Perhaps one of the users can give a hint or a moderator of this forum forwards the problems to the software development ... I tried to clarify these issues with DJI Support (German), but their answers seem to have been given to me without any expert knowledge at the moment.
First, very briefly (exact analysis was only at home by evaluating the flight data):
Our Mavic Mini crashed after a short flight time (approx. 4m 50s) and a last known charge level of 80%. Distance between starting point (Pstart) and crash point (Pcrash) approx. 430m (1.400 ft).
Phew!
Now the "Find My Drone" is called up immediately and the position (Pdrone) is shown (as an icon on the map and - in white letters - at the bottom left of the smartphone display):
What's this?
Supposedly another 1,400m (4,560 ft) distance to the Mavic Mini, between Pcrash and Pdrone.


OK. Unfortunately, this position (Pdrone) is on the other side of the (wide) river and we had to drive about 40km (about 25 miles) because we needed a bridge.
Unfortunately the Mavic Mini could not be found. The terrain at this indicated position (Pdrone) was very "simple", but there was also no radio connection to the Mavic Mini.
We searched for about 3 hours, but then we drove home to further consider.

With the help of the website "AIRDATA" came insights that did not want to match the displays of the Fly app:
There is no indication in the flight data of the position shown in the "Find My Drone" function (Pdrone)!
All the data sent from the Mavic Mini to the RC are saved in the flight data, right?

How can the Fly app display such a position (Pdrone) that was not sent by the Mavic Mini at all ???

That is the crucial point for me!
Is there a "big" bug in the RC software, along with the Fly app?
At the time of this incident, the software was the current one: 1.0.4 (both for the Mavic Mini, for all batteries, for the Android smartphone, and for the RC, of course).


The next day the weather was not so good. Because it had to be assumed that the battery had been drained in the meantime, i.e. no radio connection to the crashed Mavic Mini could be established, a new search no longer made sense.
The terrain at the crash site is much more difficult and without the help of the Mavic Mini (by beeping and blinking) the search now seemed hopeless ...


The German customer service of DJI, which received all the data (flight data in particular), confirmed the crash at position Pcrash, but said in summary (quote):
"... the chief engineer temporarily determined that the aircraft's GPS module could be damaged by the aircraft accident, and for this reason transferred the wrong GPS coordinate back to the app."

With little attention you could "believe" this, but then this position - caused by a defective GPS module - would have to be stored in the flight data, right?

Any attempt by me to contact DJI's software developers and discuss the issue with them will be blocked.
Too bad, very sad!
One should think that any concrete reference to software problems is gratefully received ...

Perhaps it will now be possible to contact the developers via this forum and provide them with all the information (data) again so that they can find out the reason for these errors and remedy them.

I would be very happy and grateful if this path could lead to success here ...

Frank

Hello and firstly sorry for your loss.
We need the flight log in order to determine what happened and investigate further.
They are stored in your phone, or you can extract them from airdata by login with your DJI login.
At 430m the mini is nearly impossible to be seen, so no flying VLOS could make you think you were closer.
Something to consider in the future is a Vifly Beacon, it would give you 30 hours extra of beeping and flashing to find your drone.
Hope this is resolved asap.

Hi Frank,

Flightlog would be great help, use this > https://www.phantomhelp.com/LogViewer/Upload/
Hope you did not sync your data in the fly app, after syncing the dat file is gone....
(DAT file the subfolder of FlightRecord called MCDatFlightRecords)


cheers
JJB

It's impossible to make any helpful suggestions without seeing the flight data.

JJB* Posted at 1-30 02:40
Hi Frank,

Flightlog would be great help, use this > https://www.phantomhelp.com/LogViewer/Upload/

Yes, I also like to use this website for the analysis of flight data!
In particular, I appreciate the possibility of being able to follow the flight route "step by step" while keeping an eye on the actuation of the control sticks!
The additional display of the "spirit level" was extremely helpful in order to be able to determine the exact position of the crash !!! (roll angle)
Thank you!

Labroides Posted at 1-30 03:14
It's impossible to make any helpful suggestions without seeing the flight data.

