in for instance cases regarding suspected uncommanded descents for the MM's.

It appear not to be fully consistent though ... some users haven't been affected but many has, this to a degree that it's been hard to get hold of MM DAT's.