Support odometry velocity in body and local frame

[kamilritz]

At the moment we expect the external vision's velocity to be expressed in world/local frame. This PR allows for the vision velocity also being expressed in body frame. For this a field is added to the vehicle_odometry message that stores the frame in which the velocity is expressed;

Related ECL PR: PX4/PX4-ECL#708
Solving issue: #13935

FYI @nicovanduijn @bresch @jkflying

[TSC21]

LGTM

[jkflying]

LGTM. NB: needs ECL merged first.

[codecov]

Codecov Report

Merging #13934 into master will decrease coverage by 0.06%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master   #13934      +/-   ##
==========================================
- Coverage   52.97%   52.91%   -0.07%     
==========================================
  Files         605      605              
  Lines       51349    51354       +5     
==========================================
- Hits        27202    27172      -30     
- Misses      24147    24182      +35     

Flag	Coverage Δ
#mavsdk	?
#python	3.15% <0.00%> (-0.01%)	⬇️
#sitl_mission_FW	29.38% <16.66%> (?)
#sitl_mission_MC_box	29.77% <16.66%> (-0.04%)	⬇️
#sitl_mission_MC_offboard_att	28.30% <16.66%> (+<0.01%)	⬆️
#sitl_mission_MC_offboard_pos	28.50% <16.66%> (-0.22%)	⬇️
#sitl_mission_Rover	26.22% <16.66%> (+<0.01%)	⬆️
#sitl_mission_VTOL_standard	33.94% <16.66%> (-0.03%)	⬇️
#sitl_mission_VTOL_tailsitter	33.94% <16.66%> (-0.03%)	⬇️
#sitl_python_FW	29.46% <16.66%> (-45.22%)	⬇️
#sitl_python_MC_box	29.84% <16.66%> (-0.04%)	⬇️
#sitl_python_MC_offboard_att	28.38% <16.66%> (?)
#sitl_python_MC_offboard_pos	28.57% <16.66%> (-0.22%)	⬇️
#sitl_python_Rover	26.30% <16.66%> (+<0.01%)	⬆️
#sitl_python_VTOL_standard	34.01% <16.66%> (-0.03%)	⬇️
#sitl_python_VTOL_tailsitter	34.01% <16.66%> (-0.03%)	⬇️
#unittest	34.81% <0.00%> (-0.01%)	⬇️

Impacted Files	Coverage Δ
src/modules/navigator/takeoff.cpp	3.92% <0.00%> (-76.48%)	⬇️
...flight_tasks/tasks/Failsafe/FlightTaskFailsafe.cpp	0.00% <0.00%> (-69.57%)	⬇️
src/modules/navigator/rtl.cpp	14.46% <0.00%> (-44.22%)	⬇️
...lib/flight_tasks/tasks/Manual/FlightTaskManual.cpp	0.00% <0.00%> (-38.89%)	⬇️
src/modules/navigator/loiter.cpp	8.16% <0.00%> (-38.78%)	⬇️
src/modules/mavlink/mavlink_timesync.cpp	72.50% <0.00%> (-15.00%)	⬇️
...der/Arming/PreFlightCheck/checks/airspeedCheck.cpp	59.25% <0.00%> (-14.82%)	⬇️
..._tasks/tasks/Utility/ManualVelocitySmoothingXY.hpp	78.57% <0.00%> (-14.29%)	⬇️
...t_tasks/tasks/Utility/ManualVelocitySmoothingZ.hpp	4.34% <0.00%> (-13.05%)	⬇️
src/lib/flight_tasks/FlightTasks.hpp	62.50% <0.00%> (-12.50%)	⬇️
... and 38 more

[kamilritz]

NOTE: Last testflights after rebase showed strange behavior. But could also be the vehicle itself. Please fly test this again before merging.

[TSC21]

@kamilritz are you able to rebase and take this further?

[kamilritz]

@TSC21 Rebasing is not a problem, but I am not able to test it myself.

[TSC21]

@kamilritz ok rebase should be enough for now. Can you do that? I will sync with @jkflying regarding how this can be tested.

[kamilritz]

@TSC21 Ok, I will start refactoring as soon as we have a solution for the testing.

[jkflying]

@TSC21 is there some way we could do ground-truth VIO in SITL?

I'm happy with this PR, but we knew it was working before and there is no test coverage, so I don't want to accidentally break something.

[TSC21]

@TSC21 is there some way we could do ground-truth VIO in SITL?

I'm happy with this PR, but we knew it was working before and there is no test coverage, so I don't want to accidentally break something.

@jkflying we seemed to have that before but it looks like it was removed - 6a4a4bd#diff-436758762aea1e43c9654faf66315930L42-L43.

[kamilritz]

I think there still is some basic testing: https://github.com/PX4/Firmware/blob/master/test/mavsdk_tests/configs/sitl.json#L18

[jkflying]

Considering the current issues I'm happy to merge this based on the SITL tests. I don't think it's realistic to get better testing than that for now.

[kamilritz]

I just remember that the last test I did on a real platform, did not went well. And there was not enough time left to investigate why. Based on that I am a bit hesitant to get this in without further testing a real platform.

[jkflying]

We found that this was due to a problem on the vehicle, so that doesn't give any information to whether this PR works or not 🙂

[kamilritz]

Nice. I will rebase it then.

[kamilritz]

Rebased the PR and reopened it here: #14703