Greetings folks,
On behalf of the people who flew the drone, I just want to ask what happened in this flight log of the Mavic 2 Pro

By the accounts of their story: They were doing a test flight prior to taking it out, after a few minutes and around the time that they wanted to return to home, the drone just hovered and did not land even when they tried to land it.
Next they initiated a CSC in an attempt to cut the power to land it (08:03). It did not work.
Then they tried to take it up to a small height, and the drone started to drift, and they tried to initiate a CSC again (08:44), but this time it seems like it lost control and plummeted in to a shallow pond.

Now the question is, could any one figure out what went wrong through the logs: DJIFlightRecord_2020-07-15_[17-01-38].zip (947.87 KB, Down times: 3)

2020-8-9 Upload

Click on the file name to download attach
Flight_record_July

https://www.phantomhelp.com/LogViewer/S0RSPUZENCKIOJJ4998W

I've tried recounting the steps on airdata and it seems at some point it stop responding to the remote.

Really need to know as we are planning to either purchase a replacement or fix the current drone.
If we fix it, there's no telling that this may happen again due to some fault currently in this drone.

Thanks for the help.

Upload the flight log here: https://www.phantomhelp.com/LogViewer/Upload/ Then copy the URL to the complete analysis and come4 back here and post it.

https://www.phantomhelp.com/LogViewer/S0RSPUZENCKIOJJ4998W

Hiya,

Interesting flight, well....

Started the flight with only 5 satellites and ZERO GPS reception.
So after take-off in OPTI mode.
At 1m6s 10 sats and 4 for reception : HP set
But GPS reception dropped to 2, thus again in OPTI mode.
RC down input shows weird behaviour, see chart3, its of/off down.

CSC shut off does not work when craft is not in a "error".

Do not know why the autolanding did not land right away, mayby waiting for user input to confirm landing site is OK?
If craft is close to the HP it will not fly back to HP but does land where it is hovering/flying when autoland start.If that is above a pond if will 'land' there.
Autolanding initiated by the smart battery.

The second CSC, while craft was in steady hover, made the craft move away [from 7.4 meters to 9.7 meters relatibe to HP ] in a spinning move....see chart1a.

So for the next flight with a new drone ; wait for GPS lock, test flight in a wide open area without ponds  and enjoy the flight.

cheers
JJB

JJB* Posted at 8-10 04:12
Hiya,

Interesting flight, well....

This is really helpful information.
They were test flying in a very confined space, basically near a converted warehouse, along with tall buildings on both sides.

I'm not too sure why the first landing attempt did not go through as it was not me that was there on the occasion but I will have to ask the person flying it. But it was very close to the home point hence they said that it didn't come down for some reason, but maybe they forgot to confirm it on the screen. I'll update when I get more info.

As for the second CSC, the veering off to the side, is it a common occurrence? See i thought it would take 3 seconds to initiate a CSC but from the recording, it was only less than a second after the controls were initiated but they lost total control after that.

Paradigm Shift Posted at 8-10 07:59
This is really helpful information.
They were test flying in a very confined space, basically near a converted warehouse, along with tall buildings on both sides.

Hi,

CSC does not shut down the motors if drone is not in a emergency....crashing into etc...
Not sure if the Mavic Pro 2 has the option to choose the CSC behaviour, MA2 has that option where CSC can be executed on a good flying craft.

If not (or not selected) drone just acts on the RC inputs, as it did in this flight, down flight in a spiral move....
They lost control, while actually they initiated or confirmed the 'stopped' autolanding.




see this video  

cheers
JJB

For the the CSC to work it probably needs the response to be in "Always" mode. (there's a similar situation with the Mavic Mini and I have CSC'd a mini onto grass, it took around 2 seconds for the CSC to stop the motors, the drop was about 18 inches)
There have been instances of drones failing to obey a throttle closed 'and land' command at point A when they are trying to RTH to point B. I am not sure if that is applicable in this situation.

JJB* Posted at 8-10 11:03
Hi,

CSC does not shut down the motors if drone is not in a emergency....crashing into etc...

Your video is excellent and helps clarify that CSC doesn't engage at all unless as stated only in an emergency.
By your demo, and flying near a lake, i think that's exactly what happened to the drone.
As for the M2P I would have to see on whether the CSC could be turn to always on next time.
These were new operators and I think in the moment of panic and near obstacles they weren't able to execute proper judgement, thus resulting in an error.

