P3P Flyaway

[Avocet]

Are you serious????
LMAO!!!
That's funny as hell.

I'll bet the friggin' gull didn't think so!

 

[Vertigo]

One example of those doesnt discredit all the others. What conceivable user error could lead to a fly away like this:

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

or this:

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

or a bazillion others on youtube ?

Yes, some of them arent fly aways and can be explained by user ignorance like the one you picked. That doesnt mean actual fly aways (out of control) dont happen or can be blamed on the user.

 

[m0j0]

I was standing next to a guy flying his P3s. He flew behind a water tower and into a pole. Immediately said, I lost control and it flew away.

Kind of looked like he flew into a telephone line then clipped the pole. He called it a flyaway. I called it flying into a telephone line. Eh... Tomato. Tomoto...


Sent from my iPad using PhantomPilots mobile app

 

[Swifteagle]

Question:

So when this so call flyaways happen, can you switch it to manual mode and take charge of the phantom? Do you lose lightbride?

Because if that is the case, it would be quite easy to bring it back.

Thanks

 

[msinger]

So when this so call flyaways happen, can you switch it to manual mode and take charge of the phantom?

You can manually switch flight modes at any point during your flight. If your Phantom is flying away because you calibrated the compass next to a large garbage truck (for example), changing the flight mode likely won't help you regain control of your Phantom.

Do you lose lightbride?

You could if your Phantom flies beyond the range of Lightbridge and/or is in an area with a lot of interference.

The best way to avoid a fly away is to read the Phantom manual and understand how your Phantom works. If you understand it, then you will know exactly what it's doing during flight and will not have to blame your lack of understanding on a fly away. Phantoms rarely legitimately fly away on their own.

 

[losman26]

hmm. Mine was a flyway. Nobody else can convince me of that. It happens, and I was one of the unlucky ones to have a defective product.

 

[Zuzua]

hmm. Mine was a flyway. Nobody else can convince me of that. It happens, and I was one of the unlucky ones to have a defective product.

Upload the flight data


Sent from my iPhone using PhantomPilots mobile app

 

[losman26]

Upload the flight data


Sent from my iPhone using PhantomPilots mobile app

already did that in another thread.

 

[Meta4]

One example of those doesnt discredit all the others. What conceivable user error could lead to a fly away like this:
or a bazillion others on youtube ?

Without an explanation of what's happening in those videos and flight data it's not possible to work out what the incident was.
But the P3 has a black-box recorder that can be interrogated to find out what actually happened rather than what a confused, misinformed or ignorant user thinks happened.
The fact that there are lots of youtube videos purporting to show flyaways doesn't mean anything.
When alleged P3 flyaways are investigated with flight data, it turns out that the majority are simply explained as flying behind obstacles or terrain, flying downwind in strong wind situations, not knowing how to activate RTH, not knowing that RTH has activated, bad compass calibrations, disorientation etc etc.

Yes, some of them arent fly aways and can be explained by user ignorance like the one you picked. That doesnt mean actual fly aways (out of control) dont happen or can be blamed on the user.

Actual flyaways are much rarer than you would expect if you accept youtube accounts at face value.

 

[Meta4]

So when this so call flyaways happen, can you switch it to manual mode and take charge of the phantom? Do you lose lightbridge?

Since these so called flyaways are caused by a whole range of things (and almost all are related to the operation of the Phantom), no simple answer can be given to your question.
There is no manual mode for the P3 anyway.
There is atti mode that ignores input from the GPS, but switching to atti mode won't make any difference in most hardware malfunctions and most user errors.

 

[Vertigo]

Without an explanation of what's happening in those videos and flight data it's not possible to work out what the incident was.

Im asking you to speculate. What possible user error could ever lead to such incidents and would be unavoidable even with proper flight control software? I'll even give you one: mounting the props backwards. But it wont fly very far that way.

When alleged P3 flyaways are investigated with flight data, it turns out that the majority are simply explained as flying behind obstacles or terrain, flying downwind in strong wind situations, not knowing how to activate RTH, not knowing that RTH has activated, bad compass calibrations, disorientation etc etc.

None of those should lead to an actual, out of control fly away (not an in-control flight towards a home- or waypoint).

