Filipok lost drone adventure

[sar104]

The unit lifted off, was 25 ft AGL in GPS and flipped into ATTI because of a sys-warning error (that was never logged). The unit had a home point and knew it's lat long from take-off. I know this because it logged it.
In this case, the drone should land immediately instead of drifting half a mile and randomly landing.

No - that's not correct. The aircraft will only autoland automatically in ATTI either due to low battery or if the uplink is lost. Neither of those happened until later. While it has a connection it will simply hold attitude and altitude and wait for stick inputs. If you don't actively control it then it will drift with the wind, exactly as you saw.

 

[Former Member07122019a]

No - that's not correct. The aircraft will only autoland automatically in ATTI either due to low battery or if the uplink is lost. Neither of those happened until later. While it has a connection it will simply hold attitude and altitude and wait for stick inputs. If you don't actively control it then it will drift with the wind, exactly as you saw.

That's incorrect. The drone was completely unresponsive to the sticks and would also not respond to a manual rth. When I found the unit, it had landed while connected at 73% battery. This is clearly a firmware bug.

 

[Keule]

The drone was completely unresponsive to the sticks

Did you verify that statement in interpreting your log files (.txt AND .dat)?

not respond to a manual rth

that's explained in post #27

it had landed while connected at 73% battery

See post #35
The aircraft will only autoland automatically in ATTI either due to low battery or if the uplink is lost.

This is clearly a firmware bug.

That is exactly how DJI vehicles work when in ATTI mode and this is on purpose and not a bug.

 

[sar104]

See post #35
The aircraft will only autoland automatically in ATTI either due to low battery or if the uplink is lost.

Did you mean to cross out the battery statement? I assume that you are not saying that the aircraft won't land due to low battery.

 

[DanMan32]

I still stand on the point that if it has a compass error but still good GPS, it still should be stable enough to fly. Sure, a static GPS position won't give you heading, but the difference between two GPS position will.
However in this case with fail safe ATTI mode due to sys error, other flight stability sensors and functions might not be there. You still need accurate gyro and accelerometer readings. It might not even be able to fly with manual stick without IMU.

I don't recall if you mentioned it, but were you able to test fly it OK after recovering? You could after all have a hardware problem.

 

[sar104]

That's incorrect. The drone was completely unresponsive to the sticks and would also not respond to a manual rth. When I found the unit, it had landed while connected at 73% battery. This is clearly a firmware bug.

Not responding to stick inputs while the uplink is good would be a fault, if that's what happened. But I'm confident that if we looked at the log file we would find that the aircraft responded to stick inputs. It cannot RTH, manually or otherwise, in ATTI mode, as explained previously. If it landed at 73% battery that will be because it lost uplink. It will all be in the log file.

 

[Keule]

Did you mean to cross out the battery statement? I assume that you are not saying that the aircraft won't land due to low battery.

I assumed by the statement:

When I found the unit, it had landed while connected at 73% battery

the aircraft did not initiate landing due to low battery. Therefore I crossed it out

 

[sar104]

I still stand on the point that if it has a compass error but still good GPS, it still should be stable enough to fly. Sure, a static GPS position won't give you heading, but the difference between two GPS position will.
However in this case with fail safe ATTI mode due to sys error, other flight stability sensors and functions might not be there. You still need accurate gyro and accelerometer readings. It might not even be able to fly with manual stick without IMU.

I don't recall if you mentioned it, but were you able to test fly it OK after recovering? You could after all have a hardware problem.

Change in GPS position gives you a track, not a heading. Navigation by observation of track is trial and error. The rest of this discussion is degenerating into speculation vs. operating principles, and, since the OP is obviously not going to share the log file, is going nowhere useful.

 

[sar104]

I assumed by the statement:

When I found the unit, it had landed while connected at 73% battery

the aircraft did not initiate landing due to low battery. Therefore I crossed it out

Oh - of course.

 

[paulatkin73]

Sure, a static GPS position won't give you heading, but the difference between two GPS position will.

it can be done for plane models, but not for drones that are expected to move in any direction. for those compass accuracy is absolutely necessary.

 

[Former Member07122019a]

I still stand on the point that if it has a compass error but still good GPS, it still should be stable enough to fly. Sure, a static GPS position won't give you heading, but the difference between two GPS position will.
However in this case with fail safe ATTI mode due to sys error, other flight stability sensors and functions might not be there. You still need accurate gyro and accelerometer readings. It might not even be able to fly with manual stick without IMU.

