First of all, let me preface the above statement by stating the entire case here.  

So about a month ago, after quite a few days of rain, it was a beautiful day outside and I wanted to fly my Spark to take a picture from the sky.  I went to a small local park nearby and set up the drone.  I remember setting the drone on the ground in front of me and turned it on.  At first there was a delay in recording the home point.  So I let it hover for a while until I got about 9 Sattelites connected and the RTH point was recorded as usually.  

I flew the drone as I usually do.... just practicing several maneuvers.  Then I wanted to take a picture of the city from the sky.  So I took it up to about 80m and the aircraft disconnected completely from the controller, thus there was no video feed, and no control of any kind.  I thought it's no big problem since the RTH function was engaged.  
The aircraft never returned to the take-off point however.  

I looked for it for two whole days and could not find it.  Eventually I noticed on the flight data that the home point was somehow erroneously recorded one block away and on top of a building (still have not been able to get to it).  I contacted DJI support, sent them the flight data (which clearly shows a discrepancy between the homepoint and where the controller was at the star of the flight.  It also clearly shows that (on the screen) the aircraft disappeared from next to the recorded homepoint, and suddenly appeared next to the controller; so there was obviously a glitch), but DJI support claimed that no error could be diagnosed.  I sent screenshots of the flight data, showing exactly the erronousely recorded home point, the flight path from the home point to where I was actually standing in the park (apparently straight through some buildings), the battery level, and connection strength.  I very clearly and with visual evidence refuted every single claim the DJI support staff made that there was no error recorded, asking for explanation every time, but they just offered me a 30% discount on a new aircraft - at the same time the 36% Spring Sale was going on, thus adding insult to injury.  

To say that I am extremely frustrated and utterly dismayed by this so-called customer service is a big understatement.  

the home point has been updated. Please check it on the map!

Home point recorded, RTH Altitude = 30m.

I think he was referring to the "Please check it on the map!" part of that statement.

Unfortunately for you, it seems like the drone functioned as it was supposed to and returned to its home point. The home point was in the wrong spot because you were flying in a place with weak GPS signal. Hopefully you can get access to the roof of that building and it will be waiting for you up there. Also check the side of the building for wreckage in case the building or anything between you and it is taller than 30m.

Thanks for your answer!  Well, there was no such statement or message on the screen.  It only said in green “Home Point Recorded”.  NOW I know that it’s better to check on the map every single time before taking it up that the correct home point is recorded - even if it was recorded correctly in the same spot every time so far up to that fateful day.  I’m also still curious why the drone lost connection completely despite having 18 satellites connected at the time when it disconnected.  
There is no damage on the side of the building as it is a slanted roof and the drone is sitting near the top on the slanted side opposite the park, thus leading me to believe that it must’ve either landed or hovered there before the battery ran out.  There doesn’t seem to be any visible damage to the roof or the building.  I am trying to organize a long enough ladder to see if it’s possible (with the homeowner’s permission, of course) to get up there and knock the drone down with a stick.  

bodhi09 Posted at 2018-4-23 17:34
Thanks for your answer!  Well, there was no such statement or message on the screen.  It only said in green “Home Point Recorded”.  NOW I know that it’s better to check on the map every single time before taking it up that the correct home point is recorded - even if it was recorded correctly in the same spot every time so far up to that fateful day.  I’m also still curious why the drone lost connection completely despite having 18 satellites connected at the time when it disconnected.  
There is no damage on the side of the building as it is a slanted roof and the drone is sitting near the top on the slanted side opposite the park, thus leading me to believe that it must’ve either landed or hovered there before the battery ran out.  There doesn’t seem to be any visible damage to the roof or the building.  I am trying to organize a long enough ladder to see if it’s possible (with the homeowner’s permission, of course) to get up there and knock the drone down with a stick.

Sorry for the unexpected experience. Please provide us your case number so we can look into the case and help you solve your problem. Thanks for your cooperation.