 

[Meta4]

Im asking you to speculate. What possible user error could ever lead to such incidents and would be unavoidable even with proper flight control software?.

Get me a description of the incident and flight data and I'll investigate any incident.
But I watched those videos and couldn't even understand what they thought the problem was from what I saw.
And since so many alleged incidents turn out to be operator error or confusion when investigated, that's the assumption I follow until some evidence is offered.

 

[Vertigo]

But I watched those videos and couldn't even understand what they thought the problem was from what I saw..

Really? The one over the jeeps, thats a pitch/roll angle you can not even command, and was clearly not commanded, it flew away at insane speed. Also read the text below.

The one in the city; just compare the sticks with the flight path. It was flying away 1000's of ft for no apparent reason (PGPS mode, and no corresponding control input) before it finally switched to RTH and continued flying somewhere quite different than the indicated home position.

I just picked two random examples on youtube, I can post these all day long. The bottom line is that it happens, and you better be prepared that it can happen to you.

And since so many alleged incidents turn out to be operator error or confusion when investigated, that's the assumption I follow until some evidence is offered

I noticed that. Just like I noticed you (incorrectly) assume the compass and GPS are immune to problems as long as you calibrate every flight. You have too much faith.

 

[Meta4]

I just picked two random examples on youtube, I can post these all day long. The bottom line is that it happens, and you better be prepared that it can happen to you.

And if you get a genuine malfunction, causing crazy uncommanded flight there's no amount of preparation or experience that will help you.
It's fortunate that these occur rarely.

I noticed that. Just like I noticed you (incorrectly) assume the compass and GPS are immune to problems as long as you calibrate every flight. You have too much faith.

Then you didn't notice anything.
I would never suggest calibration before a every flight is a good thing.
But the GPS and Compass (providing it isn't screwed up by a bad calibration or taking off from on top of reinforced concrete etc) are very reliable and responsible for far fewer incidents than they get blamed for.

 

[Vertigo]

And if you get a genuine malfunction, causing crazy uncommanded flight there's no amount of preparation or experience that will help you.

No, but your selection of where to fly may have a large impact on the outcome. If you're under the false impression your phantom is anywhere near as reliable as a manned aircraft, you're prone to do stupid things. Like flying 3 feet above your teenage kids. Treat it like the inherently unreliable consumer toy that it is, and chances of causing real damage will be dramatically reduced.

As for the "genuine malfunction"; I put it to you that not even a hardware malfunction can plausibly explain these kinds of fly aways. A malfunction of primary flight sensors (gyro and accelerometer) would most likely cause your quad to wobble and flip like a madman, and crash in seconds. It would have to be some really freaky sensor malfunction that would allow it to maintain a level attitude and altitude, let alone pull off a controlled landing miles away. A hardware malfunction or environmental issue with the secondary, navigation sensors (compass and GPS) can and should be handled in a way that doesnt involve flying 3 miles away from its take off position.

The only plausible explanation I see is soft/firmware issues, and since we tend to fly with the same firmware, its just a matter of when those bugs will bite you.

 

[Meta4]

As for the "genuine malfunction"; I put it to you that not even a hardware malfunction can plausibly explain these kinds of fly aways. A malfunction of primary flight sensors (gyro and accelerometer) would most likely cause your quad to wobble and flip like a madman, and crash in seconds. It would have to be some really freaky sensor malfunction that would allow it to maintain a level attitude and altitude, let alone pull off a controlled landing miles away. A hardware malfunction or environmental issue with the secondary, navigation sensors (compass and GPS) can and should be handled in a way that doesnt involve flying 3 miles away from its take off position.

I'm sorry but I'm not sure what you're getting at or what you mean.
If you want to talk about a real case with real data, I'd be happy to comment but it's difficult to respond to vague hypothetical situations that may or may not even happen,

 

[Vertigo]

Lets recap; OP asked if he should be afraid of fly aways. You respond by saying its extremely unlikely and almost all reported incidents are caued by user error. I ask you to explain what user error could lead to an actual fly away, and I have not gotten an answer beyond stating that many reported fly aways are not truly fly aways, which is true.

So I posted a few examples of what I consider actual fly away and asked what user error could possibly, conceivably have led to those, and numerous similar incidents. No meaningful response. I can only conclude you agree with me that actual fly aways can never be blamed on user error, and yet they do happen. So the risk is real.

