Please select Into the mobile phone version | Continue to access the computer ver.
Anyone want to Diagnose this Crash???
2017 35 2017-3-16
Uploading and Loding Picture ...(0/1)
o(^-^)o
Sestaceans
Second Officer
Flight distance : 3963865 ft
  • >>>
United States
Offline

[ Last edited by Sestaceans In 2017-4-3 14:27 Editor ]\n\n[ Last edited by Sestaceans In 2017-4-3 14:25 Editor ]\n\nQuestion for DJI Phantom Pro 4 Pilots.  So it's been 2.5 months and DJI still has not given me a reason why this drone crashed. If you own a Phantom 4 Pro and want to try and crack this case, please let me know your thoughts as DJI has not been able to (and thank you for those of you who don't own the P4Pro who have told me your thoughts, however for posterity lets try to authenticate this through the P4Pro/+ user base, but thank you for your desire to help).   Here is the flight  - https://www.youtube.com/watch?v=eflzR5FyYMk .  If you have any extra questions about the flight I will answer in order to help you diagnose the crash.   Here are the facts.  1. The Crash took place during Active Track with all 4 side sensors on (according to the menu)
2. The Crash took place on the second day it was released, flying with the DJIGo ap, not DJIGo4 (which was released later that day)
3. The Phantom Pro 4 was going less then 5mph when it crashed (opinion: should have been detected by senors).
4. As the drone was heading in to a post at 5mph,  the sticks were pushed to bail out (see youtube).  The drone did not respond.
5. I was told after the crash via customer support that I needed to calibrate the senors in the Assistant app and others also had sensor issues.
Those are the immeadiate facts.  I have not gotten a statement from DJI in 2.5 months about what caused this crash. I was charged for it.  I believe the senors failed, however at this point I don't care about the money I just want to know the truth.  As a Part 107 and 333 exempt company owner,  I am looking for answers.
Thank you for your analysis in advance (Let's all keep personal feelings out of this and stick to the diagnosis, and I'm sure a rational explantion for this crash will come to fruition.  Again please P4Pro owners only).  I will not be offering up my opinion, only answering diagnostic questions.




2017-3-16
Use props
Nigel_
Captain
Flight distance : 388642 ft
United Kingdom
Offline

Upload the log file to http://www.phantomhelp.com/LogViewer/upload/ (follow the instructions on the page) and post a link,  can't diagnose it properly without all the data.
2017-3-17
Use props
Sestaceans
Second Officer
Flight distance : 3963865 ft
  • >>>
United States
Offline

Nigel_ Posted at 2017-3-17 02:17
Upload the log file to http://www.phantomhelp.com/LogViewer/upload/ (follow the instructions on the page) and post a link,  can't diagnose it properly without all the data.

Hey Nigel,
Please refer to my prior post that you commented on as well to see the issue with DJI and the log retrieval for this crash.  Thanks for taking a look.
2017-3-17
Use props
Viridis
Captain
Flight distance : 2383556 ft
  • >>>
United Kingdom
Offline

Looks to me, that just prior to impact it was searching for track subject, was turning around immediately prior to impact and as such, struck the object side on.
Side sensors only active in tripod and beginner mode.
2017-3-17
Use props
Nigel_
Captain
Flight distance : 388642 ft
United Kingdom
Offline

Sestaceans Posted at 2017-3-17 02:31
Hey Nigel,
Please refer to my prior post that you commented on as well to see the issue with DJI and the log retrieval for this crash.  Thanks for taking a look.

So you don't have the phone/tablet logs either?  You must have them since we see them being replayed in your video...

One thing that puzzles me:
"the sticks were pushed to bail out"
Since you were in active track mode, the sticks do not move the aircraft as normal unless you exit active track mode first.  You don't appear to have exited active track until after the collision...
2017-3-17
Use props
Sestaceans
Second Officer
Flight distance : 3963865 ft
  • >>>
United States
Offline

Nigel_ Posted at 2017-3-17 03:53
So you don't have the phone/tablet logs either?  You must have them since we see them being replayed in your video...

One thing that puzzles me:

Hey Nigel, thanks for looking.  It puzzles me as well that there is absolutely no movement when trying to bail out of Active Track via stick movements.  As everyone knows,  even in Active Track movement will occur. According to DJI (not me) the logs are not able to be recovered by them (as stated in the posting you commented on earlier) for unexplained reasons.  They have confirmed the flights existence.  This is what we have to work with to find out the cause.  Again to read up on the matter with DJI and logs,  check out my post that you commented on earlier or I'm happy to post again.  Thank you.
2017-3-17
Use props
Sestaceans
Second Officer
Flight distance : 3963865 ft
  • >>>
United States
Offline

Viridis Posted at 2017-3-17 03:39
Looks to me, that just prior to impact it was searching for track subject, was turning around immediately prior to impact and as such, struck the object side on.
Side sensors only active in tripod and beginner mode.