Sorry for your loss, I failed to get the case number via your Forum account, could you please post the case number here so that we can check more details of the data analysis and make it clear for you? Thank you very much.

DJI Mindy Posted at 2018-4-23 20:29
Sorry for your loss, I failed to get the case number via your Forum account, could you please post the case number here so that we can check more details of the data analysis and make it clear for you? Thank you very much.

Hi Mindy,

Thanks so much for your reply!  Here’s the case number:

CAS-1581296-L9Z3N4 CRM:0430000006086.  

Also here are the pictures that show the aircraft clearly recorded a wrong RTH point BEFORE the weak GPS connection warning came up.  Notice that the aircraft appears at first at the wrongly recorded Home Point, and then appears next to the position of the controller where I was standings and from which point the aircraft actually took off.  
Please also note that the GPS returned to normal soon after and I was connected to about 8 or 9 Sattelites while flying, and even 17 when the complete disconnect happened.  

DJI Elektra Posted at 2018-4-23 20:25
Sorry for the unexpected experience. Please provide us your case number so we can look into the case and help you solve your problem. Thanks for your cooperation.

Thank you very much for replying, Electra.  

Please see the case number and some pictures illustrating the issue below.  

Please note the extremely straight “flight path”, and the fact that the aircraft would’ve had to go straight through at least two high rise buildings (at an altitude of about 4m nonetheless) to reach the spot in the park marked by the blue dot where I was actually standing (right next to the aircraft, btw).  It was plain and simply some kind of software glitch that recorded the home point in an erroneous location about one block from where the aircraft was actually located and took off from, thus leading to this troublesome situation for me.  I didn’t fly the aircraft much when I got the Weak GPS signal warning, just kind of hovering to see if the GPS would lock on.  When everything was green (enough satellites connected, the weak GPS warning gone, I proceeded to fly it as normal, thinking that it was all OK.  There was no more indication that the RTH point was actually recorded in a completely wrong location, nor that there was any way the aircraft could disconnect (at the time it was connected to 17 Satellites) from the controller completely.  
I’ve tried telling this to customer service via three long, and painstakingly detailed messages already.  

bodhi09 Posted at 2018-4-23 21:19
Hi Mindy,

Thanks so much for your reply!  Here’s the case number:

Thanks for the case number, I will confirm with our analysis team about the actual situation, please allow me some time, will come back to you soon, thank you.

Thanks for the patience, I have confirmed with the team, per the flight records shown the APP, the home point was recorded with strong GPS signal. If you believe the home point is wrong, please provided us more evidence like video and so on, we will forward to the team for further analysis.
At 0:08, the GPS became weak, you said the aircraft appears at first at the wrongly recorded Home Point, and then appears next to the position of the controller from 0:07-0:08. Since the weak GPS signal, the coordinate is empty, therefore, the flight record was jumped to the next coordinate that the GPS signal recovered. Please note at 0:08, the distance has become 94.6m, and during 00:08-00:25, you kept fully pushing the joystick forward until 00:25, the GPS signal got strong, which means the position of the aircraft was changed because you kept pushing the drone forward instead of there is malfunction with the drone.


At t=02:28, h=79m, d=108m, the aircraft triggered the RC LOST RTH, pilot set the return height 30m, the aircraft started to RTH with the current height, the flight record interrupted at t=02:30, h=79m, d=106m.
I think currently, your concern is that the home point is different from the place where the drone took off, but per the flight records, this point cannot be verified unless you provide more evidence, we have provided the best offer in this case, hope your kind understanding.

Thanks so much for the input, Mindy.

If you look on the first picture I provided you, the original Home Point is right on the roof of a 2 Story building.  The park is behind 2 additional buildings.   
The Height at the time shows 4.6m.   
  
I’ve also provided a video of the flight path, please have a look at that.  
  