How big, is really hard to quantify, but I think you are seriously underestimating the problem. A few years ago, there was a poll on RCG:
RC Groups - View Poll Results

yes, its an old thread, and its plausible, even likely current models are more reliable. I also realize there may be a selection bias in that poll, so those numbers should be taken with a spoon of salt. But the number (33% having suffered a fly away, half of them losing their drone as a result) was still pretty staggering. A more rigorous but similar poll about Naza fly aways gave even worse results (46% fly aways):
DJI Phantom Fly Aways Survey

And yet back then, you'd hear most users that happened to be in the 66% lucky user group claim the product was just fine, and almost all fly aways where user error. I didnt buy it then, nor do I buy it now.

My TL : DR for the OP: yes you should be worried about the potential of a fly away thats not by your own doing. Over the life time of the drone, the risk may be 1% it may be 20%, no one knows for sure, but if you cant afford to lose the investment, dont buy one. And if you have more than an infinitesimal chance of hurting someone or something valuable if your drone goes out of control, dont fly there.

 

[moswissa]

If the Phantom gets conflicting information from the compass and GPS sensors, it will drop the GPS info and slip into atti mode.
Interference might cause you to lose signal which would trigger RTH but it would be most unlikely to have any affect on your compass or GPS.
It's much more likely that your incident was caused by a compass error, possibly caused by a bad calibration.
It's easy enough to look into the flight data and see what happened.
Go to https://www.phantomhelp.com/LogViewer/Upload/
Follow the instructions to upload your flight record.

Ive noticed recently after the last update there has been an increase in compass errors mid flight, I have had one myself. I have noticed more complaints about it here on the forum. its like the last update made the GPS compass very sensitive. Add another worry to our list, first the battery voltages and now this.

 

[Meta4]

Lets recap; OP asked if he should be afraid of fly aways. You respond by saying its extremely unlikely and almost all reported incidents are caused by user error. I ask you to explain what user error could lead to an actual fly away, and I have not gotten an answer beyond stating that many reported fly aways are not truly fly aways, which is true. I can only conclude you agree with me that actual fly aways can never be blamed on user error, and yet they do happen. So the risk is real.

Ok .. I've seen 3 cases of the P3 Standard resetting its home point and RTH sending the Phantom off to the new home point or complications due to having distance limit and the new home point being outside the distance limit.
And There are occasional cases of genuine malfunctions that create an uncontrollable Phantom.
These are rare events and hugely outnumbered by cases that are labelled as flyaways but on investigation turn out to be something else.

How big, is really hard to quantify, but I think you are seriously underestimating the problem. A few years ago, there was a poll on RCG:
RC Groups - View Poll Results
you'd hear most users that happened to be in the 66% lucky user group claim the product was just fine, and almost all fly aways where user error. I didnt buy it then, nor do I buy it now.

That's not just old - that's pre-P3.
The P3 was a technological leap forward and is remarkably free of flyaways.
If your thinking is stuck back in the P2 era you need to bring yourself up to date.
P2 flyaway discussion is totally irrelevant in a P3/4 era.

My TL : DR for the OP: yes you should be worried about the potential of a fly away thats not by your own doing. Over the life time of the drone, the risk may be 1% it may be 20%, no one knows for sure, but if you cant afford to lose the investment, dont buy one. And if you have more than an infinitesimal chance of hurting someone or something valuable if your drone goes out of control, dont fly there.

If you really think it may be 20% you have no idea.
There would be hardly any posts on forums about anything else and DJI would be out of business.
It's much, much less than 1%.
The whole concept of a P3 flyaway is perpetuated by ridiculous discussions like this that cause newbies to believe that their P3 is likely to flyaway.
The real situation is quite different from forum myths and youtube would suggest.

 

[Vertigo]

It's much, much less than 1%.

So you truly believe your drone is (far) more reliable than your laptop, car or even kitchen appliance?

If you are convinced of that, you stand to make a lot of money by selling fly away insurances. I'll gladly pay you 1% of my purchase price if you refund my drone when it flies away and you can not demonstrate user error. Where do I send my €7 ? Other companies charge around 35x more for a year extended warranty.