Hey Virdis, thanks for looking at the video.   Active Track was on until it struck the object which knocked it out of  AT.   The back sensor should have worked especially at 5mph going straight back.  I fly in that location numerous times a week as it is a common area for shooting action sports (something active track usually works well for).  Never had an issue until this one.  
2017-3-17
Use props
Nigel_
Captain
Flight distance : 388642 ft
United Kingdom
Offline

If you are able to replay the flight in DJI Go, as we see in your video, then you are also able to upload the log to PhantomHelp for us to look at the details and make a decent analysis.

" It puzzles me as well that there is absolutely no movement when trying to bail out of Active Track via stick movements."

According to the manual, to bail out of active track using the sticks takes 3 seconds, you only operated the sticks for two seconds before the crash.
Alternatively you can exit Active Track immediately using the "Pause" button on the right side of the controller.  I take it that you didn't do this?
2017-3-17
Use props
Sestaceans
Second Officer
Flight distance : 3963865 ft
  • >>>
United States
Offline

Nigel_ Posted at 2017-3-17 11:51
If you are able to replay the flight in DJI Go, as we see in your video, then you are also able to upload the log to PhantomHelp for us to look at the details and make a decent analysis.

" It puzzles me as well that there is absolutely no movement when trying to bail out of Active Track via stick movements."

Hey Nigel,  I really do appreciate you looking at this.  If you don't mind, check out my previous post that explains the log issue (as I absolutely agree with you),  and then I'll definitely be happy to answer further questions of any kind that you have.  Thanks again.
As it pertains to moving the sticks.  You should see instant movement one way or another using them in Active Track.  I take it you have a Phantom Pro 4 correct?  So if you take it out and fly in Active Track, check out the movements you will get when replicating the exact same stick movements in Active Track..  Thanks
2017-3-17
Use props
Nigel_
Captain
Flight distance : 388642 ft
United Kingdom
Offline

Sestaceans Posted at 2017-3-17 12:15
Hey Nigel,  I really do appreciate you looking at this.  If you don't mind, check out my previous post that explains the log issue (as I absolutely agree with you),  and then I'll definitely be happy to answer further questions of any kind that you have.  Thanks again.
As it pertains to moving the sticks.  You should see instant movement one way or another using them in Active Track.  I take it you have a Phantom Pro 4 correct?  So if you take it out and fly in Active Track, check out the movements you will get when replicating the exact same stick movements in Active Track..  Thanks


When you move the sticks, assuming you are in the default mode 2:

First you throttle up to gain height, if you were heading towards a pole then that would make no difference to the outcome except result in a higher fall to the ground.

Second you push the right stick forward and right.  Since I can't see where the pole is I cant judge what effect the right stick had, but the full forward stick presumably flies it into the pole at full power?

The aircraft is facing you when you push full forward, however since you didn't press "Pause" to exit the active track and it is flying backwards in active track mode, right stick forwards flies further from the target (into your pole) while backwards brings it back closer to the target.

I think the real problem is that by the time you moved the right stick you only had 1 second to impact, you didn't give yourself enough space and time to be able to fly safely.  Especially if it was a new aircraft that you had little experience of.

Would be easier to check what happened if you posted the log file that we can see in your video...
2017-3-17
Use props
Geebax
Captain
Australia
Offline

Sestaceans Posted at 2017-3-17 12:15
Hey Nigel,  I really do appreciate you looking at this.  If you don't mind, check out my previous post that explains the log issue (as I absolutely agree with you),  and then I'll definitely be happy to answer further questions of any kind that you have.  Thanks again.
As it pertains to moving the sticks.  You should see instant movement one way or another using them in Active Track.  I take it you have a Phantom Pro 4 correct?  So if you take it out and fly in Active Track, check out the movements you will get when replicating the exact same stick movements in Active Track..  Thanks

'If you don't mind, check out my previous post that explains the log issue (as I absolutely agree with you),  and then I'll definitely be happy to answer further questions of any kind that you have.'