Of course, the flight data are required to analyze the events!
On the one hand, I do not know whether there is a way to attach this data here directly, on the other hand, I was not aware of whether it is common to publish this data completely.
In this first post, I was particularly interested in communicating that the Fly app obviously shows data (positions) that are demonstrably not recorded in the flight data ...
Can I personally send you the flight details if necessary?

djiuser_jJFRKC7baMLo Posted at 1-30 04:17
Of course, the flight data are required to analyze the events!
On the one hand, I do not know whether there is a way to attach this data here directly, on the other hand, I was not aware of whether it is common to publish this data completely.
In this first post, I was particularly interested in communicating that the Fly app obviously shows data (positions) that are demonstrably not recorded in the flight data ...

true, good site
But i ment to share your link on here, as the flightlog have more details not shown on phantom help.
About privacy...up to you   or just pm me.

cheers
JJB

djiuser_jJFRKC7baMLo Posted at 1-30 04:17
Of course, the flight data are required to analyze the events!
On the one hand, I do not know whether there is a way to attach this data here directly, on the other hand, I was not aware of whether it is common to publish this data completely.
In this first post, I was particularly interested in communicating that the Fly app obviously shows data (positions) that are demonstrably not recorded in the flight data ...

It's common to post flight data and there's nothing to worry about from doing it.
Or you could PM me if you prefer.
What I'd need is a link to the report that you would get following the instructions in the link in post #2, or the .txt file.

InspektorGadjet Posted at 1-30 01:53
Hello and firstly sorry for your loss.
We need the flight log in order to determine what happened and investigate further.
They are stored in your phone, or you can extract them from airdata by login with your DJI login.

Thank you for the super fast feedback!
Of course the flight details are available at any time.
This data, together with an "understandable" summary, has already been sent to DJI several times (!).
Here are the inquiries registered with DJI:
1)  2535769 from 01/07/2020 (support.ge@dji.com)
2)  2552474 from 01/15/2020 (support.ge@dji.com)
3)  31759 from 01/20/2020 (dev@dji.com)
4)  2562428 from 01/21/2020 (reply regarding 31759, again from support.ge@dji.com)
5)  2566816 from 01/23/2020 (support@dji.com)

Please let me know how I can send you this flight information again.

Your comment regarding the distance of 430m:
I am aware that flying the Mavic Mini at a greater distance is a risk that DJI is not responsible for!
My only concern is that the completely incorrect display of the supposedly last position meant that there was no chance of locating the Mavic Mini where it probably "landed".

On our flights we do the following:
While I operate the controller with the smartphone, my wife observes with binoculars of the type "Nikon ActionEX 10x50 6.5 °", because I fail to pay attention to the messages of the Fly app and also to keep an eye on the Mavic Mini ...

Please keep my main concern in mind while analyzing the circumstances:
In the "Find My Drone" function, the Fly App displayed coordinates of the supposed last position, which are not included in the flight data ...

I would also like to hand over the cover letter that I wrote (with detailed explanations) for your colleagues in China ...

Best regards from Germany
Frank
PS:
The size of the flight data file has unfortunately exceeded the size of the file attachment allowed here, so please let me know where I should upload the flight data.

Labroides Posted at 1-30 04:38
It's common to post flight data and there's nothing to worry about from doing it.
Or you could PM me if you prefer.
What I'd need is a link to the report that you would get following the instructions in the link in post #2, or the .txt file.

I hope that you can take over the data via the link given below:


https://1drv.ms/u/s!Aq-mZlQ_Lsvhg9o2XJ5nFpzYqAOTQg?e=MQPkQh

JJB* Posted at 1-30 04:20
true, good site
But i ment to share your link on here, as the flightlog have more details not shown on phantom help.
About privacy...up to you   or just pm me.

https://1drv.ms/u/s!Aq-mZlQ_Lsvhg9o2XJ5nFpzYqAOTQg?e=96ln4X

djiuser_jJFRKC7baMLo Posted at 1-30 05:21
I hope that you can take over the data via the link given below:

OK ... I've opened it and it looks OK.
Will get back when I've found something.

InspektorGadjet Posted at 1-30 01:53
Hello and firstly sorry for your loss.
We need the flight log in order to determine what happened and investigate further.
They are stored in your phone, or you can extract them from airdata by login with your DJI login.

https://1drv.ms/u/s!Aq-mZlQ_Lsvhg9o2XJ5nFpzYqAOTQg?e=0Cxz1e

djiuser_jJFRKC7baMLo Posted at 1-30 05:22
https://1drv.ms/u/s!Aq-mZlQ_Lsvhg9o2XJ5nFpzYqAOTQg?e=96ln4X

Almost done ... your drone did not cross the river

Labroides Posted at 1-30 05:53
Almost done ... your drone did not cross the river

I agree.
I knew that after analyzing the flight data at home ... But only at home!


