Need help. DJI won't cover flyaway crash. Ignores flight data.

Joined
Feb 13, 2017
Messages
10
Reaction score
2
Age
37
Hey Everyone!

I'm sorry if this isn't the correct area for this post. If it isn't just let me know. This post will be kind of long....

So I've been wanting to get into this hobby for ages now. I finally saved enough and was persuaded by my wife enough to finally buy a drone (P4). I flew it three times before it flew away on it's own and crashed into a river....

Here is the flight data:

Here's the video of the crash:

The following is the complaint I eventually sent in to the BBB. It describes the incident in detail:

I purchased a DJI Phantom 4 drone on December 31, 2016. I had flown it outdoors on two occasions before this incident, and completed all recommended firmware updates.

On January 6, 2017 I was flying the drone in a regional park, in GPS-mode, with 12-16 satellite connections. I commanded the drone to fly forwards, parallel to a nearby river approximately 20m (65ft) from the river. At that time I observed the drone roll to the right and began travelling north-west while accelerating. I briefly attempted to stop the flyaway by imputing a roll-left command to counteract the drone's trajectory, to which the drone ignored. About 10 seconds after the flyaway I initiated the Return to Home (RTH) function. The drone ignored this command and continued its looping trajectory over the river. The RTH had no effect on the drone's trajectory, nor did it attempt to ascend to the pre-set altitude of 70m. I then cancelled the RTH function and back to GPS-mode which resulted in no change in trajectory. The drone did not hold its GPS position and was now heading south-east over the river. The drone then pitched-up and impacted the river at 45km/hr. I was able to retrieve the drone from the river.

I called DJI (waited on hold for 1.5hours) and started a claim. I sent in the drone as per their instructions. On January 27, 2017 I received an email from DJI claiming among other things that "unit was responsive", "flight environment was near a house...", "user was not familiar with RTH logic", and "No fly away Non warranty Repair. They also claimed that RTH was initiated at 953 seconds and that drone "clipped an object in air" and fell to ground at 962 seconds, despite my detailed description of the event included with the returned drone. My flight data clearly shows that the drone hit the water at 962 seconds, not an in-air object. DJI quoted me $1388 to repair a defective drone when a new complete drone package costs only $1200 on their website.

It's important to note three things:

1. That DJI ignored the 10 seconds of flight data before the RTH was initiated which clearly shows a flyaway situation.

2. The nearest house is 110m (360ft) away from where the drone was flown.

3. The drone was flown in a farmland area.

I called DJI and spoke to "AJ" (CSR) who advised that I could not speak to any technicians directly but did transfer me to a supervisor. I spoke to a supervisor "Oscar L". Despite advising Oscar that the event began well before the RTH command, he continued to fixate on the RTH command and insisted it was pilot error. Oscar stated that DJI drones are not 100% failsafe or accurate but stated that my case was certainly pilot error. Oscar further stated that he was not trained to interpret flight data. Oscar indicated that a drone flying away from its operator, contrary to operator commands (this is supported by in-app flight data) does not constitute a flyaway. Oscar advised that a secondary evaluation could be done but if no discrepancy in flight data is found, that the case would be closed without any process for appeal.

Oscar sent screen-shot of flight data to me via email. The flight data clearly indicates at least two errors starting at 940 seconds, exactly when the drone began it's flyaway event. Again, three lines are highlighted in the screenshot, starting with the RTH command. No mention of the errors.


I have the flight data from my app, and the onboard video to support the events that occurred on January 6, 2017.

So that's the nuts and bolts of what happened. I originally got the following from DJI: Imgur: The most awesome images on the Internet

After my BBB complaint I got the following response:

After reviewing this flight analysis with our senior technicians, the cause of the crash will remain as the unit received normal GPS signal and was responsive. The Return to Home feature was activated and it clipped an object in air during auto landing. The conclusion was based on a combination of a poor flight environment, lack of malfunction and lack of familiarity with the Return to Home Function logic. Based on these factors, we were unable to reverse the warranty decision of this repair.

We do apologize for any inconvenience and thank you for your continued patience on this matter.

Best Regards,

DJI Support

Again, they ignored everything that happened before the RTH....