1. How did the aircraft start from a slanted roof (as indicated on the map) on a 2 story building if the home point was not recorded incorrectly?   How would someone even get up there to launch the drone from?  I thought the aircraft can only take off from an even surface? Why is the RTH point different from where the controller (as clearly indicated by the blue dot?)  Why would I climb up to a 2 story building with a slanted roof, set the aircraft down, somehow make it to the park behind two more buildings and not within direct line of sight of the controller and then take off?  Doesn’t make much sense, does it?  Not only that, but at a recorded height of 4m through two more high rise buildings somehow fly the drone back to the park in a completely straight line (obviously interfering with the building on the right) without damaging the drone on the side of the building, All without moving (the controller buttons can’t push themselves, I assume) my physical location from the park?  

2. How did the aircraft travel through 3 buildings (2 of which are high rises) at a height of 4.6m to the park (where I was standing) in the span of 2 seconds?  Please note the battery percentage and the flight time on the video between the sudden shift in position.  

3. Why is it so difficult for you guys (and by that I mean so far everyone at DJI support) to see that very clearly the home point was somehow recorded incorrectly by the software when it is as clear as day there was some kind of malfunction?  Can you imagine how utterly frustrating this is for me trying to get the company making these devices attuned to the possibility of a non-user error malfunction in their product?  Please watch the video I provided via the link.  

* Please look extra close at the flight path indicated in yellow (too straight to be normal - especially through the buildings).  
* Please note that the blue dot (indicating the controller) is in the park, not in a direct line of sight of the aircraft (through buildings, a parking area, etc.).  Keep in mind that as indicated at the time, the **height was 4.4m.**  
* Please note that the home point was recorded **BEFORE** the Weak GPS warning came up!
* Please note that the aircraft “jumped” from the RTH point to the position in the park in about 2 seconds, that **the position of the nose was exactly the same orientation just before and just after the jump**, and that the battery percentage did not change even one percent in that time frame (while it changed rather quickly in the remaining time), indicating that there was no actual distance traveled, leading to the OBVIOUS conclusion that the only possible explanation was a software error!!!   
* Please note that the aircraft disconnected at around 80m.  

So, from my side there were two fatal flaws that day...   
Number 1 the RTH point was recorded incorrectly by the aircraft.  
Number 2 for some reason the aircraft disconnected completely at 80m despite having 17 connected Satellites at the time, and triggered the RTH command, thus returning to the wrong location (which very clearly was incorrectly recorded by the aircraft before Takeoff).  

I don’t know what other evidence I can possibly provide from the flight records that there was indeed a software or some other malfunction, and that the home point recorded incorrectly, leading to this extremely frustrating situation!  Please JUST LOOK at the video and see what’s so plainly obvious in front of you!   
Sorry if I sound a bit harsh, but you have no idea how utterly frustrating this whole ordeal has been for me - especially if I have to type the same thing over and over again to DJI support and have to point out the most obvious details!  

https://www.dropbox.com/s/fpos2w ... 2021-14-47.MP4?dl=0

bodhi09 Posted at 2018-4-25 04:44
Thanks so much for the input, Mindy.

If you look on the first picture I provided you, the original Home Point is right on the roof of a 2 Story building.  The park is behind 2 additional buildings.   

The home point can be only recorded when the GPS signal which is different from take-off point. Also, the take-off altitude will be recorded as 0, so the altitude displayed in the flight records is the relative altitude. I understand your feeling, but the issue you referred cannot be verified in the flight records you offered. Therefore, the conclusion remains as inconclusive. Thanks for your understanding!

How do you explain ANY of the points I mentioned above???  NONE of those points (straight line through a building, a sudden “jump” of the aircraft from one location to another, the nose of the plane being in the exact position right before and right after the jump, no difference in battery power even though it traveled 94.6m apparently were addressed in your answer.  The homepoint not being next to the controller, etc, etc, etc.  

I took the aircraft out of the box like I’ve done every other time I’ve flown it, I put it on the ground in front of me... (IN THE PARK!!), I verified the connection to the controller, it said the RTH point was recorded, so I took off.  Then the Weak GPS warning came, so I just let the aircraft hover above ground a bit to see if the connection becomes stronger.  Then the warning disappeared, and just like every other time I’ve flown the aircraft, I thought the RTH point was recorded where the aircraft actually was at the time of take off.  
Just answer me this question... WHY would the RTH point be recorded on top of a 2 story building one block away and not in direct line of sight???  Why?  