But the coordinates were displayed on site
                Pdrone = 50.948902 ° 14.072694 °
(using the "Find My Drone" function in the Fly app) and therefore I think there is a bigger problem with the software of the Fly app or that of the controller ...

djiuser_jJFRKC7baMLo Posted at 1-30 05:22
https://1drv.ms/u/s!Aq-mZlQ_Lsvhg9o2XJ5nFpzYqAOTQg?e=96ln4X

Hiya,

Bad luck, sorry for your loss.

MM was on the way back home, but the RTH height was not high enough to fly over the highest part.
On the way to the river your MM flew left of that point, but RTH is direct line back home.
See my charts.

Yellow marking is where it hit trees i guess,  height line in GE is higher than the flown height, this to make it more visible.

DAT file reveals "impact in air", rpm of all rotors normal up to the impact.
crash : 50.96211500, 14.07156100
cheers
JJB

frankymusik Posted at 1-30 06:02
I agree.
I knew that after analyzing the flight data at home ... But only at home!

I don't think there's any problem with the software.
The data looks very clean with nothing strange about it.

Here's my interpretation:
You flew south and turned back at around 3:58.9.
At 4:32.3 you initiated RTH and left the drone to come home hands-off.
RTH is making towards home at approx 8 m/s until something goes wrong at 4:41.0 when the drone begins to spin (clockwise), tumble and lose height (still in hands-off RTH).
In three seconds it loses 30 metres of altitude and keeps tumbling and spinning.
The data stops at 4:51.3 with the drone still approx 30 metres above the ground.

Falling and spinning indicates that the drone lost a prop or a motor.
The drone will have hit the ground about 30 metres beyond the last data point which was at: 50.96155   14.07162.
The yellow arrow shows you the direction from the last point (but not the distance).
It might be less than 30 metres if it hit a tree on the way down.

JJB* Posted at 1-30 06:05
Hiya,

Bad luck, sorry for your loss.

MM was on the way back home, but the RTH height was not high enough to fly over the highest part.

Are you sure of that?
I calculate that the drone was about 80 metres above the ground level when it ran into trouble.
It's pretty hard to run into a tree when 80 metres above ground level
?????

Labroides Posted at 1-30 06:06
I don't think there's any problem with the software.
The data looks very clean with nothing strange about it.

I understand you perfectly!

But, that's not the real problem here ...
Since the function "Find My Drone" has shown the coordinates
            Pdrone = 50.948902° 14.072694°
we were forced to drive to the other side of the river (approx. 40 km detour).

Only at home we could determine - based on the flight data - where the Mavic Mini had "landed".
But it was too late for a (successful) search.
I see the problem in the position display of the "Find My Drone" function. The above coordinates (Pdrone = 50.948902° 14.072694°) are not saved in the flight data, they were only displayed in the Fly app ...

frankymusik Posted at 1-30 06:20
I understand you perfectly!

But, that's not the real problem here ...

DJI's engineer's explanation might be correct.

Labroides Posted at 1-30 06:14
MM was on the way back home, but the RTH height was not high enough to fly over the highest part.

Are you sure of that?

It's not a flat landscape there!

Yes, the RTH level was too low! No doubt...

The altitude information in the flow data is "relative" in relation to the altitude at the starting point.

I know our mistake: The RTH height should have been set higher !!!
Yes everything is clear! Our problem, of course!

But, the Fly app reports coordinates that the Mavic Mini has never flown because they are not available in the log file ...!

frankymusik Posted at 1-30 06:28
It's not a flat landscape there!

Yes, the RTH level was too low! No doubt...

Yes, the RTH level was too low! No doubt...
The altitude information in the flow data is "relative" in relation to the altitude at the starting point.
I know our mistake: The RTH height should have been set higher !!!
Yes everything is clear! Our problem, of course!

Not at all ... don't be fooled by someone that couldn't read the terrain height.
Read post #19.
You flew from a point that was higher than the terrain you flew over.
When the drone got into trouble, it was 30 metres higher than the launch point and 80 metres higher than the ground below it.

Labroides Posted at 1-30 06:14
MM was on the way back home, but the RTH height was not high enough to fly over the highest part.

Are you sure of that?

80 meters above river level?  MM at impact flew 26 meters higher than the takeoff position height.

26 meters is GPS height data in DAT file, difference between take and flying at impact.

OR - MM hit top of trees/rock, area is for sure nor flat there.
OR - MM air collision with a big bird.