Why make us go searching for your previous post? If you want people to examine the data, then provide it here.
2017-3-17
Use props
#shotxclvck
Second Officer
Flight distance : 3423297 ft
United States
Offline

alot of warnings!
2017-3-17
Use props
#shotxclvck
Second Officer
Flight distance : 3423297 ft
United States
Offline

by bail out on the "sticks" do you mean CSC if so the drone would probably just drop would explain why the drone did not respond....it was already on the ground.
2017-3-17
Use props
Sestaceans
Second Officer
Flight distance : 3963865 ft
  • >>>
United States
Offline

Geebax Posted at 2017-3-17 14:01
'If you don't mind, check out my previous post that explains the log issue (as I absolutely agree with you),  and then I'll definitely be happy to answer further questions of any kind that you have.'

Why make us go searching for your previous post? If you want people to examine the data, then provide it here.

Thank you for you feedback Geebax.  The issue at hand is why the drone crashed, not why the Active Track did not manually react (even though clearly did not with stick movements), but why it ignored the backwards enabled Active Track and instead defaulted to the Phantom 4 standard firmware and had a sensor failure while going 5 mph.  My team has reconstructed the exact circumstances and have not been able to replicate this.   Nigel has already made his voice heard on another thread which is wonderful.  He has already made up his mind through his vote and his comments which I encouraged.   For posterity I will go ahead and include the rest of the information.   Again I would encourage those commenting who have not used the Phantom 4 Pro to reserve there analysis as the P4 and P4Pro are very different (disclaimer: I own both).  Regardless, thank you for your comments.   Hopefully together we can find out why DJI was not able to diagnose this drone.  If it was in fact pilot error, that is fine.  My only personal belief that I will share on this thread is DJI should at the very least provide an official reason for a warranty denial.  Oddly they gave me 30 percent discount because they were not able to do a proper diagnosis.  
Here is the information and the entire story for you Geebax and Nigel (even though you both were  kind enough to to already comment and presumably read this,  here it is).  Please note the issue with the DJIGo application as it will answer your questions.
Thank you.

Instead here is what DJI did provide ot me in two contrasting statements  NOTE (THE DRONE WAS FLOWN THROUGH THE DJIGO APP (not litchi-I don't even know what that is) BEFORE DJIGO4 APP WAS RELEASED).  ALL specifcs are provided below in an epic novela
EMAIL 1 March 4th -
"We were able to review the information which showed CSV files that is only used with the Litchi app.
Due to that the determination from analysis and all things considered were are able to offer 30% off the invoice which brings it down to 155.00.
A new invoice will be sent and once it is paid we will begin the process.
Thank you for your understanding. "

EMAIL 2 March8th (I asked for an explanation of why a warranty was denied)

"The files were CSV files. We were not able to pull them from the Go App per out data analysis. Due to this we are able to offer a discounted repair and brought the invoice down as low as possible.
Please pay the invoice so we can begin repair or it is going to be sent back soon.
Thanks for your understanding."


Question:  Why if you can't pull your onw logs would you automatically invoice the customer?



Log Notes:   
1. The drone starts to swerve off course on Active Track at 1:36,  notice that their are absolutely no stick movements.  
2. Operator tried to kick out of Active Track manually at aprox 1:40 (Stick commands as follow, up, straight ahead, right) with no luck,  the drone keeps going and crashes right at 1:42 (youtube Time Code) and spins to the ground.
3.  Op was told a few days later that new firmware and the DJI GO4 application was "now available".  The application had completely different firmware (as many of you already know).  





















FULL EXPLANATION

My Phantom 4 Pro was received as a ticket on Jan/06/2017 13:08:24  and dropped off 4 days later for shipping by yours truly.   The unit was received on Jan/27/2017 02:28:12.  In very short order it was diagnosed (Feb/01/2017 04:18:00).   I received my final bill with a list of all the things they replaced.   My total amount due was $222.   DJI remarked simply, "Non-warranty".  That was it.


Quote Date:
01-31-17
Remarks:
Non-warranty


I immediately emailed them since it was after hours, and asked DJI why this was not covered by my warranty and how did they see this event so different then me?.  After all it said in my description exactly what happened and when...., surely they read these right ?    During this time period I sent 4 regular emails to DJI support.  I sent 2 messages to DJIsupport on Twitter, and posted 2 tweets to them.  To add to that I had about 6 to 10 online chats depending on how long one has to chat to be defined a "chat",  and 4 phone calls  (one of which was about both the Phantom 4 Pro+ issues).  I had one question in all of these messages.   I asked them to provide me information as to why I was being charged and gave them the exact Date of Accident:2016/12/9 0:03:06 .  So I think it is fair to say that I told DJI exactly when the accident happened down to the minute.   I also stated exactly what happened   


