Mavic 2 Pro.. 1 Min crash... :(

[sar104]

Granted, I wasn't in an open field in the middle of the day, I was trying to get a sunset shot, but with such a comedy of errors, loss of control right after takeoff, inaccurate RTH, Flight data and IMU, we just cannot easily account for all these anomalies.

UPDATE: DJI has finally advised me that flight data is not stored on the drone itself. Their customer service department definitely misled me repeatedly and intentionally where this is concerned. Even when I asked them bluntly they refused to give a straight answer and were content with me ending our phone calls knowing that I had the wrong information, or maybe their customer service agents don't know the answer either. One person even said he's not authorized to give me that information.

So If during a flight the Occusync connection were to fail, there would be no adequate records of the flight and hence a resulting crash would be considered as not covered, due to lack or loss of data.
If the DJI Go app were to fail, or the cell phone become disconnected from the RC and a crash was to happen, even if it's not due to pilot error, it would be considered as not covered due to lack or loss of data also.
A clear Win-Win for DJI.

So there is no onboard DAT file on the Mini? @BudWalker - can you confirm that?

 

[Meta4]

UPDATE: DJI has finally advised me that flight data is not stored on the drone itself. Their customer service department definitely misled me repeatedly and intentionally where this is concerned. Even when I asked them bluntly they refused to give a straight answer and were content with me ending our phone calls knowing that I had the wrong information, or maybe their customer service agents don't know the answer either. One person even said he's not authorized to give me that information.

The problem here is that you assume that the DJI help people that you are talking to are knowledgeable and giving you accurate information.
Most do not fly drones and have no experience and little understanding.
They read from scripts.
If you push them, what you'll get is often inaccurate and sometimes flat out wrong.
The drone does have it's own data recorder and it continues to record even if the controller is disconnected.

So there is no onboard DAT file on the Mini?

It was a Mavic 2 pro

 

[sar104]

The problem here is that you assume that the DJI help people that you are talking to are knowledgeable and giving you accurate information.
Most do not fly drones and have no experience and little understanding.
They read from scripts.
If you push them, what you'll get is often inaccurate and sometimes flat out wrong.
The drone does have it's own data recorder and it continues to record even if the controller is disconnected.


It was a Mavic 2 pro

Oh - so it was. Well those certainly have onboard DAT files.

 

[gberry]

I wonder if the micro usb was damaged in some way, contributing to the device connection problem.
It's a know problem with the square controller input and people attempting to put them in reversed.

But of course this should not have affected flyability with controller and aircraft.

I'm not sure about how to solve this with DJI, if there is not enough data for the forum analysts to give feedback.

Maybe escalate it and have some brief bullet pointed facts to give, the most important how close the aircraft was to home / you, and how you didn't have any control.
Without logs though, one can see DJIs hesitation in accepting responsibility, anything could have happened after disconnection.

There is definitely no more useful info on the drone logs ?

Perhaps the experts can work on the best list to present to DJI, but it's looking a bit shaky for them to cover.

I appreciate that DJI is in an uncomfortable position and so too am I. They are focusing on absent data and writing the entire thing off as user error, and by extension that would include the crash as well, even though the USB connection is a known issue. Basically, once you don't have the full flight records, it must be your fault and we are not covering it. "Use better cable next time".

 

[MAvic_South_Oz]

loss of control right after takeoff, inaccurate RTH,

Remember, with the RTH on the M2 drones . . .

1. RTH at more than 20m - the drone will rise to pre set RTH alt (if not at or below that alt already), return to home, and land automatically.
2. If between 5m and 20m from home point, the drone will fly directly back to home at that altitude, and land.
A couple of alternatives there if RTH is disabled.
3. If closer than 5m form home point, the drone will land where it is.

If you were in option 2 scenario, this may have meant an uncontrolled descent into whatever.
As you posted you were 100' above you, this is very possible, and if 3 above was the base (within 5m) the result would be almost the same.
in 3 above, it would not return to home point, but land above where it is. If that 5m was over a tree etc . . . well it would land in it without controller input for you to avoid things.

Though if REALLY in RTH, it should descend and land in failsafe ok, not flying erratically, but it could still hit anything below it in option 3.

 

[BudWalker]

So there is no onboard DAT file on the Mini? @BudWalker - can you confirm that?

I believe @gberry was referring to a M2 in which case there are onboard .DATs that are encrypted the same way a MA .DAT is encrypted.