After the impact, in DAT,  a large altimeter error, so not sure if baro height data in txt file is correct.

cheers
JJB

JJB* Posted at 1-30 06:42
80 meters above river level?  MM at impact flew 26 meters higher than the takeoff position height.

26 meters is GPS height data in DAT file, difference between take and flying at impact.

I won't waste much time with someone that doesn't read properly.
Read carefully what I wrote.

80 meters above river level?  
I never mentioned anything about the river
MM at impact flew 26 meters higher than the takeoff position height.
What impact?
He flew from a point that was higher than the terrain you flew over.
When the drone got into trouble, it was 30 metres higher than the launch point and 80 metres higher than the ground below it.

frankymusik Posted at 1-30 06:28
It's not a flat landscape there!

Yes, the RTH level was too low! No doubt...

Hi Franky,

Last recorded position in the DAT file =  14.072.534.411.826.800        50.952.559.128.887.400

cheers
JJB

Labroides Posted at 1-30 06:51
I won't waste much time with someone that doesn't read properly.
Read carefully what I wrote.

Hi Labroides,

Well, you are a person who does not like arguments. So i agree ; best not to react on each oher  We have a deal!  ;-)And if you get angry, calling names to others or reply with fooling etc. Read this ISBN 978-0-114789-0
Well read a post on here how to behave, o yeah  you did react to that post.

Impact, well   that`s info is in the DAT file.  Clarified.
80 meters? just aks you a question 80 meters from where ?  no answer ofcourse.
Loosing a prop on a mini?  yes possible but not likely (i think), loosing a prop means quick rapid in RPM before shuts down, not in the DAT file. Clarified as well.


So why would a MM on the way back home in a good RTH flying at a safe height suddenly dropps out of the sky ?
reason 1 = not a safe height
reason 2 = hit by bird
reason 3 = tech error on the MM
reason 4 = many more, DJI will explain.

BTW this is an open forum, some have less knowledge some have more. But the main thing is to help each other and transfer knowledge.
We (forum users), included me, makes mistakes when try to explain but the intention is always good.
Unfortunately some forum users like to abuse others when replying. Well the book makes it al clear.

Happy many landings,
cheers
JJB

JJB* Posted at 1-30 07:21
Hi Labroides,

Well, you are a person who does not like arguments. So i agree ; best not to react on each oher  We have a deal!  ;-)And if you get angry, calling names to others or reply with fooling etc. Read this ISBN 978-0-114789-0

You 80 meters? just aks you a question 80 meters from where ?  no answer ofcourse.
No answer?
I gave the answer in my initial assessment and told you to read it carefully.
Since you can't find it, I said:
80 metres higher than the ground below it.
What's so hard about that?

And if the drone was 80 metres agl, there's no problem with the RTH height and no collision with terrain, trees or other obstacles.

Spinning and falling indicates loss of a prop or a motor.
I have not speculated on what caused that, but it wasn't RTH height being too low.

Labroides Posted at 1-30 07:30
You 80 meters? just aks you a question 80 meters from where ?  no answer ofcourse.
No answer?
I gave the answer in my initial assessment and told you to read it carefully.

Dear users,
please don't be angry about my problem!

I have attached an overview picture again.
The route recorded in the log file is very clear and clearly recognizable!

The problem for me is that the "Find My Drone" function shows a position (red arrow in the attached picture) that was never flown by the Mavic Mini !!!

Yes, there was a crash due to the return height being too low. Yes, this too low return height was set by me. Therefore, no one else is to blame for the crash / collision!

I now hope that the software developers from DJI load the flight data into the app and smartphone so that they can see why this incorrect position is displayed at all.
This is the only problem that should be solved ...

Please keep in mind that we (users) cannot solve this problem. That is why I have contacted the DJI developers several times, but unfortunately I have not received a satisfactory statement from the (German) support ...

Frank

frankymusik Posted at 1-30 07:49
Dear users,
please don't be angry about my problem!

Yes, there was a crash due to the return height being too low. Yes, this too low return height was set by me. Therefore, no one else is to blame for the crash / collision!
JJB has confused you.
Read my posts.
Your RTH height was good, the drone did not crash into an obstacle on RTH at all.

That is why I have contacted the DJI developers several times, but  unfortunately I have not received a satisfactory statement from the  (German) support ...
The explanation they gave you sounds plausible.

It's really interesting to read you it looks like an airplane black box analysis

Labroides Posted at 1-30 07:56
Yes, there was a crash due to the return height being too low. Yes, this too low return height was set by me. Therefore, no one else is to blame for the crash / collision!
JJB has confused you.
Read my posts.