After my drone crashed it still flew perfectly.  It was dinged up though,  and since I was told it was going to be covered by  a warranty due to the circumstances.  Once our Phantom4Pro+ arrived to our business ,  I sent the Phantom4Pro back to repair the damage that occurred on  2016/12/9 .   All my other flights are documented and no errors occurred  However,  it looks like DJI read ONLY the last logged flight that was done on my Phantom4Pro, which was from my back porch, and happens to be 20 feet high.  Even though there are absolutely NO error messages and no reason to suspect any type of error (as the drone landed perfectly, with no movement and with the proper shut down joystick technique.) ,  what did DJI say?  Well,  two tech phone staffers said in the most ambiguous terms, that something due to a battery issue occurred and that it was a pilot error.  Seriously.  This is all I was told.  No straight answers,  no emails back, no nothing, except that something about battery power while piloting--not even a straight answer.  However I took this with a grain of salt because this drone was never crashed after  2016/12/9 ,  and I knew the exact cause of why it crashed and the craft was actually filming at the time.  So DJI's story simply doesn't jive.   When on my last call today with DJI, someone once again said it had something to do with the battery but wasn't sure, but he would find out right then and their.  Unfortunately nobody was in the "shop".  So he would get back to me IMMEDIATELY.  Now that was a promise I got hopeful about.   Unfortunately  as of this writing I have not heard anything.


Here is the issue, and most importantly for me,  nobody has given me a  reason as to why this wasn't covered under the warranty.  Nobody has directed me to which log they gathered this information from since my drone was flown several more times before it was sent to DJI, it flew perfectly however It just didn't look Shiny and New because of the initial damage.  The gimbal was banged up for sure, however it worked great and caused absolutely NO issues.  Sooo,  The issue here is not the money, and it isn't about the drone (well, a little .  It's about the fact that nobody till this hour has answered me as to one simple question as to why I'm being billed and why won't DJI provide me the logs with the flight information they used.   My first request for ANY information was acknowledged to be received on February 3rd  by DJI (# 371316.).  So this would have meant, they have not responded to a simple email request in 22 days.   However, DJI found time to send me 3 invoices for $222.  So as they were able to ask for money,  they still have not told me what log they were looking at that made them conclude it was pilot error.  So why wouldn't DJI just simply give me an easily proven reason?  Because It would be impossible based on the log they did look at.


First off everyone knows the DJIGO4 application is used in conjunction with the Phantom 4 Pro.  So obviously you would look through the flight records and look for any abnormalities,  which I imagine they did. Like I said they took the last flight that was flown in my back yard and theorized that since it was landed at 20 feet high, it must have crashed or something crazy like that.   Unfortunately for DJI,  there aren't any error messages, no abnormalities, and absolutely nothing to support it was anything but a smooth landing....ON MY DECK (20ft up)  What they forgot though is that the date of my crash, which happened at 3:06pm, was the same day DJI announced the release of the the DJIGO4 application about 30 minutes later.   So they didn't take in to consideration that my logs for the crash were actually in the DJIGo application which they couldn't have seen in the DJIGO4 app where they looked.  One reason is because they weren't even synced in to the DJIGO app until I just synced them 2 days ago.  If DJI had looked in those logs they would have seen these flights and immediately seen that the Active Track was Activated and backed up directly in to a poll.  They also would have seen a million error messages and abnormalities that are the hallmarks of crashes.  However it's my opinion that DJI forgot (just like me) that when the phantom4pro was first released,  the DJIGO was it's launch app.  HOWEVER, they were given the exact date and time down to the second.  Lastly, right after the crash,  I went back inside, and guess what.  I uploaded the firmware,  and guess what.  My drone didn't even come with the correct operating system.  So faulty sensors, faulty active track,  and the wrong firmware.   I also am TOTALLY fine if they made this mistake.  I just don't understand why they can't get back to me with ANY information in 22 days and counting

2017-3-17
Use props
Sestaceans
Second Officer
Flight distance : 3963865 ft
  • >>>
United States
Offline

#shotxclvck Posted at 2017-3-17 14:48
by bail out on the "sticks" do you mean CSC if so the drone would probably just drop would explain why the drone did not respond....it was already on the ground.

Hey @shotxclvck.  As far as a manual bailout.  I simply mean tried to use the sticks to alter the position of the drone to avoid the post.  Also the warnings were after the sensor failure.


2017-3-17
Use props
Sestaceans
Second Officer
Flight distance : 3963865 ft
  • >>>
United States
Offline

Nigel_ Posted at 2017-3-17 12:48
When you move the sticks, assuming you are in the default mode 2:

First you throttle up to gain height, if you were heading towards a pole then that would make no difference to the outcome except result in a higher fall to the ground.

