Crashed P4 - data/pictures inside - what to do next?

[John1960]

DJI is still reviewing the data. Clearly if they find malfunction, it will be a 100% coupon code. If they find pilot error, what do they do typically?

Nothing why should they give you anything for your incompitance !!!!!!!!!!!!!!

 

[Neon Euc]

Nothing why should they give you anything for your incompitance !!!!!!!!!!!!!!

What? Are you kidding? OK then... Next time you have a flight and the home point is located miles away and you lose your drone as it's happened so many times with other people due to firmware problems... Don't come crying to us

 

[sar104]

What? Are you kidding? OK then... Next time you have a flight and the home point is located miles away and you lose your drone as it's happened so many times with other people due to firmware problems... Don't come crying to us

I think you missed the question that he was replying to which was, to paraphrase - "if DJI determine that it was pilot error, what do they do?"

 

[Neon Euc]

I think you missed the question that he was replying to which was, to paraphrase - "if DJI determine that it was pilot error, what do they do?"

If that's the case then I do apologize and mean no harm [emoji4] thank you for clearing that up and I do apologize once again [emoji4]

 

[Meta4]

What? Are you kidding? OK then... Next time you have a flight and the home point is located miles away and you lose your drone as it's happened so many times with other people due to firmware problems... Don't come crying to us

I haven't seen any cases such as what you describe as happening so many times?.
Can you point me to one or two?

 

[Neon Euc]

I haven't seen any cases such as what you describe as happening so many times?.
Can you point me to one or two?

I don't have the direct links but I remember a few firmware updates a go where a number of people flew only to lose connection and then see the home point being set miles away from where they actually we're.

 

[Meta4]

I don't have the direct links but I remember a few firmware updates a go where a number of people flew only to lose connection and then see the home point being set miles away from where they actually we're.

There was a particular incident in January that caused that for owners of the Standard.
DJI jumped on it quickly and brought out a mandatory update soon after to prevent that and it's not been seen since.
Having your homepoint jump is not something that users need to be concerned about and has not been a worry at all except for a few Standard owners back in January.

 

[David_Cambridge]

I too dropped a phantom into a pond. I swore that it was a problem with the drone until I got the flight log into a spreadsheet. Exsamining it carefully at the end showed clearly that the left hand stick was in the descend range until the bird hit the water. Up until that point I continued to believe that the error was DJI. It is hard to see how any DJI issue could lead to a log showing that the control immediately before the crash was down

 

[Deleted member 99081]

DJI is still reviewing the data. Clearly if they find malfunction, it will be a 100% coupon code. If they find pilot error, what do they do typically?

I've just been offered 30% discount (drone only, no controller or charger) for a crashed and lost P4A.

DJI say..."With the record ended (sic) without any sign of abnormality, we could not verify what happened afterwards."

I have asked what this means in monetary terms, as I had ordered and received a replacement P4A from Amazon for €1211. They have not replied yet.

I'm assuming it's 30% off the current DJI price of the full P4A kit (€1400, with controller and charger) which would be about €964. I am thinking of keeping the Amazon one, as I may not save more than €150 after return costs and the delay in getting flying again (I have a commercial job tomorrow) doesn't really warrant it (IMHO).

 

[Dingoz]

I think DJI will never admit liability unless a log entry that absolutely shows a fault. If a drone is at fault but does not log it then we are left with no recourse. That all relies on the logging system and all the interconnecting systems not glitching. I think without a doubt some cases are drone or system faults but there is no proof 'recorded' so too bad for us. Its the way it is, so better to get insurance to cover "user error".

Just the other day while idling with motors spinning on my desk the go4 app had an odd quirk when turning on/off sensors and then drone just stopped the prop/motors spinning on its own. Weird stuff happens (I would not be surprised if the data - if anything was recorded to say what happened - would come out in a "user error" finding (even though the remote was set aside 1m away untouched.

My point is, did the log in my weird event create and entry "Sensor error turning on/off sensors. Go4 app failed to save sensor switch state to on. App is not working properly. Motor have been turned off as a result automatically". Of course not. But thats what actually happened. Glad I wasn't int he air at the time.

 

[sar104]

I think DJI will never admit liability unless a log entry that absolutely shows a fault. If a drone is at fault but does not log it then we are left with no recourse. That all relies on the logging system and all the interconnecting systems not glitching. I think without a doubt some cases are drone or system faults but there is no proof 'recorded' so too bad for us. Its the way it is, so better to get insurance to cover "user error".

