No costmaps being published #3034
Labels
question
Further information is requested
Comments
|
Note #3014, I think that's really the crux of this ticket where we are tracking this issue
Your comment about echoing is probably due to the QoS and since you can see the static layer in rviz means its being published. |
|
I read that bug report before posting mine. Wasn't sure if it was relevant to me or not. Key differences:
To clarify, I can view the Does it still sound like the same issue? I'm not at all familiar with the source code..
|
|
Oh neat, can you post your logs? Maybe something just didn't finish transition right. Post your parameters and the terminal output from launch Edit: I just tested the binaries and they seem to work for me, so I suspect it has to do with parameters or something failing to bringup properly. I highly doubt this issue is related to Fast-DDS, but its worth trying Cyclone real quick just given we're finding a few quirks here and there from the transition. |
|
Tyvm for your time/help btw. I was just trying Cyclone-DDS. It doesn't seem to make a difference. Will restart everything and try again just incase parts of the networks are still using Fast-DDS. Launch terminal outputFull Params fileLaunch file for good measure |
The only thing that comes to mind looking over that is maybe your static layer isn't getting the initial map? Topic is wrong or the QoS settings are wrong (e.g. Your launch file does something weird and has a second configuration file its setting for map server I'd look into that. What if you start from just the example |
|
I've tried with I've tried I believe these three lines in the launch log imply that the static layer is getting the map: The map file is 600x600 pixels at 0.05m/pix and I don't define that anywhere other than the map config so (unless I'm missing something) I'm pretty sure both global and local costmaps 'see' the original map when it's published on the You're right about that double assignment being a little weird, it was just a cheap quick way (lazy) thing I did when I merged the map_server config with the nav config from an earlier testing configuration. I'll tidy that up. I'll try the default launch file and report back. |
Agreed. I'm not actually referring to double assignment as I am that you have a separate |
So I've figured it out and it was, of course, the dumbest thing. Got the tb3_simulation_launch going (which produced costmaps fine) then copied the key files over to my package and started breaking it down.
Thank you so so much for your help, time and everything else you do for ROS. As tivial as it ended up being I wouldn't have worked it out for a long time without your advice. Bonus robot & cat tax: That's just my testbed, running on windscreen wiper motors and and old nissan leaf battery module. I'm welding up the chassis for my second bot as we speak! |
|
Had the exact same problem, thanks for the answer |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Bug? report
Required Info:
Ubuntu 22.04 Jammy
Humble
ros-humble-navigation2 1.1.0-2jammy.20220607.182532
Fast-RTPS (I believe, as it's the default)
Steps to reproduce issue
I'm running the nav2 stack pretty much as specified in the nav2_bringup package but without AMCL or obstruction layers as I currently have no LIDAR.
Expected behavior
I should be able to see (and debug) my costmaps.
Actual behavior
I can't see any costmaps in rviz2 and don't see anything when doing "ros2 topic echo" on the costmap topics.
Rudimentary navigation seems to work (robot will drive from A to B and will plan paths around objects in my static map) so the nav stack seems to have a costmap available to it. I'm not sure if they're being inflated properly though.
Additional information
Not sure if this is a bug or if I'm doing something wrong.
The launch file is almost identical to the nav2_bringup one (though I'm happy to provide my version if it would help).
These are the costmap parameters I'm using though I've tried changing almost all of them already.
The text was updated successfully, but these errors were encountered: