GPS loss

[dennyc39]

New pilot, no experience (yet).

I understand that if the drone is not happy with GPS signal, it goes into ATTI mode, and will drift with the wind.

1. Can it still stabilize altitude if within range of the ground sensors?
2. Is there any way to tell which way it is pointing besides the lights? Does the remote have a heading display?
3. If the pilot then climbs it to say 150 ft to clear obstacles, does it stop climbing when stick is centered, or does it keep drifting up due to lack of altitude awareness?
4. does the yaw stop when sticks are centered, or does it also drift in yaw?

I have no experience yet, but it seems the ATTI mode compromise is actually an emergency of some magnitude, since there seems to be no way to practice handling and control in ATTI mode, and presumably no way to return to a more benign flight mode.

 

[sar104]

New pilot, no experience (yet).

I understand that if the drone is not happy with GPS signal, it goes into ATTI mode, and will drift with the wind.

1. Can it still stabilize altitude if within range of the ground sensors?
2. Is there any way to tell which way it is pointing besides the lights? Does the remote have a heading display?
3. If the pilot then climbs it to say 150 ft to clear obstacles, does it stop climbing when stick is centered, or does it keep drifting up due to lack of altitude awareness?
4. does the yaw stop when sticks are centered, or does it also drift in yaw?

I have no experience yet, but it seems the ATTI mode compromise is actually an emergency of some magnitude, since there seems to be no way to practice handling and control in ATTI mode, and presumably no way to return to a more benign flight mode.

ATTI mode can be caused by one of two issues. One, obviously, is loss of GPS lock, but that is not the most common cause. The FC also switches to ATTI if the primary flight control system solution for heading, location or velocity (usually heading), which is derived from the IMU rate gyros and accelerometers, disagrees substantially with the magnetometer, GPS location or GPS velocity. When that happens, the sensor fusion algorithms that correct the IMU solution for drift and bias are no longer viable and, since the GPS data are too low frequency to control the aircraft, the FC ignores them and goes to ATTI.

ATTI stabilizes attitude, including yaw provided that the compass is working, and altitude. It doesn't need ground sensors to do that.

 

[dennyc39]

Thanks. It's interesting that it can still stabilize altitude. Maybe there's an internal barometric sensor?

 

[BD0G]

There is a barometric sensor that determines altitude. I gleaned this from reading forum posts.

 

[sar104]

Thanks. It's interesting that it can still stabilize altitude. Maybe there's an internal barometric sensor?

Yes - no mystery about that. It uses barometric pressure as its absolute altitude sensor.

 

[dbyantis]

New pilot, no experience (yet).

I understand that if the drone is not happy with GPS signal, it goes into ATTI mode, and will drift with the wind.

1. Can it still stabilize altitude if within range of the ground sensors?
2. Is there any way to tell which way it is pointing besides the lights? Does the remote have a heading display?
3. If the pilot then climbs it to say 150 ft to clear obstacles, does it stop climbing when stick is centered, or does it keep drifting up due to lack of altitude awareness?
4. does the yaw stop when sticks are centered, or does it also drift in yaw?

I have no experience yet, but it seems the ATTI mode compromise is actually an emergency of some magnitude, since there seems to be no way to practice handling and control in ATTI mode, and presumably no way to return to a more benign flight mode.

Regarding getting the heading of the AirCraft, AC, if the display is still active, use the "radar" in the lower left for orientation, without display, then you have to use visual cues while trying left right with the sticks and observe how the AC reacts. If the controller is still connected, you can use the telemetry, distance, alt, etc from the controller to judge when you are getting closer to home.

 

[msinger]

Regarding getting the heading of the AirCraft, AC, if the display is still active, use the "radar" in the lower left for orientation

Here are a few videos that explain how the attitude indicator works in DJI GO:

 

[theDRONEranger]

Here is another tidbit on GPS Loss:

https://www.facebook.com/UAVForecast/posts/1563201620624110

Helps for a better understanding. But remember there are also other reasons GPS signals drop!

 

[sar104]

Here is another tidbit on GPS Loss:

https://www.facebook.com/UAVForecast/posts/1563201620624110

Helps for a better understanding. But remember there are also other reasons GPS signals drop!

UAV Forecast likes to play up the significance of geomagnetic disruption to GNSS, but the basis for their assertions seems a bit weak to me. For example, they cite just one reference, Akala, A. O., et al. (2012), Impacts of ionospheric scintillations on GPS receivers intended for equatorial aviation applications, Radio Sci., 47, RS4007., which is a study of equatorial scintillation effects. It's not globally applicable and it was based on just GPS reception of 6 - 8 satellites. All the modern UAVs use dual system, single frequency GNSS, which gives a much more robust solution. Soon all these aircraft will use the newer dual frequency L1/L2 chips, which should make this a complete non-issue for navigation systems.

