DJI Mavic, Air and Mini Drones
Friendly, Helpful & Knowledgeable Community
Join Us Now

Total loss of control during roof thermal inspection

Looks like it was in ATTI mode. IMU or compass failure?
 
Would you be willing to share your Flight Logs before you send it back to DJI?

I'm moving this over to the Crash/Fly Away section.

Allen
 
  • Like
Reactions: RodPad
it seems to be a hardware problem, which is rare with Mavic 3T.

It seems that the problem first start at ard 0:47 (timing as in the video, same for below) and the gimble malfunction at first (pointing upward towards the sun and caused the sunburn protection) but by then the drone seems to be in "working" condition and can still fly straight. So, the first question is: did you manually stopped the program flight and direct it to return home?

After the program flight was stopped, the drone seems to be spinning in the air (from 01:28-02:00)...so, it may be a program error or hardware error.... and this bring us to the second question: did you do anything to stop the spin or it just stop after sometime?

And the flight path of flying back was not straight in any sense, so, am I correct that wasn't inentional?

Since drone deploy was used, would you be able to share the parameters of the flight program?
 
it seems to be a hardware problem, which is rare with Mavic 3T.

It seems that the problem first start at ard 0:47 (timing as in the video, same for below) and the gimble malfunction at first (pointing upward towards the sun and caused the sunburn protection) but by then the drone seems to be in "working" condition and can still fly straight. So, the first question is: did you manually stopped the program flight and direct it to return home?

After the program flight was stopped, the drone seems to be spinning in the air (from 01:28-02:00)...so, it may be a program error or hardware error.... and this bring us to the second question: did you do anything to stop the spin or it just stop after sometime?

And the flight path of flying back was not straight in any sense, so, am I correct that wasn't inentional?

Since drone deploy was used, would you be able to share the parameters of the flight program?
So when the gimble went out and the thermal protection kicked it I just pressed the return to home button, hoping to get it back and see what the issue is. Immediately after I pressed it it began the spinning. I cancelled that and entered sport mode to manually take over the flight, but the controls were very intermediate, the drone would occasionally just shake and spin, regardless of input. Very weird.

One thing worth noting- every once in a while when I would boot the drone up the gimble would not respond to up and down commands. Everything else functioned perfectly. I would power the drone down and back up- issue would go away. This would happen about 1 in 8 bootups. Looking back I should have reached out to DJI, not sure if it's related to the incident or not.
 
Thank you! I was legit thinking I was going to do some damage to it.. very odd behavior. Hope to get a replacement in a couple days so I can finish the job lol.
Most people never practice in ATTI mode or no GPS mode. They last about 90 seconds prior to crash. Back in the day, GPS runaway was common and if you didn't switch to ATTI, the aircraft was gone. I don't think that's what's happening here, but you made the right moves to get it back safely. Also a good example why there is a VLOS ( visual line of site) rule.
 
Most people never practice in ATTI mode or no GPS mode. They last about 90 seconds prior to crash. Back in the day, GPS runaway was common and if you didn't switch to ATTI, the aircraft was gone. I don't think that's what's happening here, but you made the right moves to get it back safely. Also a good example why there is a VLOS ( visual line of site) rule.
Indeed- my first drone was a XP2 Quadcopter- 100% manual flight, no GPS, no auto stabilization, maintaining visual and having the ability to orientate and control is a good skill to have.
 
  • Like
Reactions: VikingwithDrones
Here are the last 2 logs from the RC controller and the Drone Deploy log. I've never analyzed a flight record so no idea if there is data here that is helpful. The second log was my post incident test flight where it was doing fine, then suddenly lost its mind again and I struggled to get it back. [I clicked "attach files" but they are not showing on the message, help?]
 
  • Like
Reactions: RodPad
Hello, there are downloadable but not sure if they are usable.

Have you looked at this section?

Rod ..

Adding, I just watched the video, my stomach was? I would have failed, WOW!
And that thermal camera yes!
 
Last edited:
  • Like
Reactions: Multipleshots
Okay here is a link to the logs from DJI and Drone Deploy. I would certainly appreciate any insight!!
Here's what that flight data looks like:

I can see the drone rolls right and then left without an obvious reason from 3:03.7.
Four seconds later the gimbal starts jumping around, showing crazy pitch, roll and yaw numbers for 14 seconds.
The drone itself is wobbling around quite a bit while this is happening.
The only warning messages are to say that the gimbal is at its limits.
Things appear to have returned to normal at 3:21.5.

