Filip Sund

It does indeed look like your changes fixed the issue with free vs. occupied voxels. Great work! The `max_range` parameter for the octomap does not seem to have any effect,...

If the intention is that Free Voxels should map out space that has been scanned and verified to contain nothing, I prefer the new implementation. In that case Free Voxels...

> That boils down to whether you consider a clipped beam to represent an obstacle too far away or a measurement error. That is a good point which I hadn't...

> Can you see the message on the `move_group/monitored_planning_scene` topic when you clear the scene? Yes, the following message appears on the `monitoried_planning_scene` topic when I press "Clear octomap" in...

**EDIT**: I think I understand what you meant now. An empty `id` field should indicate an empty octomap. So I just have to figure out where the octomap message is...

**EDIT**: I've now realized that removing that check will clear the octomap every time a diff message is sent, if the octomap field is not filled. ~~To me the check...

I have made a pull request with a preliminary test, to check how CI responds to my changes.

It seems the octomap message is created in [PlanningScene::getOctomapMsg](https://github.com/ros-planning/moveit/blob/master/moveit_core/planning_scene/src/planning_scene.cpp#L827). The issue I'm having is that the `id` is set in [fullMapToMsg](https://github.com/OctoMap/octomap_msgs/blob/melodic-devel/include/octomap_msgs/conversions.h#L188), so this is not available if `map` does not...

That is obvious to me now that I have some more experience with ROS, but it was not clear from the tutorial at the time. Perhaps a notice about this...

I have a "MX Master 3 for Business" that comes with a Logi Bolt adapter, and would be willing to help with implementing support for it. Any hints to where...