I don't recall if you mentioned it, but were you able to test fly it OK after recovering? You could after all have a hardware problem.

Yes. After power cycling the mpp, it has flown without error since. I have flown over 10 hours so far without incident. As far as the connection to the drone during the malfunction, I am fortunate that it didn't disconnect as it was how I found the drone. The find my drone function led me to within 6 ft. of the craft.

 

[Former Member07122019a]

Sure, a static GPS position won't give you heading, but the difference between two GPS position will.

Exactly. As a senior software architect myself, this is an easy fix.

 

[worldtraveller]

That's incorrect. The drone was completely unresponsive to the sticks and would also not respond to a manual rth. When I found the unit, it had landed while connected at 73% battery. This is clearly a firmware bug.

Very interesting because I have recently list my Mavic 2Pro in bed similar circumstances ( 72% battery, 15 satellites and only about 800m away). Makes me even more certain that some units have a problem. Not just my Mavic 2pro with this problem.

 

[Former Member07122019a]

Very interesting because I have recently list my Mavic 2Pro in bed similar circumstances ( 72% battery, 15 satellites and only about 800m away). Makes me even more certain that some units have a problem. Not just my Mavic 2pro with this problem.

I'm very tempted to do a total factory reset and then update from there.

 

[worldtraveller]

Very important in my case is that, after the disconnection, the drone failed to return-to-home. That also seems to indicate a failure of the drone. I was also unable to reconnect the flight controller even without using the mobile phone. I had hoped to bring the drone back using the distance scale on the controller only, but that was not possible as it would not reconnect.

Very bad that DJI are so far, refusing to provide a replacement drone when it appears that their product failed.

 

[sar104]

Very important in my case is that, after the disconnection, the drone failed to return-to-home. That also seems to indicate a failure of the drone. I was also unable to reconnect the flight controller even without using the mobile phone. I had hoped to bring the drone back using the distance scale on the controller only, but that was not possible as it would not reconnect.

Very bad that DJI are so far, refusing to provide a replacement drone when it appears that their product failed.

In your case the aircraft simply stopped transmitting, and since it was nearly 1 km away and over water you were unable to see what happened or recover the aircraft. It could have been an FC failure or an improperly latched battery but, without the aircraft to examine and determine the cause or other definitive indications of what happened, DJI doesn't generally offer a replacement - just a discount.

In the case discussed in this thread the aircraft failed to return because it was in ATTI mode, and later autolanded. The OP refused to share the flight logs and so the cause is unknown.

 

[Former Member07122019a]

The OP refused to share the flight logs and so the cause is unknown.

As I've explained several times now, I'm a senior software engineer and have reviewed every line of the logs. I do not feel a compulsion to publicly post my flight logs.

My conclusion is that I will most likely not be buying another DJI drone until I see a concerted effort by DJI to remedy these fly-aways, one way or the other. Regardless of the status quo, this is an issue for many. I've already lost one Spark in which it disconnected, never came home, and logged nothing. Most likely, I'll be getting an evo or fimi x8 and selling my mpp before it freaks out again and I lose a grand.

Case closed.

 

[sar104]

As I've explained several times now. I'm a senior software engineer and have reviewed every line of the logs. I do not feel a compulsion to publicly post my flight logs.

My conclusion is that I will most likely not be buying another DJI drone until I see a concerted effort by DJI to remedy these fly-aways, one way or the other.

Case closed.

Yes - you have said that several times. You have also clearly demonstrated that you don't understand how to interpret the flight logs and, as a result, have no idea what actually happened. "Firmware bug" is not an answer. And not wanting to share the logs is completely fine - it's simply made this thread rather a waste of time.

 

[Former Member07122019a]

Yes - you have said that several times. You have also clearly demonstrated that you don't understand how to interpret the flight logs and, as a result, have no idea what actually happened. "Firmware bug" is not an answer. And not wanting to share the logs is completely fine - it's simply made this thread rather a waste of time.

Clearly demonstrated what? Brother, I hate to break it to you but your opinion isn't a final authority on anything.

Get over yourself.

 

[sar104]

Clearly demonstrated what? Brother, I hate to break it to you but your opinion isn't a final authority on anything.

Get over yourself.

As I said - you obviously don't understand what you are looking at or talking about, and the fact that you don't even know why that is obvious is, itself, very telling. But it doesn't bother me at all, so there is nothing to get over. And since I haven't claimed to be the final authority on anything that's yet another meaningless jab. However, I do recommend that you stop wasting everyone's time with your inaccurate speculation.