Just the other day while idling with motors spinning on my desk the go4 app had an odd quirk when turning on/off sensors and then drone just stopped the prop/motors spinning on its own. Weird stuff happens (I would not be surprised if the data - if anything was recorded to say what happened - would come out in a "user error" finding (even though the remote was set aside 1m away untouched.

My point is, did the log in my weird event create and entry "Sensor error turning on/off sensors. Go4 app failed to save sensor switch state to on. App is not working properly. Motor have been turned off as a result automatically". Of course not. But thats what actually happened. Glad I wasn't int he air at the time.

So have you examined the logs, rather than just speculating on what they might or might not have recorded?

 

[Dingoz]

So have you examined the logs, rather than just speculating on what they might or might not have recorded?

Unfortunately not and regrettably deleted them the same day as they were 'no fly' logs (just testing/playing on the desk). The concern did not hit me until later and I kicked myself for purging that log. I mean I purged the flight logs in the phone app (so the text files I think?). I do have a bunch of other logs (.dat files though) which I tried to look at to make sense but got nothing. Is there a way to get info from the .dat files? I really would like to know what happened or was recorded just before the motors stopped on their own.

My main point was, and maybe it does have a log fault in my case or something explainable, that I do not for second think DJI logs all its own faults/quirks so readily and that a logging system itself is only as good as its coded for what to log. Flight sensors sure (app crashes and abnormalities not so much if at all).

I created a thread here:

Phantom 4 motors randomly turned off - bug or feature?

And have all the .dat files and anything i could salvage backup up on pc. There is dat files dated the time of the on desk testing and incident. Just no in Go4 app flight record as I purged that before wanting to look into this occurrence. If anyone can help get anything meaningful I would love to know what happened.

 

[sar104]

Unfortunately not and regrettably deleted them the same day as they were 'no fly' logs (just testing/playing on the desk). The concern did not hit me until later and I kicked myself for purging that log. I mean I purged the flight logs in the phone app (so the text files I think?). I do have a bunch of other logs (.dat files though) which I tried to look at to make sense but got nothing. Is there a way to get info from the .dat files? I really would like to know what happened or was recorded just before the motors stopped on their own.

My main point was, and maybe it does have a log fault in my case or something explainable, that I do not for second think DJI logs all its own faults/quirks so readily and that a logging system itself is only as good as its coded for what to log. Flight sensors sure (app crashes and abnormalities not so much if at all).

I created a thread here:

Phantom 4 motors randomly turned off - bug or feature?

And have all the .dat files and anything i could salvage backup up on pc. There is dat files dated the time of the on desk testing and incident. Just no in Go4 app flight record as I purged that before wanting to look into this occurrence. If anyone can help get anything meaningful I would love to know what happened.

You can extract the DAT files as described here:

CsvView/DatCon

 

[MagicCarpetRide]

Nothing why should they give you anything for your incompitance !!!!!!!!!!!!!!

This sort of reply is pretty ugly and unhelpful, not to mention arrogant and judgemental. It's also based on stuff all cherry picked evidence.

The flight log video shows a sudden drop in altitude and does not show any sudden drop in the throttle - it remains very close to the neutral position - the phantom should have maintained altitude. Even if he was incompetent and accidentally put the throttle all the way down, the phantom still wouldn't have crashed into the water if it was functioning correctly; it descends at around 1m / s and would have taken 3 seconds to hit the water.

All you people shouting "pilot error!" are choosing to ignore the altitude data and exaggerate throttle data.
You are also assuming there are no bugs in DJI firmware and software, even though we see so many critical updates being rolled out.

Now, I've never met this person, but I don't think he crashed the drone, because he's basically reported it falling out of the sky - that's not a crash - that's a malfunction and it's consistent with the data.

While the VPS systems do have trouble over water, this is not a reason for the drone to fall out of the sky; the barometric sensors are very accurate.

The claims about it being risky to fly at 9 feet in Sport Mode are rubbish. Yes, the drone tilts forward - and thrust is automatically increased to cope - this is what a flight computer does. I have flown my drone low and fast plenty of times.

The claims about him not being able to see the height of the drone are also pretty frail. He was on a boat at a height between the drone and the water - you can't be in a better position to judge the height!

Finally, I myself have also witnessed the phantom 4 descending uncontrollable when in return to home mode, not properly responding to user input. I nearly lost mine right before my eyes in the river like this - and no it was not pilot error - it was something to do with RTH mode.

