Welcome Mavic Pilot!
Jump in and join our free DJI Mavic community today!
Sign up

Help find lost Mavic Pro - Flight Log missing final moments

damoj88

Member
Joined
Dec 10, 2019
Messages
6
Reaction score
0
Age
31
Location
Australia
Hi crew,

Looking for some help on how to find my missing DJI Mavic Pro drone.

I was out filming at a wedding venue on Saturday, 10th December with my DJI Mavic Pro at a vineyard when my phone lost visual feed with the drone whilst in the air. During these moments the battery quickly depleted to 14%, and as I tried to call it home it was prematurely forced to land due to a low battery without me having a visual connection to it. I tried restarting the app a couple of times, with no success of it reconnecting the vision to my phone during this time. The battery depleted completely and it landed approximately 100-200 metres away in an unknown location from my home point.

But knowing my Mavic had the 'Find my Drone' software, I thought I could access my flight logs and pinpoint the exact landing location to go and retrieve it.

Issue is, after uploading and reading the flight log, I see it has only recorded it's location up until it had 14% battery life remaining and for some reason it then cuts out with no flight info recorded beyond that, so I have no idea as to it's final landing location via the 'Find my Drone' feature in the app. The last known data point records it as being 35 metres in altitude, with it moving 12km/h as it was attempting to return home, and prematurely landed thereafter.

I went searching in the general vicinity of that last flight point logged, trying to work out a trajectory of where it may have landed, but unfortunately couldn't find it.

I know the remote controller still had connection to the drone as it landed/crashed, and there may be information on that in terms of the last known distance recorded to my home point, which I could use to map out a circumference from the home point.
Otherwise, is there any other data available on the remote controller used alongside my flight log to find out its trajectory and final landing location with GPS coordinates or otherwise?

Thanks to anyone that can provide help!

Cheers!

Here are links to the flight log:

https://www.phantomhelp.com/LogViewer/0GD6W9LSVR7WVY6JGOUA/


 

Fly Dawg

Well-Known Member
Joined
Jul 12, 2019
Messages
284
Reaction score
176
Location
Watkinsville, Ga
I went searching in the general vicinity of that last flight point logged, trying to work out a trajectory of where it may have landed,
Did you look in the trees here?? You were descending right there.

Capture.PNG
 

damoj88

Member
Joined
Dec 10, 2019
Messages
6
Reaction score
0
Age
31
Location
Australia
Did you look in the trees here?? You were descending right there.

View attachment 87804
Yep, gave it a quick look around the edges of the outside branches, but couldn’t see it up there. Was a rather windy day when I looked as well, and should have been knocked out if so.

It also appears that the coordinates have it landing just wide of the tree on the vacant land, but alas, it wasn’t there.

Was wondering if there was data I could rip from the remote control to work out if it was still moving after the drone lost connection with my flight log/DJI app on my phone. From memory it was still getting closer according to the distance getting smaller to my RTH point, which makes me think it was still flying and hadn’t got wedged in a tree upon the loss of connection to the flight log/DJI app.
 

DanMan32

Well-Known Member
Premium Pilot
Joined
Sep 1, 2018
Messages
5,670
Reaction score
2,057
Age
54
Sar would be able to figure out its landing point based on last known position, heading and descent. He would also be able to account for wind drift on falling if it lost power before landing.
 