The analysis of the flight data is really very impressive!
All errors and problems are clearly shown.

I thank you both - Labroides as well as JJB* - for the great effort with these analyzes.
That will give me - and maybe others - a lot of knowledge, without a doubt!

But if the function "Find My Drone" invents a "last position", then you have no chance to search at the actual crash site, you are "sent into the desert" ...

So I hope that DJI developers honestly explain what happened in the Fly app.
The existing bugs can certainly be solved, and that will also help other users.

Frank

djiuser_jJFRKC7baMLo Posted at 1-30 05:09
Thank you for the super fast feedback!
Of course the flight details are available at any time.
This data, together with an "understandable" summary, has already been sent to DJI several times (!).

You are very welcome and I trully hope you can find the mini!

That moment you realise that Labroides and JJB* are not drinking buddies

HomePoint Posted at 1-30 10:28
That moment you realise that Labroides and JJB* are not drinking buddies

lol I know...

frankymusik Posted at 1-30 07:49
Dear users,
please don't be angry about my problem!

nope, i am never angry....

if height is referenced to ground level it means the direct distance craft > ground where its flying.
So flying at 80 meters above ground level means 80 meters craft to surface level below craft.
That`s why they 'invented ' the above mean sea level for height in the aviation world.

Your craft did not flew 80 meters AGL but about 30 (not 27, made a simple calculation fault) meters higher than your takeoff height position.
If that was enough height to stay above obstacles, you know this better knowing your flying area.And as said ; in the DAT file is written "impact in air", so obstacle or bird or mis-use if this message.
cheers
JJB

JJB* Posted at 1-30 11:44
nope, i am never angry....

if height is referenced to ground level it means the direct distance craft > ground where its flying.

if height is referenced to ground level it means the direct distance craft > ground where its flying.
Correct

So flying at 80 meters above ground level means 80 meters craft to surface level below craft.
Correct again

That`s why they 'invented ' the above mean sea level for height in the aviation world.
That's completely irelevant to this incident

Your craft did not flew 80 meters AGL but about 30 (not 27, made a simple calculation fault) meters higher than your takeoff height position.
But the incident did not happen over the launch spot.
It was over land that was much lower.

I have to admit that I made a calculation error myself.
His drone was even higher.
At the time of the incident (4:41) the drone was about 115 metres above the ground below it.
There was no collision with terrain and his RTH height was completely safe and appropriate for the flight.

All those graphs you use to pad your analysis fooled me.
I thought you were smarter than that.
His drone was well clear of obstacles because he was flying 30 metres higher than the takeoff position over land that was much lower than the launch spot.
You seem to have trouble understanding the words so here's a graph that might help.

I always assumed the Mini would RTH at a constant elevation above sea level determined at the moment it locks in its home point.  I figure it calculates the elevation above sea level that equates to the AGL height we have set for RTH at that moment when it locks in the RTH point.

But does the Mini use GPS for all of this, or rely on its barometric sensor, or a combination of the two?

I presume it does not do any "terrain following" when executing a RTH on its own.

Both launch and crash point are approx 280m MSL. @JJB is clearly right with everthing he said.

Sigmo Posted at 1-30 14:54
I always assumed the Mini would RTH at a constant elevation above sea level determined at the moment it locks in its home point.  I figure it calculates the elevation above sea level that equates to the AGL height we have set for RTH at that moment when it locks in the RTH point.

But does the Mini use GPS for all of this, or rely on its barometric sensor, or a combination of the two?

I always assumed the Mini would RTH at a constant elevation above sea level determined at the moment it locks in its home point.  I figure it calculates the elevation above sea level that equates to the AGL height we have set for RTH at that moment when it locks in the RTH point.
Forget sea level and AGL, the drone does not know what they are.
For the drone all heights are relative to the launch point.


But does the Mini use GPS for all of this, or rely on its barometric sensor, or a combination of the two?
Your Mini uses its barometric sensor to maintain height, not GPS.

I presume it does not do any "terrain following" when executing a RTH on its own.
It has no way of knowing where the terrain is below it (beyond the short range of its VPS sensors) and flies straight and level in RTH.

Zelmani Posted at 1-30 15:01
Both launch and crash point are approx 280m MSL. @JJB is clearly right with everthing he said.
[view_image]

You need to check your information because it incorrect.
Even the broad contours on the picture you posted show that.