added documentation for symmetric_yaw_tolerance #827
Conversation
|
|
||
| Symmetric Yaw Tolerance for Goal Checking and Navigation | ||
| ******************************************************** | ||
| `PR #5795 <https://github.com/ros-navigation/navigation2/pull/5795>`_ introduces the symmetric yaw tolerance feature for goal checking and navigation, allowing symmetric robots to reach goals without unnecessary 180° rotations. |
There was a problem hiding this comment.
| `PR #5795 <https://github.com/ros-navigation/navigation2/pull/5795>`_ introduces the symmetric yaw tolerance feature for goal checking and navigation, allowing symmetric robots to reach goals without unnecessary 180° rotations. | |
| `PR #5795 <https://github.com/ros-navigation/navigation2/pull/5795>`_ introduces the symmetric yaw tolerance feature for goal checking and navigation, allowing symmetric robots to reach goals without unnecessary 180° rotations. |
configuration/index.rst
Outdated
| Advanced Topics | ||
| *************** | ||
|
|
||
| .. toctree:: | ||
| :maxdepth: 1 | ||
|
|
||
| packages/symmetric_yaw_tolerance.rst |
There was a problem hiding this comment.
I think instead, we should add this to the tuning guide as a subpage under a header for Symmetric Navigation https://docs.nav2.org/tuning/index.html
I'd make sure to mention loudly and prominently this ROSCon talk on the subject: https://vimeo.com/879000809. I don't think most users for bidirectional need the implementation you've suggested. His work makes Nav2 fully bidirectional in terms of motion in the planners / controllers for motion -- but does care about the final orientation. Yours is adding not caring about the goal completion orientation, which is a special subcase of that
There was a problem hiding this comment.
Moved section to tuning guide as suggested. Thank you for link to the video, I watched it.
I am using "Theta Star" planner, I found that by using this planner and MPPI controller with "PreferForwardCritic"
disabled and "PathAngleCritic" forward_preference = false, the robot moves in a very optimal way , it chooses
backwards or forwards depending on the current orientation and desired direction of travel, it was reorienting at the goal position , which this change fixes. In fact I am not sure why the path inversion (as described in the video) is required if the robot is truly bidirectional. Maybe I am missing something
addressed comments on PR
2 participants
Basic Info
Description of contribution in a few bullet points
Updated documentation for PR #5795 Symmetric Yaw Tolerance for
Symmetric robots which can arrive at goal in either orientation