Major crash, lost comms and RTW went sideways

[RaoulB]

So, it tried to RTW but didn’t go to 50 meters as per settings but flew sideways into the side of a ship and fell into the hold. Major damage, but at least got it back. Will buy the DJI insurance next time. Not sure if there is anything I can do, as it should have gone to 50 meters befor attempting RTW. Anyway, now need a new one...

 

[Meta4]

So, it tried to RTW but didn’t go to 50 meters as per settings but flew sideways into the side of a ship and fell into the hold. Major damage, but at least got it back. Will buy the DJI insurance next time. Not sure if there is anything I can do, as it should have gone to 50 meters befor attempting RTW. Anyway, now need a new one...

That sounds like an interesting case to look into but it might be difficult if you've got no data to show what went on because of loss of signal.
It still might be worth a shot.
Go to DJI Flight Log Viewer | Phantom Help
Follow the instructions there to upload your flight record (txt file) from your phone or tablet.
That will give you a detailed report of the flight.
Come back and post a link to the report it gives you.

 

[RaoulB]

I think this is the right file, let me know what you think.
thanks

 

[RaoulB]

That sounds like an interesting case to look into but it might be difficult if you've got no data to show what went on because of loss of signal.
It still might be worth a shot.
Go to DJI Flight Log Viewer | Phantom Help
Follow the instructions there to upload your flight record (txt file) from your phone or tablet.
That will give you a detailed report of the flight.
Come back and post a link to the report it gives you.

https://mavicpilots.com/attachments/djiflightrecord_2019-03-21_-18-28-41-txt.67413/

 

[Simmo]

@sar104 (Where do you find the initialized yaw values? Does this log show it only a few degrees off at launch?)

 

[Meta4]

So, it tried to RTW but didn’t go to 50 meters as per settings but flew sideways into the side of a ship and fell into the hold. Major damage, but at least got it back. Will buy the DJI insurance next time. Not sure if there is anything I can do, as it should have gone to 50 meters befor attempting RTW. Anyway, now need a new one...

The flight data does not shows you initiating RTH and the Mavic never started the RTH process.
The data also shows very poor GPS cover for most of the flight with GPS health being 5 (best) for seven seconds during the flight.
At 5:35.2 it drops to 0 & 1 for five seconds, again at 8:11.8 for five seconds and at 8:22 till the end of the data at 8:28.
Speeds are very slow with you having very little joystick input after 8:15.9.
It appears that
1. You did not initiate RTH
2. There was unusually poor GPS reception and the drone would have lost position holding and drifted with the wind at times.
3. At the end of the flight you were close to the ship which possibly blocked some of the skyview
4. Signal loss was after the crash which occurred at 8:27.7
5. Speed at the time of the crash was only 2 metres/sec - this seems to be due to drifting with the breeze with no GPS position holding.

The GPS reception is unusually low even when the drone was high enough that it should have had an unobstructed sky view.
Were you flying under a crane that might have blocked GPS or did you have a metallic sticker on the top of the drone?

 

[sar104]

@sar104 (Where do you find the initialized yaw values? Does this log show it only a few degrees off at launch?)

The detailed magnetometer data are only available in the DAT files. The initialized yaw is the initial OSD_yaw value in the txt log, but the actual compass data are not included.

 

[RaoulB]

thanks so much for all the feedback! interesting, I did not initiate the RTH, and also not sure why it slid sideways so quickly, I was only gaining altitude to regain reception, yes, a very metallic environment, so I guess it lost contact, but not sure why it went sideways so fast. As I said I was only going up at the time. So if no RTH initiated by me or the aircraft, doesn't explain the crash. There was no wind during the crash, it was very protected area, and not windy that night.

Interestingly, I have purchased a replacement, I have also bout the care refresh this time... lol. anyway, the new drone, doesn't have the same stability in hover, is there a setting I have missed, or is there a difference between some drones? this one moves a little in hover by itself, now I m very concerned flying near anything.

 

[RaoulB]

The flight data does not shows you initiating RTH and the Mavic never started the RTH process.
The data also shows very poor GPS cover for most of the flight with GPS health being 5 (best) for seven seconds during the flight.
At 5:35.2 it drops to 0 & 1 for five seconds, again at 8:11.8 for five seconds and at 8:22 till the end of the data at 8:28.
Speeds are very slow with you having very little joystick input after 8:15.9.
It appears that
1. You did not initiate RTH
2. There was unusually poor GPS reception and the drone would have lost position holding and drifted with the wind at times.
3. At the end of the flight you were close to the ship which possibly blocked some of the skyview
4. Signal loss was after the crash which occurred at 8:27.7
5. Speed at the time of the crash was only 2 metres/sec - this seems to be due to drifting with the breeze with no GPS position holding.

The GPS reception is unusually low even when the drone was high enough that it should have had an unobstructed sky view.
Were you flying under a crane that might have blocked GPS or did you have a metallic sticker on the top of the drone?

Thanks Meta4, no sticker, and no issues on any other flights, no explanation...

 

[Meta4]

So if no RTH initiated by me or the aircraft, doesn't explain the crash. There was no wind during the crash, it was very protected area, and not windy that night

Without the position holding ability that good GPS provides, the drone has no brakes and continues to move when you take your hands off the sticks.
It's like driving on ice.

