-
Notifications
You must be signed in to change notification settings - Fork 3.6k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
ArduPilot: Loiter item support for 0/0 lat/lon #5272
Comments
Can you try with a newer Daily Build: https://docs.qgroundcontrol.com/en/releases/daily_builds.html |
@DonLakeFlyer exactly the same problem I'm afraid (just checked) |
Can you save the .plan file and attach here. Zip it to attach. A takeoff mission item for ArduPilot should have not GPS location associated with it. |
There is is (pasted and attached): |
This is because you are loading an MP format waypoint file while not connected to a vehicle. Because of that QGC does not know what firmware and/or vehicle type you are expecting to use this mission for. QGC needs to know that because for example the different firmwares/vehicles treat various mission command differently. If you were connected to a vehicle when you loaded this mission it would have worked correctly. When not connected you need to tell QGC what kind of vehicle you are using through the Vehicle Info section fo Mission Settings. This will then be remember si you shouldn't have to do it again. |
@DonLakeFlyer I'm not sure your explanation is related to the issue I reported (although I'm sure it's true on its own). Could you reopen?
So the problem is definitely there, isn't it? |
Are you using a daily build? |
Sorry I missed the fact that your screenshot has a vehicle on it. |
@DonLakeFlyer, you asked me to try a daily build, I did and still had the problem (comment #3). Do you mind re-opening this issue? |
After you connect the vehicle and it loads the mission. From the Plan view, can you tell me that the Vehicle Info section of Mission Settings says? |
No wait. Your loiter item has a lat/lon of 0/0 and an alt of 80. The screen shot above seems correct for that waypoint file. 2 0 3 17 0 0 1 0 0 0 80 1 |
Yes, there is 0,0 in there for sure, but it's supposed to be ignored for loiter since it loiters in place, whereever it happens to be. |
0, 0 does mean use the current location, ArduPilot/ardupilot@bf5005103c although I admit I would have assumed thats: I guess at the end of the day, we explicitly support it, but I'm left with a giant question as to why :) |
Thanks @WickedShell . First it confused me a bit, but then I thought about it: |
@DonLakeFlyer another weird one you may not need to support, but I'll mention it in passing. |
@marcmerlin The primary way loiters are used is loiter at this location :) The reason I think the current position is unreasonable is more that from a path planning perspective it's hard to say where exactly the aircraft will actually loiter until it flies the mission :) But thats just my opinion as an operator and I have no intention of arguing to remove it from the autopilot :) |
Total bizarreness. I don't plan to support that.
Yes, they do. They mean go to the specified location and loiter there. All loiter commands work this way and support a gps location. |
Confirmed, all NAV_LOITER* mission items will treat lat,lng (0,0) as current location. Otherwise, what would you do; fly to 0,0? No... Never do that. |
@DonLakeFlyer I was wrong about the loiter to GPS location thing, sorry about that. @magicrub if qgroundcontrol isn't going to support loiter in place, can you confirm that there is no other way to write a generic takeoff flight plan that will just work anywhere and that I should either learn to ignore all the flightplan displays from qgroundcontrol if I'm going to use it, or just use another GCS instead? |
If you do this:
TAKEOFF
RTL
As long as you don't have a DO_LAND_START then it'll loiter above home.
Obviously, the best solution is the TAKEOFF flight mode. Not sure if we can
get that in v3.8 but maybe we can do a quick v3.8.1 with it.
…On Sun, Jun 11, 2017 at 10:38 PM, Marc MERLIN ***@***.***> wrote:
@DonLakeFlyer <https://github.com/donlakeflyer> I was wrong about the
loiter to GPS location thing, sorry about that.
@magicrub <https://github.com/magicrub> if qgroundcontrol isn't going to
support loiter in place, can you confirm that there is no other way to
write a generic takeoff flight plan that will just work anywhere and that I
should either learn to ignore all the flightplan displays from
qgroundcontrol if I'm going to use it, or just use another GCS instead?
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#5272 (comment)>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/AEj7G5AEyTRzm2Y_JWjlllRwnNqH-6nrks5sDM64gaJpZM4N13gG>
.
|
Thanks @magicrub . Sadly that's not safe for some fields (the trees at baylands are very tall) and others like Rancho San Antonio will get you a ticket from a stupid ranger for overflying the area just behind you which is no fly zone. |
@marcmerlin why is planning a mission with an actual location for the loiter to burdensome? I don't understand how if you already have a GCS with you this is to difficult to do, which has the added bonus of being able to explicitly state where the aircraft will fly? |
@WickedShell I use the GCS for track recording and sound feedback of what's going on with the flight until I can get mavlink to go through my openlrsng system via the gitsly branch and directly on my taranis. |
There all seems like an uncommon use case which I may at some point try to support. But for now, changing all the algorithms to handle this 0/0 just doesn't justify the work. |
How does the PX4 autopilot handle waypoints and/or navigation points at 0/0? Does it navigate to it? |
No special handling that I am aware of. |
As explained in ArduPilot/ardupilot#2811, both mission planner and Qgroundcontrol treat a takeoff mission as if it were navigation to gps coordinates 0,0, drawing a long line to nowhere.
Apparently only mission planner gets this right
The text was updated successfully, but these errors were encountered: