Phantom 3 malfunction and crash (flight record video)

This thread got me thinking about my days as an operations manager at a top US defense contractor and what I learned about Statistical Process Control in regards to manufacturing. The goal for major manufactures who put a strong emphasis on quality or lean manufacturing is to develop processes that perform at six sigma or better. What that means is statistically you can produce parts/features within a normal distribution that are +/- 6 standard deviations (sigma) within spec. What that boils down to (factoring in sigma shift) is that your 6sigma process will yield a success rate of 99.99966% or 3.4 defects per million opportunities (DPMO). Now that seems pretty good, but hypothetically, if you figure there are 12 critical components on a Phantom 3 (4 motors, 4 props, 4 electrical components) and if DJI produces say 500,000 Phantom 3 drones this year with a respectable 6sigma process, it could still create a potential for 20 (3.4 / 2 x 12 = ~20) brand new Phantom 3s to have a failure and fall out of the sky potentially without warning. Although that is a relatively small number it is still scary and could lead to someone getting seriously hurt.

I am certainly not implying that DJI has any sort of manufacturing issue and it’s not my intent to put them down in anyway. For now, pilot error will most likely be the cause of most drone crashes. But I wanted to illustrate how important the quality of the manufacturing process is when you consider producing products at large quantities. Even very slight margins of error can have a significant consequence, especially for systems like drones that are flying over people and property. It is statistically impossible to create a manufacturing process that has absolutely zero defects and drones will have defects that will cause them to fall out of the sky. As the number of drones goes up so will the number of failures. The question is, what does the drone community due to make the situation safer in the long-run?

I was inspired to write an article on my Blog about the topic, if you are interested in learning more, please check it out, http://mydailydrone.com/will-my-new-phantom-3-randomly-fall-out-of-the-sky.
 
This thread got me thinking about my days as an operations manager at a top US defense contractor and what I learned about Statistical Process Control in regards to manufacturing. The goal for major manufactures who put a strong emphasis on quality or lean manufacturing is to develop processes that perform at six sigma or better. What that means is statistically you can produce parts/features within a normal distribution that are +/- 6 standard deviations (sigma) within spec. What that boils down to (factoring in sigma shift) is that your 6sigma process will yield a success rate of 99.99966% or 3.4 defects per million opportunities (DPMO). Now that seems pretty good, but hypothetically, if you figure there are 12 critical components on a Phantom 3 (4 motors, 4 props, 4 electrical components) and if DJI produces say 500,000 Phantom 3 drones this year with a respectable 6sigma process, it could still create a potential for 20 (3.4 / 2 x 12 = ~20) brand new Phantom 3s to have a failure and fall out of the sky potentially without warning. Although that is a relatively small number it is still scary and could lead to someone getting seriously hurt.

I am certainly not implying that DJI has any sort of manufacturing issue and it’s not my intent to put them down in anyway. For now, pilot error will most likely be the cause of most drone crashes. But I wanted to illustrate how important the quality of the manufacturing process is when you consider producing products at large quantities. Even very slight margins of error can have a significant consequence, especially for systems like drones that are flying over people and property. It is statistically impossible to create a manufacturing process that has absolutely zero defects and drones will have defects that will cause them to fall out of the sky. As the number of drones goes up so will the number of failures. The question is, what does the drone community due to make the situation safer in the long-run?

I was inspired to write an article on my Blog about the topic, if you are interested in learning more, please check it out, http://mydailydrone.com/will-my-new-phantom-3-randomly-fall-out-of-the-sky.
OMG finally someone that understands statistics and the manufacturing process! You should go find the thread where some dude wanted to start a class action law suit because his phantom was defective. I'm actually smiling right now. Thank you for posting! And welcome aboard.
 
This thread got me thinking about my days as an operations manager at a top US defense contractor and what I learned about Statistical Process Control in regards to manufacturing. The goal for major manufactures who put a strong emphasis on quality or lean manufacturing is to develop processes that perform at six sigma or better. What that means is statistically you can produce parts/features within a normal distribution that are +/- 6 standard deviations (sigma) within spec. What that boils down to (factoring in sigma shift) is that your 6sigma process will yield a success rate of 99.99966% or 3.4 defects per million opportunities (DPMO). Now that seems pretty good, but hypothetically, if you figure there are 12 critical components on a Phantom 3 (4 motors, 4 props, 4 electrical components) and if DJI produces say 500,000 Phantom 3 drones this year with a respectable 6sigma process, it could still create a potential for 20 (3.4 / 2 x 12 = ~20) brand new Phantom 3s to have a failure and fall out of the sky potentially without warning. Although that is a relatively small number it is still scary and could lead to someone getting seriously hurt.