 

[CanadaDrone]

Flying with any less than 11-12 satellites is asking for trouble - so far that looks like the cause here. Keep an eye on those and end the flight ASAP if they drop. Assuming Meta4's analysis is correct and the most satellites you ever had were 5 you should never have even taken off. Just keep that in mind for drone #2

 

[Simmo]

@meta is on about GPS Health, not the actual number of sats aquired.

 

[peros550]

Do we know how many satellites were the drone able to see during flight time? Is this information visible in the logs?

 

[Meta4]

Do we know how many satellites were the drone able to see during flight time? Is this information visible in the logs?

Here's some of what'smshown in the recorded flight data:

DJI Flight Log Viewer | Phantom Help

Upload and view flight logs from your DJI drones

www.phantomhelp.com

The data also shows very poor GPS cover for most of the flight with GPS health being 5 (best) for only seven seconds during the flight.
At 5:35.2 it drops to 0 & 1 for five seconds, again at 8:11.8 for five seconds and at 8:22 till the end of the data at 8:28.
There were up to 15 sats and at times the number of sats dropping as low as 7
But the spread of sats must have been poor for a good fix as shown by the poor GPS health.

 

[sar104]

Here's some of what'smshown in the recorded flight data:

DJI Flight Log Viewer | Phantom Help

Upload and view flight logs from your DJI drones

www.phantomhelp.com

The data also shows very poor GPS cover for most of the flight with GPS health being 5 (best) for only seven seconds during the flight.
At 5:35.2 it drops to 0 & 1 for five seconds, again at 8:11.8 for five seconds and at 8:22 till the end of the data at 8:28.
There were up to 15 sats and at times the number of sats dropping as low as 7
But the spread of sats must have been poor for a good fix as shown by the poor GPS health.

Don't forget that "GPS health" is rather misnamed - it's not computed from the GNSS DOP but is actually some measure of the confidence in the sensor fusion results. You can find plenty of examples where yaw errors caused "GPS health" to fall to zero but DOP was low and the position solution from GNSS was very good.

 

[BudWalker]

There is a small chance the .DAT will reveal something significant. It does seem that low navHealth (aka gpsHealth) is loosely coupled with sat count, i.e. low navHealth isn't being caused by a heading issue.


What's interesting is the variation in sat count in a short interval. Here, in less than 0.1 sec the sat count goes from 13 to 10.


@RaoulB can you retrieve the tablet .DAT? Look here to see how to do that
Retrieving a V3.DAT from the tablet

 

[sar104]

There is a small chance the .DAT will reveal something significant. It does seem that low navHealth (aka gpsHealth) is loosely coupled with sat count, i.e. low navHealth isn't being caused by a heading issue.
View attachment 68269

What's interesting is the variation in sat count in a short interval. Here, in less than 0.1 sec the sat count goes from 13 to 10.
View attachment 68270

@RaoulB can you retrieve the tablet .DAT? Look here to see how to do that
Retrieving a V3.DAT from the tablet

I noticed that too. Must be due to the non-integer number of satellites¡

I guess I should have clarified above that low satellite count or high DOP are possible causes of low GPS health.

 

[BudWalker]

I noticed that too. Must be due to the non-integer number of satellites¡
......

Yeah, Yeah. CsvView is pretty lame with the vertical axis scale.

 

[pilifida]

Flying with any less than 11-12 satellites is asking for trouble - so far that looks like the cause here. Keep an eye on those and end the flight ASAP if they drop. Assuming Meta4's analysis is correct and the most satellites you ever had were 5 you should never have even taken off. Just keep that in mind for drone #2

You.re exagerating. At P3S i barely had 9 satelites ever, over 6 satellites would have been enough. The same situation for MP1, 6-7 satelites are enough for a stable flight. Don.t know why MP2 would need 11. Ofcourse the more the better, but still. Under 6 satellites you.re not allowed to go beyond 30 m height (6m height of you have VPS on!).
I am talkind about MP1, don.t know why MP2 would be more satellite hungry ?

 

[CanadaDrone]

You.re exagerating. At P3S i barely had 9 satelites ever, over 6 satellites would have been enough. The same situation for MP1, 6-7 satelites are enough for a stable flight. Don.t know why MP2 would need 11. Ofcourse the more the better, but still. Under 6 satellites you.re not allowed to go beyond 30 m height (6m height of you have VPS on!).
I am talkind about MP1, don.t know why MP2 would be more satellite hungry ?

I am not exaggerating, but feel free to do whatever you're comfortable with It also depends on where the satellites are and how they are distributed in the sky at the time of your flight - there is a mathematical way to guarantee a good signal without knowing all that information, and that is 11-12 satellites. 6 satellites all clumped together for example is not good. There should be at least 8 satellites visible to you at any given time almost anywhere on earth and the Mavic can see 20. GPS health is also a very important metric, which is an indication of signal strength and orientation. Also, if you're flying around with the bare minimum satellites and you lose one, that's a great way to force ATTI mode and potentially crash the drone. 4 would be the absolute minimum requirement for 3D positioning, but you would never want to fly on that few and DJI won't let you in positioning mode.