AP_DDS: Local Velocity Publisher#23541
Conversation
7f7fb11 to
9492b85
Compare
Should we keep it that way ? Or do everything acording to NED ? This is easy to implement but the AP community needs to agree on how to do it |
As before, we want to follow ROS rep's as best as possible. In REP-147, it says, even for aerial vehicles, to follow REP-105. That said, it doesn't clarify on NED vs ENU. For now, let's stick with ENU, since that's largely a convention in the ROS community, and this is a ROS interface. I think it's totally reasonable, later, to add an alternative subset of topics in NED convention, if it's needed. |
|
REP-103 solves the dispute between ENU vs NED:
ENU should be default and we can add NED later, as you proposed. I will change the frame to ENU and commit it as a fixup |
9492b85 to
9a915a6
Compare
|
We spent some time testing, just deciding what coordinate system the orientation should follow on the vehicle. Likely, we will match what NAV2 is using, but I had issues bringing the examples up in NAV2. |
|
Please rebase this on master to fix conflicts with the pose topic. Then, I can test it. Thanks! |
9a915a6 to
d6ec51e
Compare
d6ec51e to
25fc39e
Compare
Resolves #23279

Here is the PR working in its current state in SITL:
It is currently following:
For linear velocity, the data is earth-fixed but starts alligned with the robot body (x-forward)
For angular velocity, the gyro data is body-fixed.