[ros-users] Problems in running navigation stack

hitesh dhiman hitesh.dhiman.1988 at gmail.com
Fri Apr 16 05:29:09 UTC 2010


Hi Eitan,
Thanks for the help, I'm finally getting velocity commands out. Relating to
the application of the navigation stack for my project, i have some doubts.
The navigation stack requires a goal in order to move. My question is, if I
want to traverse an unknown environment that has obstacles: an office
corridor, for example, i can pass a goal with 1m forward..but if there is a
turn in the path, the robot will not be able to turn accordingly since the
goal represents a position in the global world.
Therefore, using the simple navigation goal will not suffice in the current
scenario. Is it possible to give the robot goal commands in the local
system?

On Thu, Apr 15, 2010 at 1:06 AM, Eitan Marder-Eppstein <
eitan at willowgarage.com> wrote:

> Hitesh,
>
> First off, it seems like something might be weird with your laser scans.
> You need to make sure of two things:
>
> 1) That the header in the laser messages sent out contain the correct
> frame_id and timestamp for tf to be able to use them.
> 2) That the transform betweeen odom and [your_laser_frame_name] exists in
> your tf tree. You can "rosrun tf view_frames" to generate a pdf of your tf
> tree which you can check.
>
> To verify that things are working, try to view laser scans in rviz with
> both fixed and target frame set to odom. If you see them show up... things
> are good.
>
> Next, to use the navigation stack in the odometric frame, you'll have to
> change some of the parameters in the global_costmap_params.yaml file to tell
> it that you want it to operate in an odometric frame rather than using a
> static map. Something like the following configuration should work:
>
> global_costmap:
>   global_frame: odom
>   robot_base_frame: base_link
>   update_frequency: 5.0
>   publish_frequency: 0.0
>   static_map: false
>   rolling_window: true
>
>
>   width: 20.0
>   height: 20.0
>   resolution: 0.05
>
> This will give you a 20m x 20m around the robot in which you can send goals
> in the odometric frame.
>
> Hope this helps a bit,
>
> Eitan
>
>
> On Wed, Apr 14, 2010 at 6:48 AM, hitesh dhiman <
> hitesh.dhiman.1988 at gmail.com> wrote:
>
>> Hi,
>> actually i did use the tutorial to configure the robot. My setup is almost
>> the same as the files listed in the tutorial, except that the costmap
>> configuration is set to a differential robot.
>> From what i could see using rxgraph, i'm getting the following messages
>> into move_base:
>>
>> 1) Odometry
>> 2) Odometry tf
>> 3) laser tf (base_link to base_laser)
>> 4) laserscan
>>
>> I'm not using a static map, so I removd amcl from move_base.launch.
>> When i start the move_base.launch file, i get the following:
>>
>> [ INFO] 1271252548.078700000: Subscribed to Topics: laser_scan_sensor
>> [ INFO] 1271252548.178918000: MAP SIZE: 199, 199
>> [ INFO] 1271252548.185354000: Subscribed to Topics: laser_scan_sensor
>> [ WARN] 1271252563.370080000: MessageNotifier [topic=scan,
>> target=/base_laser /base_laser ]: Dropped 100.00% of messages so far. Please
>> turn the [ros.costmap_2d.message_notifier] rosconsole logger to DEBUG for
>> more information.
>>
>> Using rxloggerlevel to debug, here's the output:
>>
>> [DEBUG] 1271251959.320831000: MessageNotifier [topic=scan, target=/odom
>> /base_laser ]: Added message in frame laser at time 1271251959.264, count
>> now 50
>> [DEBUG] 1271251959.320973000: MessageNotifier [topic=scan, target=/odom
>> /base_laser ]: Removed oldest message because buffer is full, count now 50
>> (frame_id=laser, stamp=1271251958.043920)
>>
>>  [DEBUG] 1271251959.599588000: MessageNotifier [topic=scan,
>> target=/base_laser /base_laser ]: Successful Transforms: 0, Failed
>> Transforms: 4051196, Discarded due to age: 0, Transform messages received:
>> 67611, Messages received: 26918, Total dropped: 26868
>>
>> Also, to view the laser scan image in rviz, i get the following error:
>>
>> TOPIC : No messages received
>> Transform[sender=/laser] : For frame [laser]: Frame [laser] does not exist
>>
>>
>> Any suggestions?
>>
>> On Wed, Apr 14, 2010 at 1:01 AM, Eitan Marder-Eppstein <
>> eitan at willowgarage.com> wrote:
>>
>>> Hitesh,
>>>
>>> If you're trying to run the navigation stack on your own robot, I'd
>>> suggest reading the following tutorial first,
>>> http://www.ros.org/wiki/navigation/Tutorials/RobotSetup/, as it will
>>> explain how the pieces of the navigation stack work and how you'll need to
>>> configure your robot.
>>>
>>> From the errors you listed, it looks like your robot may not be
>>> publishing odometry information to the navigation stack. Also, amcl will not
>>> work without a static map... so that's expected.
>>>
>>> Hopefully, the tutorial will help, feel free to post again if you have
>>> other questions,
>>>
>>> Eitan
>>>
>>> On Tue, Apr 13, 2010 at 5:49 AM, hitesh dhiman <
>>> hitesh.dhiman.1988 at gmail.com> wrote:
>>>
>>>> Hi all,
>>>> I've been trying to run the navigation stack. However, upon running
>>>> move_base.launch, i get the following message:
>>>>
>>>> [ WARN] 1271162502.543875000: Costmap2DROS transform timeout. Current
>>>> time: 1271162502.5438, global_pose stamp: 1271162501.8677, tolerance: 0.3000
>>>>
>>>> I tried changing the tolerance value to 1. move_base.launch starts with
>>>> the following:
>>>> [ INFO] 1271162733.968657000: Subscribed to Topics: laser_scan_sensor
>>>> [ INFO] 1271162734.066727000: MAP SIZE: 199, 199
>>>>
>>>> A second later, i start getting these messages:
>>>>
>>>> [ WARN] 1271162749.245438000: MessageNotifier [topic=scan, target=/odom
>>>> /base_laser ]: Dropped 100.00% of messages so far. Please turn the
>>>> [ros.costmap_2d.message_notifier] rosconsole logger to DEBUG for more
>>>> information.
>>>>
>>>> rxconsole gives the following:
>>>> Request for map failed, trying again...  Warn /amcl
>>>> This could be because i'm not using any static map. Even so, move_base
>>>> is not publishing any messages out.
>>>>
>>>> Any help would be appreciated. I'm not able to figure out where the
>>>> problem is.
>>>>
>>>> --
>>>> Regards,
>>>> Hitesh Dhiman
>>>> Electrical Engineering
>>>> National University of Singapore
>>>>
>>>>
>>>> ------------------------------------------------------------------------------
>>>> Download Intel® Parallel Studio Eval
>>>> Try the new software tools for yourself. Speed compiling, find bugs
>>>> proactively, and fine-tune applications for parallel performance.
>>>> See why Intel Parallel Studio got high marks during beta.
>>>> http://p.sf.net/sfu/intel-sw-dev
>>>> _______________________________________________
>>>> 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
>>>
>>>
>>
>>
>> --
>> Regards,
>> Hitesh Dhiman
>> Electrical Engineering
>> National University of Singapore
>>
>> _______________________________________________
>> 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
>
>


-- 
Regards,
Hitesh Dhiman
Electrical Engineering
National University of Singapore
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ros.org/pipermail/ros-users/attachments/20100416/cef615e1/attachment-0003.html>


More information about the ros-users mailing list