In any case DJI's job is to create a predictable, reliable drone that is easy to fly. If they can't do that then they better find something else to do - because that's what their whole business is built on. Dismissing your fellow flyers on the forum as incompetent to elevate your own self esteem as a gifted pilot instead of listening to them is dumb. Try being kind and offering support instead.

 

[Meta4]

The flight log video shows a sudden drop in altitude and does not show any sudden drop in the throttle - it remains very close to the neutral position - the phantom should have maintained altitude. Even if he was incompetent and accidentally put the throttle all the way down, the phantom still wouldn't have crashed into the water if it was functioning correctly; it descends at around 1m / s and would have taken 3 seconds to hit the water.

All you people shouting "pilot error!" are choosing to ignore the altitude data and exaggerate throttle data.

In any incident investigation you have to work out what evidence can be relied on and what is potentially unreliable.
The recorded flight data has been proved reliable hundreds of times.
Eyewitness accounts on the other hand, can be affected by confusion, misunderstanding and even deception and where they contradict the recorded data, I would accept the data almost every time.

We seen few posts like yours lately suggesting that the data could be inaccurate but none have ever shown an example of faulty flight data to support their claim.
Just an assertion that it might be faulty.

You are completely ignoring the relevant data and your faith in the accuracy of barometer for altitude recording is misplaced.
It is very common for the altitude shown by the barometric sensor to be out by a few metres from the start of a flight to the end due to heating up during the flight and possible small changes in air pressure during that time.
The absolute height shown by the barometer cannot be relied on but the trend can.
The assumption that his Phantom was 29 feet above the water because the barometer indicates 29 ft is likely to be in error.
Even the OP didn't claim that his Phantom fell 29 feet into the water. His guesstimate was 8 feet.
The data shows the left stick pulled down to an average position of 843 from 7:16.1 and still in that position when the flight record ends.
This is matched by a steady descent from an indicated altitude of 51.5 ft until the end of the flight record at 7:22.7 when it shows 28.5 ft.
Over this time the rate of descent has varied between 3-5 feet per second.
Now if you are flying at 40+ mph just a few feet above the water and descending at 3-5 ft per second .... what do you suppose is the most likely outcome?

The claims about it being risky to fly at 9 feet in Sport Mode are rubbish. Yes, the drone tilts forward - and thrust is automatically increased to cope - this is what a flight computer does. I have flown my drone low and fast plenty of times.

Flying close to obstacles (the ground is an obstacle) is risky at any time.
Flying close to obstacles at 40 mph is very much more risky.and many times more risky at 40 mph.
Continue to enjoy your risky flying. Pay attention to how well your Phantom maintains altitude in sport mode and what it does when you ease off the sticks.

Finally, I myself have also witnessed the phantom 4 descending uncontrollable when in return to home mode, not properly responding to user input. I nearly lost mine right before my eyes in the river like this - and no it was not pilot error - it was something to do with RTH mode.

Something to do with RTH mode is extremely vague and explains nothing.
If you want to learn from the incident to help you avoid it in future, it would benefit you to investigate the actual flight data to find out what really happened.

In any case DJI's job is to create a predictable, reliable drone that is easy to fly. If they can't do that then they better find something else to do - because that's what their whole business is built on.

Do you have any reason to believe that DJI aren't producing predictable, reliable drones that are easy to fly?
From my experience, Id say that is exactly what they do and that they do it very well.

Dismissing your fellow flyers on the forum as incompetent to elevate your own self esteem as a gifted pilot instead of listening to them is dumb. Try being kind and offering support instead.

There are only a few members here that analyse flight data and if that's what you think they do, you've not read many of their incident reports.
Sometimes they are able to show that the cause of an incident was a malfunction and advise the pilot to help them in dealing with DJI, sometimes not.
But the result is always of great educational benefit to all pilots that want to learn from unfortunate flight incidents and prevent repeating them in the future.
That's offering real support.

 

[MagicCarpetRide]

In any incident investigation you have to work out what evidence can be relied on and what is potentially unreliable.
The recorded flight data has been proved reliable hundreds of times.
Eyewitness accounts on the other hand, can be affected by confusion, misunderstanding and even deception and where they contradict the recorded data, I would accept the data almost every time.

We seen few posts like yours lately suggesting that the data could be inaccurate but none have ever shown an example of faulty flight data to support their claim.
Just an assertion that it might be faulty.

I believe I said that people had exaggerated throttle data while ignoring altitude data. I may have been incorrect, but please don't misrepresent what I said.