Sean-newbie Posted at 8-10 12:11
For the the CSC to work it probably needs the response to be in "Always" mode. (there's a similar situation with the Mavic Mini and I have CSC'd a mini onto grass, it took around 2 seconds for the CSC to stop the motors, the drop was about 18 inches)
There have been instances of drones failing to obey a throttle closed 'and land' command at point A when they are trying to RTH to point B. I am not sure if that is applicable in this situation.

I'm not sure if the return to home was initiated but what they did mention was they were about a meter or so above the ground, near to the operator and they tried to land it (throttle closed) but it didn't respond.
But as to what JJB said, it may have a command initiated that asked them to confirm landing and they might not have seen it, hence failing to land.

Also as I recall or seen in a few videos.
If you're taking off, to get home point to set in precisely, you'd have to fly above 10m upon take off. Now I'm not sure if it for certain but that's what they demo.
If you're less than 10m, then a rough estimation of home point will be set. I think in a 5m radius to the take off point?
Could anyone confirm this?

Paradigm Shift Posted at 8-10 17:57
Also as I recall or seen in a few videos.
If you're taking off, to get home point to set in precisely, you'd have to fly above 10m upon take off. Now I'm not sure if it for certain but that's what they demo.
If you're less than 10m, then a rough estimation of home point will be set. I think in a 5m radius to the take off point?

No, that is not correct. In a normal situation, once you start the motors, the aircraft will attempt to set a home point immediately, and you should wait until it is set before attempting to take off. In your case, the tall buildings probably made it difficult to obtain a good satellite lock. The home point should ALWAYS be set before you leave the ground.

Hello there Paradigm Shift. I am sorry to read and to know what happened to your DJI Mavic 2 Pro. Since this unfortunate incident happened and there was a crash that happened before. It would be best to contact our DJI support team at https://www.dji.com/support . We have a professional team that will do there best to assess the said drone and give out the best resolution for this issue. If the said drone is still under the warranty period and it is evaluated as a non-pilot error, the warranty service will be provided. Again, I am sorry for this trouble and please stay safe always.

It is my understanding that in order to get the "precision" RTH (it's very good) the first bit of the flight MUST be a pure verticlal up or above to either 7m or 10m, I can't remember which.
If the drone is moved laterally before the necessary ceiling has been reached or crossed then precision RTH is not set and can not be set and the 'normal' accuracy of GPS is applicable. That said the latter accuracy seems to be within maybe 3m for me but it is rare that I would actually land via RTH, I normally use it as a get teh drone overhead tool.
In what follows I am not arguing with Geebax's comment, but of late my drones are struggling to get GPS lock at my at-home takeoff point so I am taking off before they have a good lock. That take off point is surrounded by trees etc. so there is nowhere for the drone to go but up, I quite frequently see it set or reset the homepoint as it climbs into sight of more GPS satellites and gets a better lock. I am deaf so can not hear any verbal messages given out and do not screen watch during launch or landing so I can not say what written messages appear duing takeoff.

With regard to refusing to land, there are 'two' other things that might have been in play (they may be part and parcel of the same behaviour). Firstly the drone will not descend from 'height' and land all in one continuous constant speed descent, it will at least pause/hover for maybe a second or so and assess the landing area, your friends might have encountered this assessment pause. If the drone does not like the landing area it will refuse to land but asks for confirmation that it is to land, I have never followed through with that process so do not know how it completes, I shift the drone by a few feet to a better spot.  Your friends might have encountered this and not seen messages or perhaps heard them if there are any.

I would like to thank all for the support and comments on the mishap of the flight.
We've come to the conclusion that further and more precise training is needed and safety measures will be taken into a pre-flight and during flight plan.
It was a costly mistake but the users now know that in case of an emergency, panicking won't solve anything.
Once again, to all the amazing replies, thank you for helping solve this.

Paradigm Shift Posted at 9-8 18:38
I would like to thank all for the support and comments on the mishap of the flight.
We've come to the conclusion that further and more precise training is needed and safety measures will be taken into a pre-flight and during flight plan.
It was a costly mistake but the users now know that in case of an emergency, panicking won't solve anything.

Many happy landings in the future!

cheers
JJB