[ros-users] navigation stack goal problem
eitan at willowgarage.com
Wed Jul 14 20:19:41 UTC 2010
The navigation stack expects the actual velocities on the odometry topic.
One issue could be that the acceleration limit parameters are set too high
for your robot which could lead to the planner thinking that the robot can
slow down and stop when it actually cannot. See the following page for
documentation on the relevant parameters:
If that suggestion doesn't help, could you post the configuration files that
you're using for the navigation stack... that'll help give insight into what
might be going on.
Also, when you say that the robot never stops, do you mean that it
oscillates while trying to achieve the goal orientation? Or that it actually
drives past the goal position, turns around, and attempts to achieve the
Hope this helps,
On Wed, Jul 14, 2010 at 1:12 PM, chrisb
<cburbridge+ros at gmail.com<cburbridge%2Bros at gmail.com>
> I am new to ROS and am trying to use it with a MetraLabs Scitos-G5 robot,
> which is a differential drive non-holonomic robot. In particular, I am
> trying to run the navigation stack.
> I have a player driver for the robot, which is a bit messy so I decided the
> quickest solution to get started would be to create ros nodes that connect
> to a player server running on the robot and republish the laser and sonar,
> and subscribe to cmd_vel and pass it on to player server. I made these
> nodes in a package, and they seem ok. I use a static transform to publish
> the transformation from laser to base.
> I eventually got it going enough to log data, and create a map with
> slam_gmapping. I then set up the navigation stack following the tutorial at
> http://www.ros.org/wiki/navigation/Tutorials/RobotSetup, and got rviz
> working to show the map, the laser, the odometry etc. The map is good, the
> localisation (amcl) works nicely and I can see a reasonable particle cloud
> using rviz.
> However, when I set a goal location the robot drives to the goal but never
> stops. It goes past it and tries to get back to it, never reaching the
> finished state. Is this the normal behaviour? Is there anything that I
> likely missed in my player nodes - my odometry node publishes the
> that player reports, could this be a problem? ie. Position2d::GetXSpeed(),
> GetYSpeed() and GetYawSpeed() I'm not fully certain what velocity is
> expected here - the commanded or the actual?
> Thanks in advanace!
> View this message in context:
> Sent from the ROS-Users mailing list archive at Nabble.com.
> This SF.net email is sponsored by Sprint
> What will you do first with EVO, the first 4G phone?
> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
> ros-users mailing list
> ros-users at lists.sourceforge.net
> ros-users mailing list
> ros-users at code.ros.org
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the ros-users