Also, this sentence was not finished....
"The home point can be only recorded when the GPS signal which is different from take-off point”.  It would also be helpful if you could walk us through the usual steps the aircraft goes through to record the RTH point once it’s turned on.  Is it possible to turn it on remotely without physically pressing the power button?  Would you mind answering IF it is possible for a DJI Spark to take off from a slanted surface?   Why or why not?  Thanks for your understanding!

Can we agree that the RTH point was recorded before the Weak GPS warning came up?  Yes, or No?   
  
Can you please explain what the blue dot (in the park) signifies?

Can we agree that the controller is in the Park behind the buildings at the time of take-off?  Yes, or No?

Is there a clear line of sight between the controller and the aircraft in the beginning of the flight records video?   
Yes, or No?

Can we agree that in the flight records there is a clear and completely straight line (in yellow) from the recorded Home Point to the place in the park with the controller?  Yes, or No?  

It is also very clear that this line goes straight through and completely close to two buildings, as indicated on the screenshots above.   
Can we agree that you see this in the screenshot and that the recorded height goes from 4.4-4.7m?  Yes, or No?   
Would you mind defining the term relative altitude?  

Can you please describe in your own words, in detail, what happens between flight time 00:06-00:08 on the flight records video I provided the link to?   

Can you explain how the aircraft is able to travel a supposed distance from 1.2m to 96.4m in such a short time, and accelerate from 0.4km/h to 4.8km/h and then travel in a completely straight line through buildings, at a height of 4.6m within 1 or 2 seconds, and keep the nose pointed in the same direction during that time frame with only a 1% drop in battery power?  Also, what was happening on controller at the time?  Was there any input in any direction during those two seconds?  

Listen, I don’t think I’m being unreasonable to want to know the answers to these questions when it is so blatantly clear that there was some kind of malfunction with the aircraft.  I understand that you were not there that day in the park and that perhaps your company does not want you to admit on a public forum there was an error, but did you not see the video I provided?  How do YOU explain all these idiosyncrasies
of the aircraft in the video right after takeoff?  Would you not do the exact same thing were you in my situation?

Thanks for your continued support in this matter!  

bodhi09 Posted at 2018-4-26 00:31
Can we agree that the RTH point was recorded before the Weak GPS warning came up?  Yes, or No?   
  
Can you please explain what the blue dot (in the park) signifies?

bodhi, I totally understand your concern, here is the explanation of the points you mentioned.
1. The take-off point was recorded around the building per the flight record, the GPS signal may be affected by the building, the home point could shift a little bit, therefore, it is not weird that the home point was displayed in the 2 story building as you said.
2. Because of the poor signal, the position of the drone jumped to the next point that the signal recovered, this is normal. It is also normal that the flight records show a straight line since there was no GPS signal, we could not verify the actual flight path during this period.
3. It is user's responsibility to verify the home point on the map to ensure the flight performance, if it is recorded wrongly, the pilot should stop flying.
We could understand how frustrated you are, but the flight records were interrupted, we could not figure out the actual reason of the accident, your case will be escalated to the appropriate department to see if there is any better resolution could be provided. Thanks for your patience.

Hi bodhi, we should have someone already contacted you, please check the email and keep us updated if there is any further issue. Sorry again for the trouble caused.

DJI Mindy Posted at 2018-4-27 01:43
Hi bodhi, we should have someone already contacted you, please check the email and keep us updated if there is any further issue. Sorry again for the trouble caused.

Thank you very much for your help and your patience!  I really appreciate it!  

bodhi09 Posted at 2018-4-28 03:05
Thank you very much for your help and your patience!  I really appreciate it!

You are welcome, hope we can serve you next time with positive experience.