Meta4

Well-Known Member
Joined
Oct 30, 2016
Messages
4,793
Reaction score
5,478
Age
64
Did you look in the trees here?? You were descending right there.
Yep, gave it a quick look around the edges of the outside branches, but couldn’t see it up there. Was a rather windy day when I looked as well, and should have been knocked out if so.
It also appears that the coordinates have it landing just wide of the tree on the vacant land, but alas, it wasn’t there.
To have snagged the drone at that point, those trees would have to be more than 64 metres (210 ft) high ... and they aren't .
The drone is most unlikely to have come to rest in those trees.
The battery depleted completely and it landed approximately 100-200 metres away in an unknown location from my home point.
1-200 metres is a guess and it's likely to be quite wrong.
But knowing my Mavic had the 'Find my Drone' software, I thought I could access my flight logs and pinpoint the exact landing location to go and retrieve it.
The Find My Drone feature can only show you the last place the app had contact with the drone.
But your drone was still flying after that last data point so FMD won't know where it went.
It also appears that the coordinates have it landing just wide of the tree on the vacant land, but alas, it wasn’t there.
You have data showing the drone flying slowly with the right stick pushed full forward, fighting a significant headwind.
But it shows nothing of any landing or coordinates of any landing spot.
The last known data point records it as being 35 metres in altitude, with it moving 12km/h as it was attempting to return home, and prematurely landed thereafter.
The last data at 19:56.4 shows the drone 250 metres away and 37 metres up, and only making 3.5 metres/sec (7.8 miles/hr) toward home with you driving.
If you lost control signal after that point, as the drone was set to RTH.
Its RTH speed would have topped out at approx 10 metres/sec.
BUT the drone was battling a headwind of approx 12 m/s.
The end result would be for the drone to have been slowly blown backwards until after a short time, the battery level would have reached critical low level and the drone autolanded to prevent reaching zero battery and falling.
The point where it came to rest should be further away and downwind.
Was wondering if there was data I could rip from the remote control to work out if it was still moving after the drone lost connection with my flight log/DJI app on my phone.
The app will have no data for anything which happens after losing signal.
From memory it was still getting closer according to the distance getting smaller to my RTH point, which makes me think it was still flying and hadn’t got wedged in a tree upon the loss of connection to the flight log/DJI app.
Your drone was only getting (slowly) closer while you held the right joystick full forward.
After losing signal .... see above.
 
  • Like
Reactions: damoj88

Fly Dawg

Well-Known Member
Joined
Jul 12, 2019
Messages
284
Reaction score
176
Location
Watkinsville, Ga
To have snagged the drone at that point, those trees would have to be more than 64 metres (210 ft) high ... and they aren't .
The drone is most unlikely to have come to rest in those trees.
Where do you get those numbers? The altitude at that point is 120ft (36meters).
 

Meta4

Well-Known Member
Joined
Oct 30, 2016
Messages
4,793
Reaction score
5,478
Age
64
Where do you get those numbers? The altitude at that point is 120ft (36meters).
Yes ... his drone was 120 ft higher than the launch point.
But the drone wasn't at the launch point, it was 250 metres away from it and the earth is not always pancake-flat.
 

Fly Dawg

Well-Known Member
Joined
Jul 12, 2019
Messages
284
Reaction score
176
Location
Watkinsville, Ga
Yes ... his drone was 120 ft higher than the launch point.
It was a quick and mistaken judgement call from the "apperance" of the actual site. Looking back, there was an 80 + ft drop in the terrain profile from the launch point to the trees. ( From the view, I didn't see that sort of elevation change ) That would put the actual AGL @~ 200ft. So, your theory is correct.
 

damoj88

Member
Joined
Dec 10, 2019
Messages
6
Reaction score
0
Age
31
Location
Australia
The last data at 19:56.4 shows the drone 250 metres away and 37 metres up, and only making 3.5 metres/sec (7.8 miles/hr) toward home with you driving.
If you lost control signal after that point, as the drone was set to RTH.
Its RTH speed would have topped out at approx 10 metres/sec.
BUT the drone was battling a headwind of approx 12 m/s.
The end result would be for the drone to have been slowly blown backwards until after a short time, the battery level would have reached critical low level and the drone autolanded to prevent reaching zero battery and falling.
The point where it came to rest should be further away and downwind.
Thank you for your help.

Given there was 14% battery life left, and the direction and speed at which the drone was travelling, is it all possible to pinpoint a relative location as to where it may have landed?

I recall still having control of the drone with my remote control after my app went black (presumably when the flight log ended at 14% battery). Could the Black Box data in the remote controller be useful in trying to work out the direction/deviation until the battery died completely and the controller lost total connection with the drone?
 

damoj88

Member
Joined
Dec 10, 2019
Messages
6
Reaction score
0
Age
31
Location
Australia
Beaming a Bat Signal into the sky pleading for @sar104 to help with the theoretical location of my drone? If I can pinpoint it to a reasonable circumference/radius that would be amazing!

Thank you to anyone out there that can help analyse the data and provide advice.
 

sar104

Dic mihi solum facta, domina.
Premium Pilot
Joined
May 27, 2017
Messages
9,851
Reaction score
11,645
Location
Los Alamos, NM
Beaming a Bat Signal into the sky pleading for @sar104 to help with the theoretical location of my drone? If I can pinpoint it to a reasonable circumference/radius that would be amazing!