The Mavic Mini also has onboard .DATs. I'm pretty sure these are encrypted as well but using a different scheme than the MA/M2 encryption.

 

[MAvic_South_Oz]

Basically, once you don't have the full flight records, it must be your fault and we are not covering it. "Use better cable next time".

This is the scripted response they are told to do no doubt.
And without decent review like you can get here, many would accept it and move on.

If you can put together a factual bullet pointed list of why this should be covered, then escalate it, you will get someone that knows more, and understands the listed arguments, if they are helpful to your case.

 

[RayOZ]

I absolutely appreciate all input and I ensure that you have my utmost respect as well as all the other members of the forum. May I humbly ask what the effects of open field and better lighting would have on the loss of control ability of the drone?
I'm definitely seeking answers that we may all learn more as I have since joining the group, not necessarily trying to poke holes in the reliability of the drone (though there may be some) because even if DJ I were to replace the drone completely, we would still not know why I lost complete control of the drone seconds after takeoff. I hate the uncertainty and it may cause me to pack my running shoes on my next flight.

A lot of new drone owners have been in the same position, me included. Trying to get their drone up as soon as they can, got home late and getting dark, so they do it at home or in their back yard.
How would it have helped flying in an open field in broad daylight? Loss of comms, RTH means it’ll attempt to land. It’s visual sensors would be working and it wouldn’t have clipped a tree branch.

 

[gberry]

Oh - so it was. Well those certainly have onboard DAT files.

This is getting weirder by the minute, as I mentioned that I had an update from DJI this morning to confirm that the drone does not save flight data, persons in the forum commented that it does. So I was just going back to my Yahoo for an email from the representative named "Kamil", Where he or she stated specifically that the drone does not save flight data. To my astonishment that email from Kamil Is missing! I still have my lengthy reply to this person but the incoming email is missing and I did not delete it. I'm specific at gathering all my correspondence with them and I would not delete anything that they sent.

 

[Meta4]

This is getting weirder by the minute, as I mentioned that I had an update from DJI this morning to confirm that the drone does not save flight data, persons in the forum commented that it does. So I was just going back to my Yahoo for an email from the representative named "Kamil", Where he or she stated specifically that the drone does not save flight data. To my astonishment that email from Kamil Is missing! I still have my lengthy reply to this person but the incoming email is missing and I did not delete it. I'm specific at gathering all my correspondence with them and I would not delete anything that they sent.

That's not weird at all.
You are only confusing yourself if you assume that DJI's help people know what they are talking about.
They can deal with common, easy incidents most of the time, but when you get away from their scripts, what they'll tell you is a lucky dip.
They are not a reliable source for accurate information.

Going over the short flight record that we have for this flight, we have:
8.2 seconds of data showing you launching vertically and climbing to 66 ft
No data for 20 seconds due to losing downink
One second of data when downlink was restored.
It shows that you continued climbing during the 20 seconds "blackout".

From 32 ft to 66 ft, the drone should have maintained horizontal position because you only pushed the left stick up.
But the drone is showing slight horizontal motion, suggesting windy conditions?

We have nothing to show what happened after 0:29.3 when the data stops, except that you've said that it descended a short distance from the launch point.

We can't tell what control inputs you made after that point that might have been a factor in the incident, or what the drone was doing.
Or how or where RTH was initiated
But DJI's data analysis people could and what they have seen would factor into their decision.

Additional confusion comes from DJI's data analysis people not communicating much (but they are usually right) and the help people giving misleading information.

We can't give you a conclusion because of the lack of data for the critical part of this flight, but given what we do have, user error is a plausible assessment.

 

[gberry]

That's not weird at all.
You are only confusing yourself if you assume that DJI's help people know what they are talking about.
They can deal with common, easy incidents most of the time, but when you get away from their scripts, what they'll tell you is a lucky dip.
They are not a reliable source for accurate information.

Going over the short flight record that we have for this flight, we have:
8.2 seconds of data showing you launching vertically and climbing to 66 ft
No data for 20 seconds due to losing downink
One second of data when downlink was restored.
It shows that you continued climbing during the 20 seconds "blackout".

From 32 ft to 66 ft, the drone should have maintained horizontal position because you only pushed the left stick up.
But the drone is showing slight horizontal motion, suggesting windy conditions?

We have nothing to show what happened after 0:29.3 when the data stops, except that you've said that it descended a short distance from the launch point.