So here I am... What do I do now?

Thanks!!
 
Upload your DJI flight log to healthydrones and/or phantomhelp log viewer and post a link back in this thread so we can have a look at it.
 
...what a mess. I hope they make good with you eventually! I'd love to hear what the experts on here will have to say about the data
 
  • Like
Reactions: Helihover
The first thing that catches my eye is all that metal in the bridge. The initial flyaway was probably caused by interference with the compass caused by the bridge. Just my impression. I personally would not have flown my bird there. Especially underneath it like that.


Sent from my iPhone using PhantomPilots
 
I can tell you from experience , that once DJI makes ( up ) their conclusion , they will not change it , no matter what proof you bring forward .
 
  • Like
Reactions: Dennis3257
I don't see any farmland and see a bridge full of rebar and I-beams. Could have been the compass went haywire and to me it looks like the bird started to go out of control at the start, the camera was almost facing the landing gear. Did you calibrate the compass from the same location maybe?
 
  • Like
Reactions: Suhail78
Need to see the phantomhelp.com and/or healthydrones logs.
Didn't notice any compass error warnings (or any warnings) in the app and the flight trajectory with no recorded stick input is a worry.

Also, the altitude stayed at 2m after RTH - very odd

Edit again - the Phantom direction doesn't match the map direction arrow so a compass error looks quite likely - even though its not shown in the app.

Lots of steel in that bridge.
 
Last edited:
  • Like
Reactions: sinxa
Was that vid the whole flight? It definitely hit a branch of the bush at the side of the water, didn't it? What were you doing flying so close to the ground?
 
That sucks! I hope that a look at the actual logs shows some good news. Hard to tell what happened from just the videos. Could be interference from the road/bridge or temperature difference from the water/land boundary. I know that the RTH function can work differently if you are close to the bird.

I have had a fly away and was lucky enough to control it and had space above me to gain enough height to avoid some trees as I had switched to atti mode when the RTH function seem to fail (or behave different than I expected).

I hope DJI helps you out after further review.

Following this one.
 
The first thing that catches my eye is all that metal in the bridge. The initial flyaway was probably caused by interference with the compass caused by the bridge. Just my impression. I personally would not have flown my bird there. Especially underneath it like that.


Sent from my iPhone using PhantomPilots

Are bridges like this that much of an influence over the Phantom?
 
It says RTH was at t=953 in the DJI report, assuming that's in seconds that would be almost 16 mins into a 1:29 flight? I don't get that?
 
  • Like
Reactions: sinxa
Phantom Log Viewer - PhantomHelp.com

Location of incident: Google Maps

As for the video. It was corrupted off the the SD card. This was all I was able to recover (thankfully).

I hear the concerns with the bridge, however DJI never mentions the bridge (besides noting the Home point was next to it). Their concerns are the nearby "houses: and "wood". In google earth you can clearly see this is farmland and the nearest house is 110 meters away.
 
Are bridges like this that much of an influence over the Phantom?

Oh my yes!! The bridge has lots of steel in it. The metal can cause the compass to malfunction. Once that happens the bird then has no idea which way home is. It is never a good idea to get that close to a bridge. Flying underneath it is also going to block gps signal. Hate to say it but the more I watch this and think about it, it is pilot error.


Sent from my iPhone using PhantomPilots
 
Also, it didn't hit the grass stalks at the side of the river. It was close but didn't impact. At the beginning of the flyway I was nowhere near those stalks.

And why would a compass error cause a roll command?
 
Also, here what DJI sent me after the first call. It's a copy of their logs:Imgur: The most awesome images on the Internet

E9ZJn
 
And why would a compass error cause a roll command?

Perhaps not a roll command per se, but it couldn't keep itself from rolling away...
 
Also, it didn't hit the grass stalks at the side of the river. It was close but didn't impact. At the beginning of the flyway I was nowhere near those stalks.

And why would a compass error cause a roll command?
Did you calibrate the compass near the bridge by any chance?
 

Recent Posts

Members online

No members online now.

Forum statistics

Threads
143,094
Messages
1,467,600
Members
104,980
Latest member
ozmtl