Lost GPS = No RTH, why? Not an actual incident, just wondering.

[Yorkshire_Pud]

If a drone started out with good GPS but subsequently loses it I am wondering why it has DJI have decided that the drone should land ?
I think the drone maintains a 'direction to home is' value but presume that is only whilst GPS is available. I am wondering, if RTH was initiated 'as soon as' GPS was lost, why the drone can't fly on that heading for a while to see if things improve ?

I AM NOT suggesting that this would be feasible if the entire flight had been without GPS nor am I suggesting it would be possible if the drone had been moved from the loss of GPS point.

I also realise that with out GPS the drone is likely to be blown about by side winds.

 

[old man mavic]

@Yorkshire_Pud how can it fly back to its set home point ,without GPS ,I say set home point, because that position could be different from its original take off point, if it was altered by the pilot during flight ,when it looses GPS it has no way of knowing where it is in,relation to where the set home point is ,and as it will be, as you say, being blown along by the wind ,it is basically out of control, so having it land would appear to be the safest option
of course this is only becomes an issue if the drone is BVLOS, because it can be flown back to where you the pilot are ,if you are able to tell its orientation,even though there is no GPS lock
where does it state that upon loss of GPS ,the drone will just land ?

 

[slup]

If a drone started out with good GPS but subsequently loses it I am wondering why it has DJI have decided that the drone should land ?

It doesn't, just by losing the GPS... it will land when losing RC connection when it doesn't have GPS.

...I am wondering, if RTH was initiated 'as soon as' GPS was lost, why the drone can't fly on that heading for a while to see if things improve ?

"To see if thing's improve"... compared to what?

If the drone loses the GPS reception it loses all positional awareness, it doesn't have a clue where it is relative the HP & don't know how to manoeuvre/counteract outside forces, to get there/somewhere... the drone doesn't have a sixth sense that tells it that it's probably safe & OK to continue with a certain attitude... for a while.

 

[Yorkshire_Pud]

Ahh I'm mixing up Go app logs and the logs of the other apps, plus vague recollections of the logs contents and disconnection behaviour.
Homeever Go, Go4 and fly all have HOME.courseLockAngle columns but only Go populates it.

I will need to check in the morning whether or not that angle changes as a P3 moves 'sideways' but the thinking is, if the drone knows the angle for home it could fly on that course based on the compass.

But perhaps I am up a gum tree lol.

 

[StevenBrodsky]

If a drone started out with good GPS but subsequently loses it I am wondering why it has DJI have decided that the drone should land ?
I think the drone maintains a 'direction to home is' value but presume that is only whilst GPS is available. I am wondering, if RTH was initiated 'as soon as' GPS was lost, why the drone can't fly on that heading for a while to see if things improve ?

I AM NOT suggesting that this would be feasible if the entire flight had been without GPS nor am I suggesting it would be possible if the drone had been moved from the loss of GPS point.

I also realise that with out GPS the drone is likely to be blown about by side winds.

I didn't know that drones land if they lose their GPS location mid flight. I thought they just hover and move around in ATTI mode until GPS re-links.

 

[Meta4]

I didn't know that drones land if they lose their GPS location mid flight.

They don't.
They land if they have no GPS reception and RTH is initiated.

 

[StevenBrodsky]

They don't.
They land if they have no GPS reception and RTH is initiated.

So to clarify, Even if you still have a connection with the drone and can see the camera transmission, if you lose GPS and RTH is initiated it will land. That is only if RTH gets initiated, correct?

So if the RC and Drone don't disconnect RTH is unlikely to activate, thus the drone won't land and you can fly home in atti mode. Yes?

 

[Meta4]

So to clarify, Even if you still have a connection with the drone and can see the camera transmission, if you lose GPS and RTH is initiated it will land. That is only if RTH gets initiated, correct?

So if the RC and Drone don't disconnect RTH is unlikely to activate, thus the drone won't land and you can fly home in atti mode. Yes?

You can fly as long as your battery lasts without GPS.
If you lose signal without GPS, that would initiate RTH (if Failsafe Actions is on the default setting of RTH), and the drone would land.

 

[slup]

...the thinking is, if the drone knows the angle for home it could fly on that course based on the compass.

A compass isn't a geographical point locking device... it shows a direction to the target when measured, based on where you are when measuring. If you later stray away from that direction sideways... the direction will not change as it's not locked to a point & as you doesn't have access to GPS data you doesn't know that you've strayed away sideways... this will make you miss your target.

Try it yourself with a map & a ordinary compass... take out a direction to a known object & instead of following that you walk 100m to the side, then continue in the compass direction. You will miss your target... & furthermore you will not know when you need to stop walking as you don't have any measurement of the distance, neither to the object initially or how far you've walked.

DJI will never consider this to be safe...

 

[Yorkshire_Pud]

A compass isn't a geogrImaaphical point locking device... it shows a direction to the target when measured, based on where you are when measuring. If you later stray away from that direction sideways... the direction will not change as it's not locked to a point & as you doesn't have access to GPS data you doesn't know that you've strayed away sideways... this will make you miss your target.

