Question to a developer, why not make it so that when a GPS signal is lost in RTL mode transition control mode stabilize?
Because of the loss of GPS signal recently flew a quadrocopter. After enabling RTL for the lost GPS copter gained altitude of 200 m and left in an unknown direction.
I've started a developer discussion to check for a couple radio failsafe and RTL things in the code.
I did review the code and you are right, we check only if we have a home position when switching to RTL.
This mean that if there was a GPS fix at take off (home position was stored) and later the GPS fix is lost, you can have a flyaway when asking for RTL.
This will be fixed in version 3.0 with a GPS failsafe feature.
Thanks for pointing this problem and please report if you still have this problem when version 3.0 will be out.
You guys are amazing. Working with all that code and still be able to find little flaws and fix them instantly.
REALLY AMAZING !!!
Thanks for all your work !!!
When I remember back my first Arducopter ... it barely flew, was a pain to config and now. Near perfection...
I´ve ask for it in the past. Is it possible to implement a logical check in RTL?
when switching to RTL
if current_Position-> go further afar Home_Position-> loiter or land
Thanks Cybercrash. This is an effect of opensource and collaborative work. The world is changing.
@Meiser : I don't understand your last sentence.
Actually it does land if home distance is < waypoint radius, or RTL if home distance > waypoint radius.
And there is the new Geofence as well in version 3, with it you can define a max altitude and max radius distance. When breaking over those limits, RTL is trigerred.
I tried to explain a possible RTL verification.
When RTL is triggered the copter should move to it´s home position.So the distance from "RTL-trigger event" position to home should become smaller. If this is not the case the apm could stop moving, loiter or land. That would prevent "fly aways" on RTL.
Thank you all for your responses and rapid response to emerging issues.
This is a good idea, switching to Land if RTL is not sucessfull.
I will discusse that with the dev group.
Keep in mind that we have something similar actually : battery failsafe. If battery failsafe is triggered, then we Land immediately.
But would be better to Land as soon as possible if RTL do not work because of strong wind for example.
Landing when gps lock is not available?
I think that is a very bad idea. What if the copter is just hovering over a tree, river, car, or street and suddenly it's going to land!
I wouldn't want that.
Better would be just to try to hold position anyway possible (stabilize) and Alt-hold.
This will give me enough time to fly my copter manually.
Holding position is not possible more than a couple seconds without GPS.
I think it's better to Land. At least it does prevent a high energy crash that you could have trying to keep a hazardous position or altitude.
Remember that those actions are last chance recovery attempts. In this regards they should be as safe as possible for ground people.
To reach this level of failsafe, you need a lost radio link + lost GPS. Obviously in those terrible conditions, the best is to Land.
If there is a GPS fix available and enough battery, then RTL will do the recovery job.
Okay, lost radio link + lost gps...then landing is the most wise thing to do.
And what will the copter do, when I let him fly and I loose gps?
or radio link?
can I decide what the copter should do?
"Kraut" Rob came up with this idea long ago .RTL, GeoFence, Failsafe absolute great, if RTL leads the copter back and not further afar ;-D