"Second you push the right stick forward and right.  Since I can't see where the pole is I cant judge what effect the right stick had, but the full forward stick presumably flies it into the pole at full power?"

Wouldn't this contradict your thought about the sticks not being able to react unless pushing pause (mine actually has a red "stop button" that is far more prevalent)?   Remember after the drone leaves Active Track the drone has already crashed.  So I'm trying to follow your logic I truly am,  however if you can provide a time code maybe I can try and understand what you are referring to.  Thank you
2017-3-17
Use props
Nigel_
Captain
Flight distance : 388642 ft
United Kingdom
Offline

Sestaceans Posted at 2017-3-17 15:46
"Second you push the right stick forward and right.  Since I can't see where the pole is I cant judge what effect the right stick had, but the full forward stick presumably flies it into the pole at full power?"

Wouldn't this contradict your thought about the sticks not being able to react unless pushing pause (mine actually has a red "stop button" that is far more prevalent)?   Remember after the drone leaves Active Track the drone has already crashed.  So I'm trying to follow your logic I truly am,  however if you can provide a time code maybe I can try and understand what you are referring to.  Thank you

There is no way you pressed the red stop button and then moved your hands back to the sticks in the time available, I didn't need to ask about that.  The only way you could have exited Active Track was using the pause button, I asked about it because without the log I couldn't be sure if you had used it or not.

Since you didn't exit Active Track, I think pushing the right stick forward would have flown the aircraft into the post instead of away from it which is presumably what you expected it to do when you had to react with only 2 seconds available and not enough time to think about what you were doing...
2017-3-17
Use props
Nigel_
Captain
Flight distance : 388642 ft
United Kingdom
Offline

I guess the reason that the rear obstacle avoidance didn't work was that by the time the post moved behind the aircraft, the post was already too close to the aircraft for the 3D vision to be able to measure the distance.  Like your eyes, the stereo vision does have a minimum distance at which it works for depth perception.
2017-3-17
Use props
Sestaceans
Second Officer
Flight distance : 3963865 ft
  • >>>
United States
Offline

Nigel_ Posted at 2017-3-17 16:21
There is no way you pressed the red stop button and then moved your hands back to the sticks in the time available, I didn't need to ask about that.  The only way you could have exited Active Track was using the pause button, I asked about it because without the log I couldn't be sure if you had used it or not.

Since you didn't exit Active Track, I think pushing the right stick forward would have flown the aircraft into the post instead of away from it which is presumably what you expected it to do when you had to react with only 2 seconds available and not enough time to think about what you were doing...

Respectfully,  I am still confused by your analysis.  Are you saying I could not use my joysticks regardless or are you saying that I did and therefore went forward (or backward) and in to a pole after (or before) the crash?  if I'm reading this correctly you are saying that I did back in to the pole, but also went forward in to it through my joystick movements?   Perhaps you could clarify by using time-codes since you are interested in the analysis (which is greatly appreciated).   Just for clarity,  are you using the P4Pro or the P4Pro+  and did what app were you using when you first flew your bird?

HOWEVER,  I think we are putting the cart before the horse as we must find the reason the drone swerved off course in the first place when backwards Active Track was enabled,  Since I know you already read the full description I wrote,  to summarize for others, the  DJIGO4 app had not yet been released,  and it looks as if the Phantom 4 firmware was installed NOT the Pro version.  So I'm wondering if the P4Pro arrived with incorrect firmware and therefore had an impact on the P4Pro sensors (the drone acted like a P4 instead of a P4Pro).  I have a full theory but am interested in others analysis first and foremost.  

Keep them coming.  It's really interesting because DJI couldn't diagnose it.
2017-3-17
Use props
Sestaceans
Second Officer
Flight distance : 3963865 ft
  • >>>
United States
Offline

Nigel_ Posted at 2017-3-17 16:24
I guess the reason that the rear obstacle avoidance didn't work was that by the time the post moved behind the aircraft, the post was already too close to the aircraft for the 3D vision to be able to measure the distance.  Like your eyes, the stereo vision does have a minimum distance at which it works for depth perception.

I very much doubt this theory.  I've seen drones going 30-40mph stop on a dime with the sensors.  Going 2mph,  forget about it,  the drone would have easily had time to stop if the sensors were in fact working.  In my opinion they could not have been (my opinion of course).

New pic.jpg
2017-3-17
Use props
Nigel_
Captain
Flight distance : 388642 ft
United Kingdom
Offline

Sestaceans Posted at 2017-3-17 18:11
Respectfully,  I am still confused by your analysis.  Are you saying I could not use my joysticks regardless or are you saying that I did and therefore went forward (or backward) and in to a pole after (or before) the crash?  if I'm reading this correctly you are saying that I did back in to the pole, but also went forward in to it through my joystick movements?   Perhaps you could clarify by using time-codes since you are interested in the analysis (which is greatly appreciated).   Just for clarity,  are you using the P4Pro or the P4Pro+  and did what app were you using when you first flew your bird?