I was aware of that which, to my mind, was included in "I also realise that with out GPS the drone is likely to be blown about by side winds." but thanks anyway.

Anyhow I have had a looks at some of my P3 logs and it appears that HOME.courseLockAngle does not record what I thought it to be so the whole idea is probably base less, whoops.

To quote Battery Sergeant Major Williams "Oh dear, how sad, never mind".

 

[Droning on and on...]

A compass isn't a geographical point locking device... it shows a direction to the target when measured, based on where you are when measuring. If you later stray away from that direction sideways... the direction will not change as it's not locked to a point & as you doesn't have access to GPS data you doesn't know that you've strayed away sideways... this will make you miss your target.

Try it yourself with a map & a ordinary compass... take out a direction to a known object & instead of following that you walk 100m to the side, then continue in the compass direction. You will miss your target... & furthermore you will not know when you need to stop walking as you don't have any measurement of the distance, neither to the object initially or how far you've walked.

DJI will never consider this to be safe...

In addition to only being able to hold a heading and not a course, more critically it has no idea how fast it is going, so can't indirectly estimate when it gets there.

 

[EssenYVR]

In addition to only being able to hold a heading and not a course, more critically it has no idea how fast it is going, so can't indirectly estimate when it gets there.

That's not actually quite true. The drone has acceleration and rotation sensors. There's no fundamental reason why it couldn't attempt to navigate back to the home point using dead reckoning based on its sensor inputs, in much the same way that INS systems used to guide aircraft across the Atlantic way before the days of GPS.

It wouldn't be nearly as accurate as GPS, of course, but it's better than just landing where ever it happens to be. And anything that gets the drone closer to the home point increases the odds that it will be able to re-establish connection to the controller.

Not something they bothered to code into the flight software, I guess.

 

[slup]

...Not something they bothered to code into the flight software, I guess.

No, of course... as DJI needs to consider safety instead of letting a drone fly on it's own without any accurate means of precise navigation & without a pilot supervising it & can intervene.

 

[EssenYVR]

No, of course... as DJI needs to consider safety instead of letting a drone fly on it's own without any accurate means of precise navigation & without a pilot supervising it & can intervene.

Sure, because landing in a random place is so much safer than trying to get back into radio range.

 

[slup]

Sure, because landing in a random place is so much safer than trying to get back into radio range.

Yep... in most cases.

 

[Yorkshire_Pud]

No, of course... as DJI needs to consider safety instead of letting a drone fly on it's own without any accurate means of precise navigation & without a pilot supervising it & can intervene.

I don't wish to start a fight but if DJI were solely governed by safety considerations they wouldn't permit way point flights that continue after disconnection or flights beyond the average persons' VLOS and to be honest I can well imagine that EU bureaucrats will eventually mandate a hard limit on range.

I don't see how a dead reckoning flight is any more risky that some of the current automated behaviours and we have seen some very risky automated flights in here, notably 'continued waypoint flights', at least one of which very narrowly avoided cables strung over a busy street and ended in a crash into a tree, all due to bad programming by the pilot.
That said, my original thinking was flawed and I am no longer know if the drone knows the 'bearing to home'. I don't know if there is anything relating this in those DAT's that are readable.

The drone has acceleration and rotation sensors.

In addition, I am pretty sure I have seen Sar104 say the drone knows its speed, from the angle of tilt, although, in the absence of GPS, I think that would be airspeed and not ground speed.

 

[Yorkshire_Pud]

One related question, somewhere I have seen it said that OA does not work in the absence of GPS.
I can quite well imagine that a drone can not maneuver around an obstacle with out GPS but can they brake with out GPS?

 

[Kilrah]

Only possibly if VPS was available...

 

[slup]

...if DJI were solely governed by safety considerations...

I don't see how a dead reckoning flight is any more risky that some of the current automated behaviours and we have seen some very risky automated flights in here.

Of course they aren't...

But it's a big difference between...

Having a reliable navigation system even without a connection & a pilot that can intervene (happens every time you failsafe for instance, which can initiate an automated RTH)... or putting the faith into mainly only height hold.

Dead Reckoning can't be compared with GPS positioning... it doesn't even come close.

A drone that is in ATTI & loses RC connection... but still continue to fly will likely never be seen again as it could go the rest of the battery in the wrong direction. If it lands instead, you will have the last picture of the surroundings on the screen device + some kind of awareness where on the route it was... & only a wind drift on the way down.

 

[StevenBrodsky]

One related question, somewhere I have seen it said that OA does not work in the absence of GPS.
I can quite well imagine that a drone can not maneuver around an obstacle with out GPS but can they brake with out GPS?

OA works without GPS. It doesn't need GPS. I can fly in my house and around the yard without GPS. It uses the obstacles in such a way, that it knows where it is relative to the obstacles. so it stays put or can move a little bit, if nothing is in its way. It will just hover if blocked by an obstcle.