International Date Line - any experience crossing this with a 2Pro?

[sar104]

This definitely requires the mobile device DAT file (FLY045.DAT).

How to retrieve a mobile device DAT file: How to retrieve a V3.DAT from the tablet

It does look like a coding error, but the IMU data are also strange.

 

[redge57]

There's something quite strange going on in this flight.
Although the GPS position never changes after hitting the wall and the GPS Speed shows zero after 0:45, your IMU is registering horizontal speed in response to your inputs with the right joystick.
It's as if the GPS has stuck.
At 0:45.5, just after hitting the wall you see the message: Switched to Backup IMU
The GPS Position NoMatch messages that follow, are usually meaningless.
At 0:45.6 there is a momentary indicated speed of 482.6 metres/sec with corresponding crazy speeds showing for Velocity Y & Velocity Z.
These figures are a glitch and the drone did not move that fast.
But the pitch data suggests that the drone did not stop when it "hit the wall" and kept going.


Looking at your description, you say the drone was out of sight, yet the distance shown was only 100 metres away?
I'm wondering if it was perhaps somewhere else and it was moving as the IMU and pitch data indicates?
This is also supported by he IMU speed data corresponding to your joystick inputs.
But the GPS position was locked at -16.79197232 179.9999967 which caused the screen map position to look frozen.
Do you remember what you saw on screen?
Did it look like the drone was flying, did the view change?

Hi, Been offline for a while and appreciate all the responses. The view changed in response to my rotation on the same plane but I could not determine environmental change in response to my (apparent) forward movement. further updates below...

 

[redge57]

With more head-scratching and staring at data, I think that your drone's GPS froze as you passed through the 180° meridian.
The drone kept flying and responding to your joystick inputs.
But you weren't watching when this happened and were confused by the false position and distance the app was showing you.
If you had been watching the drone, you should have been able to bring it back as the data indicates it was responding appropriately to your controls and crossing the 180° meridian had no effect on the drone's flying - just the GPS.
RTH would have not worked, but manual flying using the screen for guidance would have been unaffected.

This is definitely a strange flight and there aren't many places on land where it's possible.
It looks like something DJI never anticipated or corrected in their programming.
It would be interesting to test - but only keeping the drone close and over land.

You're correct - RTH when activated, indicated about to land, however the position of the drone was ~100 m from me, thus I aborted this on a few occasions. All happening in such a short time and with the adrenalin flowing some actions now considered were far from top of mind! Given time of day and lowering light towards the end of what was planned as a 5 minute flight, turning on the downward-facing lights may have assisted in locating the a/c, possibly spurring an attempt at retrieval by dinghy if the manual control attempts to fly home had not worked. All speculation.

The upside is that DJI carried out analysis. I indicated to them prior to the conclusion of their analysis that I had viewed data through both Airdata and DJI Flight Log and cut and pasted some of the tabular data. They have provided a 100% coupon for online replacement and I await delivery.

Thanks for your responses and head scratching - which app would you recommend to view the .dat file for iMac?

 

[redge57]

You're correct - RTH when activated, indicated about to land, however the position of the drone was ~100 m from me, thus I aborted this on a few occasions. All happening in such a short time and with the adrenalin flowing some actions now considered were far from top of mind! Given time of day and lowering light towards the end of what was planned as a 5 minute flight, turning on the downward-facing lights may have assisted in locating the a/c, possibly spurring an attempt at retrieval by dinghy if the manual control attempts to fly home had not worked. All speculation.

The upside is that DJI carried out analysis. I indicated to them prior to the conclusion of their analysis that I had viewed data through both Airdata and DJI Flight Log and cut and pasted some of the tabular data. They have provided a 100% coupon for online replacement and I await delivery.

Thanks for your responses and head scratching - which app would you recommend to view the .dat file for iMac?

Further comment - strong suspicions that the code does not cater for longitude changes from - to + in a matter of metres. I raised this as a question but it was not directly answered by the (very sympathetic and supportive) support person. She offered to replace my SD card if I could supply the proof of purchase, which I couldn't, but appreciated the gesture.

 

[Deleted member 94047]

You're correct - RTH when activated, indicated about to land, however the position of the drone was ~100 m from me, thus I aborted this on a few occasions. All happening in such a short time and with the adrenalin flowing some actions now considered were far from top of mind! Given time of day and lowering light towards the end of what was planned as a 5 minute flight, turning on the downward-facing lights may have assisted in locating the a/c, possibly spurring an attempt at retrieval by dinghy if the manual control attempts to fly home had not worked. All speculation.

The upside is that DJI carried out analysis. I indicated to them prior to the conclusion of their analysis that I had viewed data through both Airdata and DJI Flight Log and cut and pasted some of the tabular data. They have provided a 100% coupon for online replacement and I await delivery.

Thanks for your responses and head scratching - which app would you recommend to view the .dat file for iMac?

Thanks for the update and glad you got DJI to own up to their blunder. If you could upload the .DAT file here, it would make for a really interesting analysis and subsequent read. Sar104 and Budwalker will have a look I am sure with the scenario being as exceptional as it is and they are the best when it comes to doing postmortems on ill fated flights.

 

[redge57]

Any further analysis on these files would be interesting - I have a number of contacts who will be sailing on cats in the location of the loss and have mentioned this issue as a warning. Attached are .txt and .dat files

 

[Deleted member 94047]

@sar104, @BudWalker really looking forward to see your analysis with this one.

 

[Meta4]

Given time of day and lowering light towards the end of what was planned as a 5 minute flight, turning on the downward-facing lights may have assisted in locating the a/c, possibly spurring an attempt at retrieval by dinghy if the manual control attempts to fly home had not worked. All speculation.

I suspect that your drone's flight was not affected and it could have easily been flown home ... if you knew where it was.
It was just the displayed position and distance that malfunctioned.
But that's a good result with DJI for you.

 

[sar104]

Any further analysis on these files would be interesting - I have a number of contacts who will be sailing on cats in the location of the loss and have mentioned this issue as a warning. Attached are .txt and .dat files

None of those files is the relevant DAT file. See post #21 above.

 

[redge57]

Thanks @sar104. Procedure detailed in #21 followed - the MCDatFlightRecords directory is empty.

 

[sar104]

Thanks @sar104. Procedure detailed in #21 followed - the MCDatFlightRecords directory is empty.

That's most unusual under iOS - are you quite sure it's empty? In any case, without the DAT file this is going to remain unexplained.

 

[gnirtS]

Next obvious question, does it do the same here with latitude as you pass the equator?.....

 

[Thomas B]

Next obvious question, does it do the same here with latitude as you pass the equator?.....

No.

 

[gnirtS]

Someones tested this?

 

[sar104]

Someones tested this?

I haven't seen such a test, but the significant difference between those two cases is that latitude is continuous as you cross the equator, while longitude is not continuous across the 180° meridian.

 

[redge57]

Update - Warranty replacement (not DJICare replacement) arrived on Monday. DJI had offered to replace the 128Gb card if I had a receipt - thought that was a good touch.

I also got remaining DJICare transferred from the old A/C to the new one. Many thanks for assistance and opinions - great forum!