I am certainly not implying that DJI has any sort of manufacturing issue and it’s not my intent to put them down in anyway. For now, pilot error will most likely be the cause of most drone crashes. But I wanted to illustrate how important the quality of the manufacturing process is when you consider producing products at large quantities. Even very slight margins of error can have a significant consequence, especially for systems like drones that are flying over people and property. It is statistically impossible to create a manufacturing process that has absolutely zero defects and drones will have defects that will cause them to fall out of the sky. As the number of drones goes up so will the number of failures. The question is, what does the drone community due to make the situation safer in the long-run?

I was inspired to write an article on my Blog about the topic, if you are interested in learning more, please check it out, http://mydailydrone.com/will-my-new-phantom-3-randomly-fall-out-of-the-sky.
I love this guy.
 
  • Like
Reactions: pjw2565
This thread got me thinking about my days as an operations manager at a top US defense contractor and what I learned about Statistical Process Control in regards to manufacturing. The goal for major manufactures who put a strong emphasis on quality or lean manufacturing is to develop processes that perform at six sigma or better. What that means is statistically you can produce parts/features within a normal distribution that are +/- 6 standard deviations (sigma) within spec. What that boils down to (factoring in sigma shift) is that your 6sigma process will yield a success rate of 99.99966% or 3.4 defects per million opportunities (DPMO). Now that seems pretty good, but hypothetically, if you figure there are 12 critical components on a Phantom 3 (4 motors, 4 props, 4 electrical components) and if DJI produces say 500,000 Phantom 3 drones this year with a respectable 6sigma process, it could still create a potential for 20 (3.4 / 2 x 12 = ~20) brand new Phantom 3s to have a failure and fall out of the sky potentially without warning. Although that is a relatively small number it is still scary and could lead to someone getting seriously hurt.

I am certainly not implying that DJI has any sort of manufacturing issue and it’s not my intent to put them down in anyway. For now, pilot error will most likely be the cause of most drone crashes. But I wanted to illustrate how important the quality of the manufacturing process is when you consider producing products at large quantities. Even very slight margins of error can have a significant consequence, especially for systems like drones that are flying over people and property. It is statistically impossible to create a manufacturing process that has absolutely zero defects and drones will have defects that will cause them to fall out of the sky. As the number of drones goes up so will the number of failures. The question is, what does the drone community due to make the situation safer in the long-run?

I was inspired to write an article on my Blog about the topic, if you are interested in learning more, please check it out, http://mydailydrone.com/will-my-new-phantom-3-randomly-fall-out-of-the-sky.
Loved the blog man. I'll keep an eye on it. Welcome to the board!
 
  • Like
Reactions: pjw2565
Loved the blog man. I'll keep an eye on it. Welcome to the board!

Thanks Adam! That means a lot. Hopefully everything has worked out with your situation and you are taking care of. I am anxiously awaiting for my Phantom 3 to arrive and hopefully I don't have similar issues. But at least I know to look out for the Jello issue as an indicator of a problem. Thank you again for sharing what happened with your drone. We all learned a great deal from the discussion.
 
Thanks Adam! That means a lot. Hopefully everything has worked out with your situation and you are taking care of. I am anxiously awaiting for my Phantom 3 to arrive and hopefully I don't have similar issues. But at least I know to look out for the Jello issue as an indicator of a problem. Thank you again for sharing what happened with your drone. We all learned a great deal from the discussion.
From what I understand, it will be quite awhile until I know the outcome of their internal investigation but I've done everything I could to document it and I do have faith they'll issue me a new unit...

...but yeah, guinea pig lesson from me is absolutely pay attention to "jello". I had someone PM me a video of theirs and it was awful. There should be NONE. The gimbal is remarkable and their stability technology is second to none. If you're experiencing repeated jello under during pretty standard movements and heights, as much as it sucks to be without your bird for a few weeks, contact them and send it in.
 
  • Like
Reactions: ctp and pjw2565
If someone can explain what the hell just happened, I'd be awfully appreciative.

Trying to troubleshoot the "jello" issue, I had my P3P in the air floating around 90 feet while changing settings. I go to bring it a little closer and it goes absolutely crazy. With nothing around it, it spins, rocks from side to side and as you can see from the video below, crashes. It's completely totaled (gimbal broken, shell busted, ripped apart). Thankfully no one was hurt! My neighbors were scared, but really just felt bad for me. They saw it just go crazy.

Thankfully there's flight records so I can prove it went completely crazy with ZERO input from me... however does that matter? Is this considered user error? I'm just out?

Stunned.


After watching your video, I suspect it took an RF hit. Seen it and have had it happen many times - see my test -
 
The odd thing is the spinning. I have lost a couple of quads and they all tipped over sliding into the ground, even with prop loss. The spin seems software driven or errant radio signal. Also, once a spin starts serious prop wash begins and the quad fallss through very disturbed air with little lift.
 