Thank you to anyone out there that can help analyse the data and provide advice.
Apologies - I've was out of commission most of last week. The aircraft was around 25 seconds from starting autoland at 11%:

Battery.png

The problem is not knowing what stick inputs it received during that time. You had full forward throttle at disconnect - do you recall if you continued that?
 

damoj88

Member
Joined
Dec 10, 2019
Messages
6
Reaction score
0
Age
31
Location
Australia
Apologies - I've was out of commission most of last week. The aircraft was around 25 seconds from starting autoland at 11%:

View attachment 88177

The problem is not knowing what stick inputs it received during that time. You had full forward throttle at disconnect - do you recall if you continued that?
Oh wow, how did you build this graph? Thanks for your help mate!

If I can recall correctly, I was at full throttle.

I was wondering, would it be possible for me to rip data from the remote controller/black box file to work out the stick inputs after the flight log disconnected at 14%? I still had connection to the drone with my remote controller after this happened, so if I can grab a record of the stick inputs, I can use it to work out which direction it was flown beyond 14%.

I noticed in DJI Assistant you can rip the Black Box data, but it just persists with a spinning wheel when I try and save it to my computer.

Thanks again for your help @sar104 ! :)
 
Last edited:

Fly Dawg

Well-Known Member
Joined
Jul 12, 2019
Messages
284
Reaction score
176
Location
Watkinsville, Ga
I was wondering, would it be possible for me to rip data from the remote controller/black box file to work out the stick inputs after the flight log disconnected at 14%? I still had connection to the drone with my remote controller after this happened, so if I can grab a record of the stick inputs, I can use it to work out which direction it was flown beyond 14%.
The data you posted is from the remote ( Device ). There is another file ( a .dat ) but as that file won't help either. You only lost downlink from the aircraft ( Video and Telemetry ) which means there is no more data available other than on the aircraft itself. You still had aircraft control ( Presumably ), but no data was being transmitted back to be recorded. As mentioned by @sar104, you were issuing full forward elevator at the end of the data stream, so bottom line is that there is not a way to know how long and/or if you were still in control of the aircraft so determining a location would be a most difficult task and only a guess.

Data End.jpg
 

damoj88

Member
Joined
Dec 10, 2019
Messages
6
Reaction score
0
Age
31
Location
Australia
The data you posted is from the remote ( Device ). There is another file ( a .dat ) but as that file won't help either. You only lost downlink from the aircraft ( Video and Telemetry ) which means there is no more data available other than on the aircraft itself. You still had aircraft control ( Presumably ), but no data was being transmitted back to be recorded. As mentioned by @sar104, you were issuing full forward elevator at the end of the data stream, so bottom line is that there is not a way to know how long and/or if you were still in control of the aircraft so determining a location would be a most difficult task and only a guess.

View attachment 88326
Oh ****! So the 'blackbox' file located within the remote control won't have any additional info on it? I assumed it may have had the throttle controls logged within it, despite the info not being logged in the flight log due to the disconnect.

Thanks for your explanation!

I am but a rookie with extrapolating data (direction, acceleration etc) and guesstimating it's location, can someone help me? @sar104 and @Fly Dawg I would most appreciate your knowledge here.

I have gone ahead and drawn up a possible flight path based on the assumption the drone didn't massively deviate from it's current flight path, and maintained it's last known speed of 7.8mph. As reference, the purple marker is it's last recorded points a second before it lost connection. The red marker is the last recorded point as connection was lost. The blue markers on the yellow line is my extrapolated guesstimate of it's potential flight path and distance traveled. The green marker is the point the drone would have roughly hit 11% battery and been forced to land on that spot. The middle yellow line follows it's last recorded flight path direction, whilst the other two yellow lines branching off either side are ones that help build a (rough) 45 degree perimeter either side to map out the area it could have landed.

DRONE FLIGHT PATH.jpg

Under @sar104's advice that the Mavic was around 25 seconds from auto landing once it reached 11%, I've mapped out approximately 1 second with each marker (25 markers for 25 seconds), using the distance traveled between the purple marker (19m 55.4 secs of flight time) and the last known recorded red marker (19m 56.4 secs) as the average distance traveled between two points - which I then extrapolated out. In addition, the altitude lost between the purple marker and the red marker was 2 feet in that second (down from 122.7 feet to 120.7 feet). Extrapolating this out, if it's going down at 2 feet a second across the 25 seconds, it would mean that the drone lost 50 feet in altitude, and was hovering at 70 feet (21.33 metres) when it hit 11% battery and was forced to auto-land.