You are completely ignoring the relevant data and your faith in the accuracy of barometer for altitude recording is misplaced.

Ok, enlighten me on which data is relevant if the throttle and altitude are not.

It is very common for the altitude shown by the barometric sensor to be out by a few metres from the start of a flight to the end due to heating up during the flight and possible small changes in air pressure during that time.
The absolute height shown by the barometer cannot be relied on but the trend can.

I think we agree on this. The phantom should have known it was descending based on the "trend". Now I watched the video of the stick positions - was this not a reliable visual representation of the data?

The assumption that his Phantom was 29 feet above the water because the barometer indicates 29 ft is likely to be in error.
Even the OP didn't claim that his Phantom fell 29 feet into the water.

I didn't claim that. Where did you pull that from?

Something to do with RTH mode is extremely vague and explains nothing.

Well it's not the incident being discussed here; my intention was merely to state that I have witnessed malfunctions. We could go into the specifics of other incidents in other posts.

Do you have any reason to believe that DJI aren't producing predictable, reliable drones that are easy to fly?

Yes. While they are much easier to learn to fly than racing drones in acrobatic mode, they can behave more unpredictably at times, due to the complexity of the software involved. Often people complain about this here, only to be told they are just incompetent. I challenge the notion that everyone is incompetent and that DJI have a perfect product.

There are only a few members here that analyse flight data and if that's what you think they do, you've not read many of their incident reports.
Sometimes they are able to show that the cause of an incident ... ...the result is always of great educational benefit to all pilots that want to learn from unfortunate flight incidents and prevent repeating them in the future.
That's offering real support.

Yes it is - but that's not what the person I replied to was doing at all.

 

[Meta4]

I believe I said that people had exaggerated throttle data while ignoring altitude data. I may have been incorrect, but please don't misrepresent what I said.

I have no idea where I may have misrepresented you at all. Rereading, I can't see it at all.

I believe I said that people had exaggerated throttle data while ignoring altitude data. I may have been incorrect, but please don't misrepresent what I said.
Ok, enlighten me on which data is relevant if the throttle and altitude are not.

In my extensive analysis for the OP and in restating parts to you, I made it clear that altitude data and throttle data are both of critical value.
Why would you think I suggested anything else?
I put equal value and credibility on both, neither ignoring nor exaggerating either.

I think we agree on this. The phantom should have known it was descending based on the "trend". Now I watched the video of the stick positions - was this not a reliable visual representation of the data?

I think we agree on this. The phantom should have known it was descending based on the "trend". Now I watched the video of the stick positions - was this not a reliable visual representation of the data?

Should have known?? The Phantom knew it was descending. The data makes that very obvious and I gave you the numbers to back that up.
The actual numbers tell a lot more than tiny graphic joystick representations do.

I didn't claim that. Where did you pull that from?
You said >>the barometric sensors are very accurate<<
The barometric sensor gave an altitude reading of 28.5 ft. It's obvious that it was not how high the Phantom was.

Well it's not the incident being discussed here; my intention was merely to state that I have witnessed malfunctions. We could go into the specifics of other incidents in other posts.

If you can't back up your assertion that flight data can be unreliable, all I can say is that you believe you have witnessed malfunctions.
I deal with flyers every week that believe their Phantom malfunctioned only to have the recorded data tell a very different and much more credible story.

Yes. While they are much easier to learn to fly than racing drones in acrobatic mode, they can behave more unpredictably at times, due to the complexity of the software involved. Often people complain about this here, only to be told they are just incompetent. I challenge the notion that everyone is incompetent and that DJI have a perfect product.

Except in the case of fairly uncommon hardware malfunctions, I don't believe I've seen any cases of Phantoms behaving unpredictably.
People often complain about such things but the data usually shows a quite reasonable explanation and can show the flyer was confused, disoriented, lacked proper understanding of how the Phantom works or occasionally was intentionally deceptive.

Your notion that I think everyone is incompetent is nonsense.
I think you don't appreciate the time and effort that goes into a proper investigation here.
We give the owners a far better understanding and explanation of what really happened than DJI ever do.
We also frequently offer a likely search area and have been responsible for finding quite a few lost Phantoms all over the world. This is just more proof of the value of the flight data.
Having someone like yourself dismissing investigations like this as unreliable but not bothering to back up your assertion is quite offensive to the people that put in the time to do it.

 

[MagicCarpetRide]

Um... you keep misunderstanding. You don't answer my questions. You're argumentative. You take things personally that were not even directed at you. Just forget it please.