  • Like
Reactions: Adam Kontras
The odd thing is the spinning. I have lost a couple of quads and they all tipped over sliding into the ground, even with prop loss. The spin seems software driven or errant radio signal. Also, once a spin starts serious prop wash begins and the quad fallss through very disturbed air with little lift.

Yeah it certainly wasn't RF related. I mean, I guess we don't know that but if you watch the way it spins (and since I actually watched it rock back and forth), it was fighting itself in a sense, violently, all the way down for a good amount of time. For whatever reason one of the arms wasn't working and it was trying to make up for it, but couldn't. If it were RF there wouldn't have been the crazy rocking back and forth and spinning, I assume it would've just got its signals crossed and moved one way or the other.

Time will tell as DJI now has it.
 
Yeah it certainly wasn't RF related. I mean, I guess we don't know that but if you watch the way it spins (and since I actually watched it rock back and forth), it was fighting itself in a sense, violently, all the way down for a good amount of time. For whatever reason one of the arms wasn't working and it was trying to make up for it, but couldn't. If it were RF there wouldn't have been the crazy rocking back and forth and spinning, I assume it would've just got its signals crossed and moved one way or the other.

Time will tell as DJI now has it.

Haven't posted much here. I mainly post on RCG. This thread caught my eye because I am an owner of a P3 Pro and I also purchased from Advexure. I have had a pleasant experience thus far with Advexure. I spoke to Travis the evening before placing my pre order for the P3. They ensured me that they have good relations with DJI and do everything they can solve warranty issues directly with the customer. With that said....are you happy with how they handled this situation? Did they attempt to help you directly before referring you to DJI? Did you purchase your new P3 with Advexure?
 
Haven't posted much here. I mainly post on RCG. This thread caught my eye because I am an owner of a P3 Pro and I also purchased from Advexure. I have had a pleasant experience thus far with Advexure. I spoke to Travis the evening before placing my pre order for the P3. They ensured me that they have good relations with DJI and do everything they can solve warranty issues directly with the customer. With that said....are you happy with how they handled this situation? Did they attempt to help you directly before referring you to DJI? Did you purchase your new P3 with Advexure?
I've literally never been happier with a company in my life. I have been stunned at how well they've handled this. The president wrote me on a Sunday and called me the morning after the holidays to walk me through the process. I will certainly write more about them as we continue through the process.
 
I've literally never been happier with a company in my life. I have been stunned at how well they've handled this. The president wrote me on a Sunday and called me the morning after the holidays to walk me through the process. I will certainly write more about them as we continue through the process.

Great good to know. I just got my extra battery from them yesterday. I am impressed with their service for sure. I'm waiting for them to have preorders on the GPC backpack, that'll be my next purchase from Advexure.
 
  • Like
Reactions: Adam Kontras
I was changing camera settings. 1080, 30 fps, 24 fps. When I said without "user input" I meant anything that had to do with flying.

Yeah, this question had me thinking. What are the settings you are not supposed to change while flying. Whether you are changing camera settings (we all do this from time to time) to flying mode settings, (also done quite often by many of us) the bird should stay on the air.
My drone never had an issue like this until yesterday. Battery went from 40% to 7% in just one second. I had a hard but successful extra fast landing. These things can go crazy at any time without much of an explanation.
 
Did we ever get the FLY???.DAT file from this flight? It would be interesting to put it through http://www.uavforecast.com/djilogs and see what log messages come out.
 
Did we ever get the FLY???.DAT file from this flight? It would be interesting to put it through http://www.uavforecast.com/djilogs and see what log messages come out.

I just ran it through that site and haven't the slightest idea what the spreadsheets show, but I assume the video of the animated flight path pretty much covers it... have at it if you can find anything good!
 

Attachments

  • DJIFlightRecord_2015-05-23_[09-58-55].txt
    363.5 KB · Views: 315
HOLY *@#$*&% that kml file is BRILLIANT!!!!! It proves I didn't hit anything! I wish I could overlay the animation on it so you could see the speeds. But that 3D tracking line along with the video:


Is perfect. Thank you so much for sharing that site I've been looking for it forever.
 
You're welcome. If you could get the FLY???.DAT file from off the SD card inside the Phantom, that would be even better :)
 
Nothing obviously wrong looking at the .txt file - just looks like motor/ESC/prop failure. The Phantom dropped out of GPS mode because the speed error was too large, but that happened a couple of seconds after things started to go wrong.

Only the FLY???.DAT file will reveal more at this stage. Instructions here http://www.phantompilots.com/threads/how-can-i-read-fly-dat-files.42994/
 

Recent Posts

Members online

No members online now.

Forum statistics

Threads
143,099
Messages
1,467,637
Members
104,986
Latest member
dlr11164