HOWEVER,  I think we are putting the cart before the horse as we must find the reason the drone swerved off course in the first place when backwards Active Track was enabled,  Since I know you already read the full description I wrote,  to summarize for others, the  DJIGO4 app had not yet been released,  and it looks as if the Phantom 4 firmware was installed NOT the Pro version.  So I'm wondering if the P4Pro arrived with incorrect firmware and therefore had an impact on the P4Pro sensors (the drone acted like a P4 instead of a P4Pro).  I have a full theory but am interested in others analysis first and foremost.  

I'm saying that in Active Track mode, pushing the right stick forward (which is what you did as shown on your video just before the crash) moves the aircraft further away from you and towards the location of the crash.

The aircraft did what you told it to do - fly into the pole.
2017-3-18
Use props
Nigel_
Captain
Flight distance : 388642 ft
United Kingdom
Offline

Sestaceans Posted at 2017-3-17 18:11
Respectfully,  I am still confused by your analysis.  Are you saying I could not use my joysticks regardless or are you saying that I did and therefore went forward (or backward) and in to a pole after (or before) the crash?  if I'm reading this correctly you are saying that I did back in to the pole, but also went forward in to it through my joystick movements?   Perhaps you could clarify by using time-codes since you are interested in the analysis (which is greatly appreciated).   Just for clarity,  are you using the P4Pro or the P4Pro+  and did what app were you using when you first flew your bird?

HOWEVER,  I think we are putting the cart before the horse as we must find the reason the drone swerved off course in the first place when backwards Active Track was enabled,  Since I know you already read the full description I wrote,  to summarize for others, the  DJIGO4 app had not yet been released,  and it looks as if the Phantom 4 firmware was installed NOT the Pro version.  So I'm wondering if the P4Pro arrived with incorrect firmware and therefore had an impact on the P4Pro sensors (the drone acted like a P4 instead of a P4Pro).  I have a full theory but am interested in others analysis first and foremost.  

The reason that it swerved "off course" is that you had backwards flying enabled so it was backing away from the target.

Backwards flying does not mean that it looks into the future to see what direction the target is going to travel and then puts itself in front of where the target is going to be, instead it just moves backwards in whatever direction it is already pointing, that means that if the target started moving towards its right (targets left) then it would have reversed around to its left (targets right) and towards the accident scene.

It did what I would expect for backwards flying in active track mode.
2017-3-18
Use props
Nigel_
Captain
Flight distance : 388642 ft
United Kingdom
Offline

Sestaceans Posted at 2017-3-17 18:13
I very much doubt this theory.  I've seen drones going 30-40mph stop on a dime with the sensors.  Going 2mph,  forget about it,  the drone would have easily had time to stop if the sensors were in fact working.  In my opinion they could not have been (my opinion of course).

If it sees an obstacle 10 meters away then certainly it will stop pretty fast.  The problem here was that it was turning and the pole was a lot closer than anything behind the pole, it couldn't see the pole until the pole came into its field of view, which according to the manual is 50 degrees for the rear sensors.  A 50 degree FOV starting at the rear obstacle sensors means that it can't see things that are very close unless they are directly behind the aircraft, since the pole came into view from the side and was already close it wouldn't have see it until very late.  Further more, the FOV for the 3D perception doesn't start immediately behind the sensors, it starts where the FOV of the two stereo sensors starts to overlap which is some distance behind the aircraft.   I don't know what the closest point it can detect objects is but I wouldn't expect it to be able to judge 3D distance to anything within the first meter, I suspect that by the time the pole was within the FOV it may have been less than a meter away.

Also when the obstacle detection puts the brakes on, you can't expect it to stop instantly, it does take a little time and from watching the video it doesn't look like it would have had more than about half a second from seeing the pole even if the pole was still sufficiently far away for it to be able to judge distance.
2017-3-18
Use props
Sestaceans
Second Officer
Flight distance : 3963865 ft
  • >>>
United States
Offline

Nigel_ Posted at 2017-3-18 01:55
I'm saying that in Active Track mode, pushing the right stick forward (which is what you did as shown on your video just before the crash) moves the aircraft further away from you and towards the location of the crash.

The aircraft did what you told it to do - fly into the pole.