We can't tell what control inputs you made after that point that might have been a factor in the incident, or what the drone was doing.
Or how or where RTH was initiated
But DJI's data analysis people could and what they have seen would factor into their decision.

Additional confusion comes from DJI's data analysis people not communicating much (but they are usually right) and the help people giving misleading information.

We can't give you a conclusion because of the lack of data for the critical part of this flight, but given what we do have, user error is a plausible assessment.

I appreciate your input and analysis of this. I did go up after launch because there's not much to do in terms of horizontal manoeuvering. I was aiming to get a panorama sunset shot with brightly coloured clouds, so all I needed was pretty much altitude, then I would take about eight photos and come back down. I appreciate that the community has resources to get to the nitty-gritty of data, It's unfortunate that I don't have more data or insight to offer.

 

[MAvic_South_Oz]

I did go up after launch because there's not much to do in terms of horizontal manoeuvering.

This could be the signal connection issue with the drone and controller.
It can be difficult maintaining connection directly above, note usually ok if moving directly above, because the two don't lose connection for long if at all, but just going straight up can be a problem unless very close attention is given to the antenna positioning.
Lying on a sun lounge may help 90 degrees to the drone, otherwise it's a bit unnatural keeping the controller in such a position.

Looking at @Meta4 input above, it was only 34' when you lost the device view, and you still had stick control to 64' when it was totally disconnected from your control ?
That should be reasonable in any case, but still could have caused loss of drone - controller connection.

Coupled with the blackout of the device, it was very unfortunate.

 

[JAW]

I believe what you were inferring was that DJ I somehow deleted one of their emails off of your computer, no?

 

[gberry]

This could be the signal connection issue with the drone and controller.
It can be difficult maintaining connection directly above, note usually ok if moving directly above, because the two don't lose connection for long if at all, but just going straight up can be a problem unless very close attention is given to the antenna positioning.
Lying on a sun lounge may help 90 degrees to the drone, otherwise it's a bit unnatural keeping the controller in such a position.

Looking at @Meta4 input above, it was only 34' when you lost the device view, and you still had stick control to 64' when it was totally disconnected from your control ?
That should be reasonable in any case, but still could have caused loss of drone - controller connection.

Coupled with the blackout of the device, it was very unfortunate.

Another insightful analysis that I can definitely wrap my head around. The drone was about 10 feet in front of me at the time of spin up, and I was seated about the hight of two cinder blocks laying flat, I just tried to recreate my seating position by sitting on the edge of my bathtub with RC in hand and looking skyward, and it is naturally tilting upwards about 30 to 40°, and the antenna angle may vary of course. To hold the RC level while seated like that would be very uncomfortable on my wrists, but this is still plausible as an explanation for signal trouble. Something to think about if I ever get the chance to fly again.

 

[bonmot]

I have used RTH on this drone very seldom, and it is very accurate, but I do prefer to land manually, nothing to do with safety or anything, just like the feeling of being in control. Even launching the drone I prefer to use the sticks.

I am like you, only used RTH twice so far. Yes, very accurate. Amazing job DJI has done. But after reading more about problems/confusions about RTH, I guess I need to use it more often and try to understand it better when conditions allow.

 

[gberry]

I believe what you were inferring was that DJ I somehow deleted one of their emails off of your computer, no?

Not wanting to draw that inference, cause I can't see how this would be done.

 

[JAW]

Not wanting to draw that inference, cause I can't see how this would be done.

Understood

 

[gberry]

I am like you, only used RTH twice so far. Yes, very accurate. Amazing job DJI has done. But after reading more about problems/confusions about RTH, I guess I need to use it more often and try to understand it better when conditions allow.

Please do, We're all here to learn from each other and improve our skills. It's a pity that I had to come here with a brain teaser as opposed to some helpful hints or something more productive, but I'm sure there are a few small nuggets here that are worth keeping.

 

[MAvic_South_Oz]

Not wanting to draw that inference, cause I can't see how this would be done.

Not just a closed email 'tree' I guess ?
You probably know this if using some email applications, this one MS outlook.




vs open

 

[gberry]

UPDATE: I found screenshots that I had of the email in question. It is of no consequence now, but this further drives my point that DJI customer service will frustrate you for no reason. I offer my apologies to the community here because I said earlier that DJI stated that flight records are not stored on the drone, when in fact what they did was to tactfully evade that question again. The first screenshot is my email to them and the second is their reply.