This looks like one for DJI to look into.
 
Here's what that flight data looks like:

I can see the drone rolls right and then left without an obvious reason from 3:03.7.
Four seconds later the gimbal starts jumping around, showing crazy pitch, roll and yaw numbers for 14 seconds.
The drone itself is wobbling around quite a bit while this is happening.
The only warning messages are to say that the gimbal is at its limits.
Things appear to have returned to normal at 3:21.5.

This looks like one for DJI to look into.
Thank you- that was the short test flight I did after the Drone Deploy incident, purposely over some shrubs in the event I had to set it down while it was behaving poorly. I sent the flight logs to DJI, replacement should be here tomorrow.
 
  • Like
Reactions: RodPad and BigAl07
I'm missing something.

If somebody knew how to convert the data I figured it would happen. The link that Meta4 shared works, the data rolls like normal, etc, but the aircraft doesn't show up visually at all. I tried a different browser also.

I'm not wanting take over this thread, send me a PM?

Rod ..
 
Last edited:
  • Like
Reactions: BigAl07
I'm missing something.

If somebody knew how to convert the data I figured it would happen. The link that Meta4 shared works, the data rolls like normal, etc, but the aircraft doesn't show up visually at all. I tried a different browser also.

I'm not wanting take over this thread, send me a PM?

Rod ..

Zoom in on the map a bit, it should show the triangle (aircraft) during playback.
 
  • Like
Reactions: RodPad
I'm missing something.

If somebody knew how to convert the data I figured it would happen. The link that Meta4 shared works, the data rolls like normal, etc, but the aircraft doesn't show up visually at all. I tried a different browser also.
I see (don't see) that sometimes.
Ifind that refreshing the browser window usually fixes it.
 
  • Like
Reactions: RodPad
There are 3 logs in the google drive: (listed in chronological order)

  1. DJIFlightRecord_2023-10-70_[12-05-29]
    This appears to be the record of a test flight, immediately before the incident flight. Everything seems normal.
  2. Drone Deploy.log
    This log covers the incident flight.
  3. DJIFlightRecord_2023-10-70_[12-24-19]
    Another flight conducted immediately after flight 2.

Flight log 2 was converted into zipped csv (by AirData) and attached below. Here is some highlight of incident in Flight 2:

14866300 (466.3 sec) Gimbal roll up
14903500 (503.5 sec) flight mode change from GPS_CL to GPS_HomeLock and the gimbal continue to act like Linda Blair in Exorcist.
14908500 (508.5 sec) drone elevated without stick input
14922100 (522.1 sec) flight mode change from GPS_HomeLock to AutoLanding
14926200 (526.2 sec) First stick input, but subsequent input value looks abnormal (value exceeding 364-1684, and never 1024, not tally with the movement of the drone)
15063600 (663.6 sec) Landing

For sure the problem was not caused by pilot as there were no stick input at that time. As similar problem persists even in manual mode, program error is also unlikely. Hardware problem is the most likely cause then.

May I further ask if the OP had paired the same remote controller with other drones before? Did that happen to other drones too? Or just with this drone?

Also, there is a DAT file in the controller which may further help investigating into the cause of incident. Here is how to retrieve it (extracted from sar104's original post):

The DAT file naming convention, based on the date and time of the start of the file, is: YY-MM-DD-hr-min-sec_FLYXXX.DAT., where XXX is the flight recorder file index from the HOME_dataRecorderFileIndex field in the txt log.

These are retrieved by the same method as the TXT logs. Under both iOS and Android they are in a subfolder, MCDatFlightRecords, in the folder that contains the TXT logs.
 

Attachments

  • Oct-7th-2023-12-09PM-Flight-Airdata.zip
    168.3 KB · Views: 7
Last edited:
  • Like
Reactions: mavictk and RodPad
Lycus Tech Mavic Air 3 Case

DJI Drone Deals

New Threads

Forum statistics

Threads
131,133
Messages
1,560,156
Members
160,105
Latest member
anton13