Crash - Flight Log interpretation help please.

[Eastcoast]

Ok, finally after 3 and a bit years I've had my first full drone loss! I was doing some video of the beaches near me in Essex for a client to include in a promotional video. I'd launched from a wide open space on the grass clifftop and had flown out offshore and down a tracking pass of the beach and promenade. A return pass doing the same at a different angle and then a couple of panoramic shots. I spun around to view passing yacht and the started to return onto the cliff.

Without any warning the Mavic flipped on its back, seems to do a roll and then just dropped straight to the ground with its forward momentun carrying it over the tideline into a seawater pool by a groin with a hell of an impact.

The log shows two obstacle avoidances at take-off, which I think were me as I was doing a full control movement test and a stick movement during a pano sequence. Then the last few seconds are all new error message to me.
I'd be grateful for any interpretation help anyone can offer. I assume the insurance company will want to see this file?

Flight Log

I haven't tried it yet but hoping the SD card is still readable to get todays images off.

This is what I'm left with........

 

[Starz]

Is it possible that a prop came off?

 

[sar104]

Okay - that one is very weird. It doesn't appear to be characteristic of a prop or motor problem. There's a hint of an IMU issue but it is definitely going to require the mobile device DAT file, ending FLY065.DAT, to figure it out.

 

[Eastcoast]

I guess so yes. It was my 4th flight of the day in between 3 sites, so I'll be honest that my pre-flights weren't as through as they could have been ☹ I only recovered the props still attached 2 x sets of both blades and 1 x 1 blade, the 4th set I couldn't find in the impact area, so that would make sense.

 

[Eastcoast]

Okay - that one is very weird. It doesn't appear to be characteristic of a prop or motor problem. There's a hint of an IMU issue but it is definitely going to require the mobile device DAT file, ending FLY065.DAT, to figure it out.

If I understand the retrieval guide correctly I may have balked that avenue up as I've synched the DJI app data with my drone logbook. Does that delete the DAT files?

 

[sar104]

If I understand the retrieval guide correctly I may have balked that avenue up as I've synched the DJI app data with my drone logbook. Does that delete the DAT files?

Unfortunately yes - since you are running iOS DJI GO 4 v 4.3.37, that will no longer be available.

This event didn't have any of the characteristics of motor failure or a lost prop, and a couple of indicators of IMU failure. Looking at the pitch, roll and yaw at the end, we see that during the descent there is no evidence of the typical 3 Hz oscillation when thrust is lost from one motor:



Even more interesting is the reported vertical speed:



40 m/s seconds downwards is unphysical, and means at least partial IMU failure. There was subsequently a switch to the backup IMU at 499 seconds, but too late to avoid the crash.

 

[Eastcoast]

Thanks very much for your help.
On the down side, I got complacent and having done flights only 30 minutes earlier with the same unit, didn't do all the checks as through as I should. If a prop did come off then that's perhaps partly on me I guess.
If it was a hardware glitch then there's nothing I could have done, so some comfort there.

I was very lucky in that (a) it hit no-one, (b) I found it (or rather somebody spotted it it and told me were to look) and (c) the SD card survived with the work from the previous flights intact.

Big learning day, DO the pre-flights every time, and use a different card for every flight.

Thanks again.

 

[sar104]

Thanks very much for your help.
On the down side, I got complacent and having done flights only 30 minutes earlier with the same unit, didn't do all the checks as through as I should. If a prop did come off then that's perhaps partly on me I guess.
If it was a hardware glitch then there's nothing I could have done, so some comfort there.

I was very lucky in that (a) it hit no-one, (b) I found it (or rather somebody spotted it it and told me were to look) and (c) the SD card survived with the work from the previous flights intact.

Big learning day, DO the pre-flights every time, and use a different card for every flight.

Thanks again.

As I said - it wasn't a prop.

 

[Meta4]

Without any warning the Mavic flipped on its back, seems to do a roll and then just dropped straight to the ground with its forward momentun carrying it over the tideline into a seawater pool by a groin with a hell of an impact.

The log shows two obstacle avoidances at take-off, which I think were me as I was doing a full control movement test and a stick movement during a pano sequence. Then the last few seconds are all new error message to me.
I'd be grateful for any interpretation help anyone can offer. I assume the insurance company will want to see this file?

It really doesn't matter what the exact explanation for the incident was, there's enough in the data to show that it was a hardware error.
The flight seemed normal until 7:55.5 when it started to show a slight loss of altitude holding and over 10 seconds, lost 25 ft altitude without joystick input.

After 8:04.7 the drone rolled hard and rapidly gained (not believable) speed without joystick input.
The hard roll to one side persisted until the crash.
According to the data, the speed after 8:15.2, it exceeded 100 mph, still without any joystick input !!
This looks like a fault in the way GPS data is being processed.
Together with the altitude issue, it looks like some kind of IMU failure.

At 8:13.9 the drone had wandered back up to 191 ft and started losing height quickly and crashing.

The characteristic feature of loss of a prop or motor is a spiraling descent.
There is no hint of that in the data.

Whatever failed and caused the crash, it wouldn't have been picked up in any pre-flight inspection.