So you are saying that the drone during Active Track did have a response to the sticks?  I'm confused because earlier you said it didn't. .  Could you provide a TC and the drone you are using to analyize this as I'm sure you have a PhantomPro4 if you are diagnosing this.  As we have found in several attempts to replicate this instance,  Active Track continues on the path back, even to the subject.  As you can see this is a road and the drone would be goping backwards parallel to the road.  Are you implying that the object went to the right or left of the drone?  Please use time code.  I appreciate you taking the time to look at this, however if you don't provide time code how can you substantiate anything?
I also have a video I will post WITH the log of the exact same flight with all same circumstances and settings (actually the same drone I got back, still same settings)/.  Perhaps that will give us a clue to the reality of your claim and the drones actual movements.   Right now I have to disagree completely with you until I see either a Timecode and what drone you are basing this off of.  However I appreciate you taking the time to look at the flight log (unlike DJI).
2017-3-18
Use props
Nigel_
Captain
Flight distance : 388642 ft
United Kingdom
Offline

Sestaceans Posted at 2017-3-18 13:14
So you are saying that the drone during Active Track did have a response to the sticks?  I'm confused because earlier you said it didn't. .  Could you provide a TC and the drone you are using to analyize this as I'm sure you have a PhantomPro4 if you are diagnosing this.  As we have found in several attempts to replicate this instance,  Active Track continues on the path back, even to the subject.  As you can see this is a road and the drone would be goping backwards parallel to the road.  Are you implying that the object went to the right or left of the drone?  Please use time code.  I appreciate you taking the time to look at this, however if you don't provide time code how can you substantiate anything?
I also have a video I will post WITH the log of the exact same flight with all same circumstances and settings (actually the same drone I got back, still same settings)/.  Perhaps that will give us a clue to the reality of your claim and the drones actual movements.   Right now I have to disagree completely with you until I see either a Timecode and what drone you are basing this off of.  However I appreciate you taking the time to look at the flight log (unlike DJI).

If you post the video from the aircraft's main camera then we should be able to see when the subject moves left and then we can give you an exact timecode, if indeed the subject did move left.  Also if you post the log file that we can see being replayed in the above video then we will see from the log data the exact timecode when it starts to turn and also when you operate the sticks.

I think it is irrelevant to apportioning blame however.  The real problem is that you only operated the sticks to avoid the collision 2 seconds before the collision and didn't move the right stick until 1 second before the collision.  I believe you then pushed the stick in the wrong direction thus causing the crash.

There may have been a fault in the aircraft although I don't see any evidence of one, if there was one then DJI would be responsible for fixing that fault under warrantee.  However they are not responsible for the crash or the damage caused by the crash, that is pilot error and should be dealt with by insurance.
2017-3-18
Use props
Sestaceans
Second Officer
Flight distance : 3963865 ft
  • >>>
United States
Offline

Nigel_ Posted at 2017-3-18 15:25
If you post the video from the aircraft's main camera then we should be able to see when the subject moves left and then we can give you an exact timecode, if indeed the subject did move left.  Also if you post the log file that we can see being replayed in the above video then we will see from the log data the exact timecode when it starts to turn and also when you operate the sticks.

I think it is irrelevant to apportioning blame however.  The real problem is that you only operated the sticks to avoid the collision 2 seconds before the collision and didn't move the right stick until 1 second before the collision.  I believe you then pushed the stick in the wrong direction thus causing the crash.

Understood.  What is the timecode you see though in which you are refering to on the video I posted?
2017-3-18
Use props
Nigel_
Captain
Flight distance : 388642 ft
United Kingdom
Offline

Sestaceans Posted at 2017-3-18 16:40
It's a brand new machine with Active Track function that enables you to go backwards and claims sensors work.


Did you read the last sentence under "Enable backward flying" in your photo above?




2017-3-18
Use props
Sestaceans
Second Officer
Flight distance : 3963865 ft
  • >>>
United States
Offline

Nigel_ Posted at 2017-3-18 23:59
Did you read the last sentence under "Enable backward flying" in your photo above?

https://forum44.djicdn.com/data/attachment/forum/201703/17/220017qaedyyea1elu9uyd.jpg[/img]

Yes but I didn't see a timecode.  It seems you think the crash was at a different time then  me.
Since you have a P4PRo (otherwise I'm sure you wouldn't be in a the P4Pro forum ) you should simply take your P4Pro drone outside and do the experiment for yourself .  You will find your answer right there.
Simply put your drone in these settings, have your finger on the stop button, go backward at 2MPH using the same diagnostics and see if it is possible to push it back in to a post or tree (or in your case get close to it ).

