On Wed, Dec 8, 2010 at 2:24 PM, Laura
<laurago894@gmail.com> wrote:
Hello Eric,
Thank you for your answer.
1. I am actually using gmapping to make the map->odom transform, I have it
in a separate launch file I didn't include. The reason I used gmapping to
make this transform is that I couldn't make it work with amcl. Perhaps the
problem lays there. Could you tell me if this transform has to necessarily
be executed by amcl? My odom source is RosAria.
I'm not really familiar with the RosAria driver, but for AMCL to work you need to make sure that three things are happening:
1) The AMCL is getting the map correctly from the map_server
2) That AMCL is getting odometry information over tf
3) That AMCL is getting scans in from the laser you have on your robot and that those scans can be transformed to the odometric frame
I also wanted to know if the transform tree I am looking forward to is
actually useful for the execution of the navigation stack:
map->odom->base_link->base_scan.
This transform tree should work for the navigation stack. Though, we'd normally use the name "base_laser" instead of "base_scan" for the frame_id of the laser scans to differentiate them from the topic name.
I am a little bit confused because in amcl node's code I understood that the
attempted transform was odom->map and no map->odom.
I believe AMCL publishes the map frame as a parent of the odom frame.
2. I have already set /map as my fixed frame and the particle cloud isn't
displayed. In fact, the particle cloud never appears, even when I
teleoperate the robot with my keyboard.
This seems like something is not hooked up correctly with AMCL, I'd suggest checking the 3 requirements I listed above. Also, is AMCL printing any warnings? If you set the log level to debug, do you see anything from AMCL?
Hope this helps and that all is well,
Eitan
Thank you very much for your help.
Laura
--
View this message in context: http://ros-users.122217.n3.nabble.com/Particle-cloud-isn-t-displayed-on-rviz-tp2052811p2054220.html
Sent from the ROS-Users mailing list archive at Nabble.com.