[ros-users] Navigation stack not working

Eitan Marder-Eppstein eitan at willowgarage.com
Mon Sep 20 17:49:40 UTC 2010


Sorry for joining the discussion a bit late.

A couple of things:

1) The voxel grid tracks unknown space, but its not published in the
visualization of the grid. It was just too many cells to render. If you want
to see the 2D projection of unknown space in the voxel grid, you can
subscribe to the unknown_space topic of the relevant costmap.

2) Publishing visualization information from very large maps can slow the
system down. As Eric suggested, setting publish_frequency to 0 for the
global_costmap should help things.

3) Changing the update_frequency from 5.0Hz to 8.0Hz actually makes the map
update more frequently so I'm not sure why this resolved anything for you.

4) For large maps, on some machines, I can see it taking some time for AMCL
to start up. Do the errors your experiencing go away consistently after AMCL
is successfully up and running?

Hope all is well,


On Mon, Sep 20, 2010 at 4:07 AM, Prasad Dixit <abhimohpra at gmail.com> wrote:

> Hello Eric,
> I have changed the update rate from 5.0 to 8.0.
> Though It is taking lot of time initially to load the map and configuring
> likehood model; it worked finally..
> Is it something because i have not specified position in AMCL? Is this
> taking lot of time to spread initial particles and get exact global
> position
> of robot in the map?
> OR is it completely related to hardware which i should update it
> finally...??
> - Prasad
> --
> View this message in context:
> http://ros-users.122217.n3.nabble.com/Navigation-stack-not-working-tp1519053p1528384.html
> Sent from the ROS-Users mailing list archive at Nabble.com.
> _______________________________________________
> ros-users mailing list
> ros-users at code.ros.org
> https://code.ros.org/mailman/listinfo/ros-users
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ros.org/pipermail/ros-users/attachments/20100920/0b068be9/attachment-0003.html>

More information about the ros-users mailing list