2017-3-19
Use props
Nigel_
Captain
Flight distance : 388642 ft
United Kingdom
Offline

Sestaceans Posted at 2017-3-19 12:55
Yes but I didn't see a timecode.  It seems you think the crash was at a different time then  me.
Since you have a P4PRo (otherwise I'm sure you wouldn't be in a the P4Pro forum ) you should simply take your P4Pro drone outside and do the experiment for yourself .  You will find your answer right there.
Simply put your drone in these settings, have your finger on the stop button, go backward at 2MPH using the same diagnostics and see if it is possible to push it back in to a post or tree (or in your case get close to it ).

I have no intention of replicating your crash
2017-3-19
Use props
Sestaceans
Second Officer
Flight distance : 3963865 ft
  • >>>
United States
Offline

Nigel_ Posted at 2017-3-19 14:04
I have no intention of replicating your crash

Because you couldn't
2017-3-20
Use props
Nigel_
Captain
Flight distance : 388642 ft
United Kingdom
Offline


Until you provide the log file and video from the aircraft's main camera, nobody can replicate the crash accurately.  Why do you not provide these?  What are you trying to achieve by only providing selected data?

Even if I did have the necessary information, I still do not want to intentionally crash my aircraft
2017-3-20
Use props
Glimmerman
Banned

United States
Offline

Tip: the author has been banned or deleted automatically shield
2017-4-2
Use props
Brian Fist
lvl.4
Flight distance : 19928 ft
United States
Offline

It is obvious that Sestaceans only wants to represent *HIS* point of view and wants someone (anyone???) to agree with him on something.

If he provided ALL the logs and data, it would show that he was to blame, exactly what DJI found.  DJI does NOT need to provide you with any more information than what they already have, that YOU were at fault.  To come on here and bemoan the issue is moot.  You will not "shame them" into anything.  You won't get anyone else to take your side either with limited, and faulty, information.  Your case is dead, the point is moot, move on, it is all over, yet again.

Now, I won't state you are a bad pilot, good pilot or anything else, since I don't know.  What is obvious is that you are attempting to manipulate things to build your "case".  You don't have one.

If you supply all the requested information, then people can do true forensic analysis of the situation and then offer a valid viewpoint.  But be prepared to be told you are at fault.  And if that happens then it just shows you had no reason to start this thread in the first place.  If not, then it might just give others a bit of insight on how to handle their own cases, but as it stands, for the dollars you mentioned, nothing more will happen for you.  And for you to press it, even with others stating whether or not it was your fault or not, to try to further things with DJI is just ignorant.  The best outcome would be for the information gleaned here to be useful for someone else.
2017-4-3
Use props
Sestaceans
Second Officer
Flight distance : 3963865 ft
  • >>>
United States
Offline

Brian Fist Posted at 2017-4-3 10:58
It is obvious that Sestaceans only wants to represent *HIS* point of view and wants someone (anyone???) to agree with him on something.

If he provided ALL the logs and data, it would show that he was to blame, exactly what DJI found.  DJI does NOT need to provide you with any more information than what they already have, that YOU were at fault.  To come on here and bemoan the issue is moot.  You will not "shame them" into anything.  You won't get anyone else to take your side either with limited, and faulty, information.  Your case is dead, the point is moot, move on, it is all over, yet again.

DJI and I came to a resolution on this matter that was mutally beneficial.  They ended up listening to my concerns. Thank you for your interest though
2017-4-3
Use props
Brian Fist
lvl.4
Flight distance : 19928 ft
United States
Offline

Sestaceans Posted at 2017-4-3 11:59
DJI and I came to a resolution on this matter that was mutally beneficial.  They ended up listening to my concerns. Thank you for your interest though

That is a joke, "mutually beneficial", as in if you had ANY chance EVER of doing them "harm", or that they "owed" you in any way...

You come off like such a joke...

There is nothing you could ever do except throw the temper-tantrum you usually do.  It is just funny that your ilk believe they are "due" and everyone else should pander to your stupidity.

It is obvious that YOU broke your P4P... Then you tried a temper tantrum here to try to get them to pander to your stupidity.

Do you ever realize that maybe, just MAYBE they gave you a discount to make you go away?

But, then again, morons like you think you are constantly winning at something just because people are tired of dealing with you.

Hard telling you don't have many friends, or associates that care to deal with you, or in any way shape or form are held with any respect by anyone.  Thus your need to attempt to "force" any and all situations.  And somehow you feel you are "right"... What a joke!
2017-4-3
Use props
Glimmerman
Banned

United States
Offline

Tip: the author has been banned or deleted automatically shield
2017-4-3
Use props
Advanced
You need to log in before you can reply Login | Register now

Credit Rules