[ros-users] gmapping errors

David Feil-Seifer david.feilseifer at gmail.com
Fri Jul 23 06:20:20 UTC 2010


Rishi-

William is right about the transform between base_link and base_scan.
If you look in the p2os_launch package, you will see a number of
launch files. In particular, you might be interested in the
tf_hokuyo_p2os_teleop_joy.launch or p2os_teleop_joy_tf_hokuyo.launch
which creates a hokuyo node, a p2os node, and joystick and teleop
nodes. This launch file should give you all the data that you need.
Bear in mind that you will have to adjust the static transform to
match your laser/pioneer configuration.

Good luck, and let me know if this does not fix things for you.

-Dave

On Thu, Jul 22, 2010 at 1:04 PM, William Woodall <wjwwood at gmail.com> wrote:
> I am seeing your tf tree as this: /map -> /odom -> /base_link.  Which is ok,
> but you do not have a laser link so the laser data is not being transformed.
>  The laser data's frame is laser, but you don't have a joint between the
> base_link and the laser link, this causes it to not be viewable in rviz or
> be used by gmapping.  You need to define the geometric relationship between
> the position of the laser sensor and the base of the robot (/base_link ->
> /laser).
> I would recommend reading this bit of the
> wiki: http://www.ros.org/wiki/tf/FAQ#Can_I_declare_static_transforms.3F
> Hope that helps,
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> William Woodall
> Graduate Software Engineering
> Auburn University
> w at auburn.edu
> wjwwood at gmail.com
> 256-345-9938
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>
>
> On Thu, Jul 22, 2010 at 10:34 AM, Rishi Bedi <rbedi100 at gmail.com> wrote:
>>
>> David - I'm running Ubuntu 10.04, ROS 1.1.13-12 (installed through
>> synaptic), revision 107 of p2os (modified 2010-07-13).
>> Full SVN info output for p2os:
>>
>> Path: .
>> URL:
>> https://usc-ros-pkg.svn.sourceforge.net/svnroot/usc-ros-pkg/trunk/p2os
>> Repository Root:
>> https://usc-ros-pkg.svn.sourceforge.net/svnroot/usc-ros-pkg
>> Repository UUID: 157d20ef-4c6e-4bb2-9484-52dd339cd29c
>> Revision: 107
>> Node Kind: directory
>> Schedule: normal
>> Last Changed Author: dfseifer
>> Last Changed Rev: 106
>> Last Changed Date: 2010-07-13 17:11:49 -0400 (Tue, 13 Jul 2010)
>>
>> Here is the bagfile that I just recorded:
>> www.rbedi.net/2010-07-22-11-29-38.bag
>>
>> William - I didn't set anything up for my transforms -- I just set ROS
>> to record the /tf topic -- is there any setting up I need to do?
>>
>> Thanks for all your help,
>> Rishi
>>
>>
>> On Thu, Jul 22, 2010 at 1:53 AM, David Feil-Seifer
>> <david.feilseifer at gmail.com> wrote:
>> > We at USC have been using P2OS concurrently with gmapping for months
>> > now, so I know that it CAN work. Let me help you debug this problem in
>> > more detail. Can you do me a favor and send me a link to that bag file
>> > that you recorded? Also, can you fill me in on what OS, ros release
>> > and version of the p2os stack (preferably the output of svn info from
>> > the p2os stack directory). That will help me figure this out in more
>> > detail.
>> >
>> > On Wed, Jul 21, 2010 at 10:02 PM, William Woodall <wjwwood at gmail.com>
>> > wrote:
>> >> I can't think of any reason why gmapping could not run concurrently
>> >> with the
>> >> p2os driver.  All that the gmapping slam node needs is published laser
>> >> scans
>> >> and correct TF data.  I have not used the p2os, but I would think it
>> >> gives
>> >> this information already, if the gmapping system isn't working
>> >> correctly
>> >> then I would look at the laser topic and the transforms to make sure
>> >> that
>> >> they are publishing the correct data.  Specifically make sure your
>> >> transforms are setup correctly.  Judging from your error:
>> >> "...TF_OLD_DATA
>> >> ignoring data from the past for frame /base_link at time..." I would
>> >> say
>> >> that there is something wrong with the transformations you have setup.
>> >> Hope that helps,
>> >> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>> >> William Woodall
>> >> Graduate Software Engineering
>> >> Auburn University
>> >> w at auburn.edu
>> >> wjwwood at gmail.com
>> >> 256-345-9938
>> >> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>> >>
>> >>
>> >> On Wed, Jul 21, 2010 at 10:41 PM, nitinDhiman <nitinkdhiman at gmail.com>
>> >> wrote:
>> >>>
>> >>> Hi Rishi,
>> >>> Then does that mean we cannot do concurrent mapping with currently ROS
>> >>> packages. For pioneer, I heavily relying on p2os drivers.
>> >>>
>> >>> Has any one tried concurrent localisation and mapping? May be using
>> >>> some
>> >>> other setup??
>> >>>
>> >>> Thank you,
>> >>> Nitin
>> >>> --
>> >>> View this message in context:
>> >>>
>> >>> http://ros-users.122217.n3.nabble.com/gmapping-errors-tp981480p986050.html
>> >>> 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
>> >>> https://lists.sourceforge.net/lists/listinfo/ros-users
>> >>> _______________________________________________
>> >>> 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
>> >>
>> >>
>> > _______________________________________________
>> > 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
>
>
> _______________________________________________
> ros-users mailing list
> ros-users at code.ros.org
> https://code.ros.org/mailman/listinfo/ros-users
>
>



More information about the ros-users mailing list