Hope that makes sense!

Of course, there are a number of variables that could have occurred that I haven't factored in (ie. battery could have suddenly depleted massively, causing it to land earlier than predicted; dropped altitude suddenly; wind conditions etc.), but this is what I have mocked up not accounting for them, and assuming consistency.

As noted above, you'll see I've added other lines (very loose 45 degrees from the original predicted flight path) in order to map out a circumference area it could have landed, had the drone deviated in any way. But given I searched along each row of the grapevines to no avail (the area between the first and middle yellow potential flight path lines), I am thinking it landed in the tree lines either side of the grapevine rows (as indicated by the second yellow flight path, which follows it's original flight path direction according to last known coordinate).

Would love to see if either of you gurus could help me further, or if anyone else has an opinion on where it could be.

It's been 21 days since I lost it, but I'm not yet giving up hope. Trouble is the location is an hour and 40 minutes drive from my home, so the more data and info I can collect to help map out it's potential location the better it is to equip me for another search.

Many thanks in advance! :)
 

sar104

Dic mihi solum facta, domina.
Premium Pilot
Joined
May 27, 2017
Messages
9,851
Reaction score
11,645
Location
Los Alamos, NM
Oh ****! So the 'blackbox' file located within the remote control won't have any additional info on it? I assumed it may have had the throttle controls logged within it, despite the info not being logged in the flight log due to the disconnect.

Thanks for your explanation!

I am but a rookie with extrapolating data (direction, acceleration etc) and guesstimating it's location, can someone help me? @sar104 and @Fly Dawg I would most appreciate your knowledge here.

I have gone ahead and drawn up a possible flight path based on the assumption the drone didn't massively deviate from it's current flight path, and maintained it's last known speed of 7.8mph. As reference, the purple marker is it's last recorded points a second before it lost connection. The red marker is the last recorded point as connection was lost. The blue markers on the yellow line is my extrapolated guesstimate of it's potential flight path and distance traveled. The green marker is the point the drone would have roughly hit 11% battery and been forced to land on that spot. The middle yellow line follows it's last recorded flight path direction, whilst the other two yellow lines branching off either side are ones that help build a (rough) 45 degree perimeter either side to map out the area it could have landed.

View attachment 89162

Under @sar104's advice that the Mavic was around 25 seconds from auto landing once it reached 11%, I've mapped out approximately 1 second with each marker (25 markers for 25 seconds), using the distance traveled between the purple marker (19m 55.4 secs of flight time) and the last known recorded red marker (19m 56.4 secs) as the average distance traveled between two points - which I then extrapolated out. In addition, the altitude lost between the purple marker and the red marker was 2 feet in that second (down from 122.7 feet to 120.7 feet). Extrapolating this out, if it's going down at 2 feet a second across the 25 seconds, it would mean that the drone lost 50 feet in altitude, and was hovering at 70 feet (21.33 metres) when it hit 11% battery and was forced to auto-land.

Hope that makes sense!

Of course, there are a number of variables that could have occurred that I haven't factored in (ie. battery could have suddenly depleted massively, causing it to land earlier than predicted; dropped altitude suddenly; wind conditions etc.), but this is what I have mocked up not accounting for them, and assuming consistency.

As noted above, you'll see I've added other lines (very loose 45 degrees from the original predicted flight path) in order to map out a circumference area it could have landed, had the drone deviated in any way. But given I searched along each row of the grapevines to no avail (the area between the first and middle yellow potential flight path lines), I am thinking it landed in the tree lines either side of the grapevine rows (as indicated by the second yellow flight path, which follows it's original flight path direction according to last known coordinate).

Would love to see if either of you gurus could help me further, or if anyone else has an opinion on where it could be.

It's been 21 days since I lost it, but I'm not yet giving up hope. Trouble is the location is an hour and 40 minutes drive from my home, so the more data and info I can collect to help map out it's potential location the better it is to equip me for another search.

Many thanks in advance! :)
You can certainly try to bound the final location like that. At full forward elevator the aircraft was moving at 8.5 m/s. If it had continued to do that for another 25 seconds, without rudder or aileron inputs to deviate from it's ground track, the result would be something like you showed - the red line below. I'd start by looking along that line.

1577722157789.jpeg