There are other studies that demonstrate very little effect outside the polar and equatorial regions and better performance from dual system chips. For example, for an easily understood presentation:

Equatorial Scintillation Impact on GNSS Precise Positioning Services

 

[theDRONEranger]

I’ve not experienced any GPS significant drop ever. It was posted more as a possible point of interest for those who have experienced flyaway as a distant possibile occurance asa result of low GPS grabs. I read of quite a few flyaways, but seemingly almost never see a repost as to the defined cause. Be nice to see, or learn, some of the underlying causes that perhaps others may learn to avoid.

 

[sar104]

I’ve not experienced any GPS significant drop ever. It was posted more as a possible point of interest for those who have experienced flyaway as a distant possibile occurance asa result of low GPS grabs. I read of quite a few flyaways, but seemingly almost never see a repost as to the defined cause. Be nice to see, or learn, some of the underlying causes that perhaps others may learn to avoid.

I'm puzzled - most of the reported "flyaways" on this forum and PP are fully explained in the threads by analysis of the logs. I can recall just a handful where the cause was not determined. It's almost always magnetic interference at takeoff causing subsequent yaw errors. Occasionally it's ATTI mode due to locations with very poor sky view and hence no position lock. There have been a few due to faulty GPS systems - characterized by random drops to zero. I cannot remember any that looked like candidates for compromised position lock due to geomagnetic storms.

 

[Meta4]

I’ve not experienced any GPS significant drop ever. It was posted more as a possible point of interest for those who have experienced flyaway as a distant possibile occurance asa result of low GPS grabs. I read of quite a few flyaways, but seemingly almost never see a repost as to the defined cause. Be nice to see, or learn, some of the underlying causes that perhaps others may learn to avoid.

If it was a thing, we'd see plenty of it at times of high KP-index.
But we don't.
The next flight incident we see that can be attributed to high KP-index will be the first I've seen in 5 years of DJI flying.

 

[DanMan32]

Off topic but related: you would think that if there was a compass offset due to mag interference on launch, it would use the GPS to correct for the offset for the flight. With a few GPS coordinates, it should be able to figure out where north really is. At least allow enough to get back home safe.

 

[sar104]

Off topic but related: you would think that if there was a compass offset due to mag interference on launch, it would use the GPS to correct for the offset for the flight. With a few GPS coordinates, it should be able to figure out where north really is. At least allow enough to get back home safe.

Using GPS it can figure out which it is moving, but not which way it is facing.

 

[DanMan32]

Once it moves laterally, it can calculate the compass offset from the GPS delta and use the offset temporary. Or ignore the compass altogether for that time and rely on IMU and GPS only. That's pretty much how they did it heading to the moon and back, except instead of GPS, they used stars, even when once they accidentally hit the code to reset the IMU to ground zero.

 

[sar104]

Once it moves laterally, it can calculate the compass offset from the GPS delta and use the offset temporary. Or ignore the compass altogether for that time and rely on IMU and GPS only. That's pretty much how they did it heading to the moon and back, except instead of GPS, they used stars, even when once they accidentally hit the code to reset the IMU to ground zero.

No - that's not correct. Compass data don't give direction of travel - only the orientation of the aircraft, and GPS position data don't give orientation - only direction of travel. Those two values are independent, and so one does not inform the other.

 

[Prismatic]

Compass data don't give direction of travel - only the orientation of the aircraft, and GPS position data don't give orientation - only direction of travel. Those two values are independent, and so one does not inform the other.

And, I'd add, the direction of travel for an airborne drone is not entirely determined by the stick inputs; wind speed and direction play in. So trying to calculate orientation by comparing GPS-detected motion to stick inputs can't be reliable.
That said, wouldn't it be possible to do that in zero-wind? That is, in windless conditions, the drone (with a known compass error, but with working GPS) notes that forward stick moves the drone in, say, a SE direction. Wouldn't that determine true north? (But again, the notion is wholly unreliable in the real world.)

 

[sar104]

And, I'd add, the direction of travel for an airborne drone is not entirely determined by the stick inputs; wind speed and direction play in. So trying to calculate orientation by comparing GPS-detected motion to stick inputs can't be reliable.
That said, wouldn't it be possible to do that in zero-wind? That is, in windless conditions, the drone (with a known compass error, but with working GPS) notes that forward stick moves the drone in, say, a SE direction. Wouldn't that determine true north? (But again, the notion is wholly unreliable in the real world.)

Yes - a trial and error method could be used to determine orientation, or even used for navigation, but it's going to result in a very drunken flight path and nothing like that has been implemented.