[ros-users] Error not reading scans on tf

Tully Foote tfoote at willowgarage.com
Wed Aug 11 21:47:03 UTC 2010


Renato,
Looking at your bag file.  The tf data and laser scans all look up to date
and are synced nicely.  However, the error above comes from the
observation_buffer.cpp in the costmap2d package.  Which I don't believe is
used in the linked launch file.  Is there something else running?  You can
get more info on ROS_ERROR messages from rxconsole.

As for the error message it suggests that data is being stored for more than
10 seconds.  And the default tf cache is 10 seconds so as soon as the data
is 10 seconds old it errors as above.

Tully

On Wed, Aug 11, 2010 at 11:38 AM, David Feil-Seifer <
david.feilseifer at gmail.com> wrote:

> Renato, I don't know about the ROS ARIA node for the pioneer, however,
> there is a ROS driver for the pioneer in the usc-ros-pkgs repository.
> The stack has tutorials for using the gmapping package with a pioneer
> and laser.
>
> http://www.ros.org/wiki/p2os/Tutorials/Getting%20Started%20with%20p2os
>
> On Wed, Aug 11, 2010 at 6:07 AM, Renato Samperio
> <mintaka_alnitak at hotmail.com> wrote:
> > Dear ROS users,
> >
> > I would like to know if somebody can help me with my problem. I want to
> map
> > with a Pioneer robot (with a ROSARIA node) and Sick laser (with a
> modified
> > version of sicktoolbox_wrapper). They are producing a tf tree as showed
> in:
> >
> > http://yfrog.com/f/jkrobotframesp/
> > *In case could be necessary I can share the launch file.
> >
> > The error I am obtaining is:
> >
> > "TF Exception that should never happen for sensor frame: , cloud frame:
> > /base_scan, You requested a transform that is 9.976 seconds in the past,
> > but the tf buffer only has a history of 9.971 seconds.
> >  When trying to transform between /base_scan and /map."
> >
> > I thought the problem was the tf was not publishing the laser scans but
> the
> > static scan transform seems to be faster than the C++ broadcaster. The
> last
> > output from tf_monitor is:
> >
> > RESULTS: for all Frames
> >
> > Frames:
> > Frame: /base_link published by /ROSARIA_client Average Delay: 0.00239482
> Max
> > Delay: 0.028034
> > Frame: /base_scan published by /sscrovers_tf_static Average Delay:
> > -0.0179362 Max Delay: 0.010168
> > Frame: /odom published by /Gmapping Average Delay: 0.00222625 Max Delay:
> > 0.026887
> >
> > All Broadcasters:
> > Node: /Gmapping 20.0193 Hz, Average Delay: 0.00222625 Max Delay: 0.026887
> > Node: /ROSARIA_client 10.0096 Hz, Average Delay: 0.00239482 Max Delay:
> > 0.028034
> > Node: /sscrovers_tf_static 45.3095 Hz, Average Delay: -0.0179362 Max
> Delay:
> > 0.010168
> >
> >
> > The problem is when I want to map as the scans are not representing the
> > robot surroundings. The same tf tree structure works with simulated data
> on
> > Stage for Gmapping but not in the robot. How can I assure Gmapping is
> > obtaining scans and tf tree correctly?
> > *In case could be useful, the bag I produced is here.
> >
> >
> > Many thanks.
> >
> > Renato.
> >
> > _______________________________________________
> > ros-users mailing list
> > ros-users at code.ros.org
> > https://code.ros.org/mailman/listinfo/ros-users
> >
> >
> _______________________________________________
> ros-users mailing list
> ros-users at code.ros.org
> https://code.ros.org/mailman/listinfo/ros-users
>



-- 
Tully Foote
Systems Engineer
Willow Garage, Inc.
tfoote at willowgarage.com
(650) 475-2827
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ros.org/pipermail/ros-users/attachments/20100811/1774992e/attachment-0004.html>


More information about the ros-users mailing list