Hi Advait,

I'm pretty new to ROS, so can you please give me a bit more guidance for editing the hrl_hokuyo to work with URGs?

As far as the servo is concerned, I'm still thinking between rewriting the package or buying the new servo. I'm slightly closer to the second one because the current servo is not really satisfying.

Thanks a lot,
Sanja
________________________________________
From: ros-users-bounces@code.ros.org [ros-users-bounces@code.ros.org] On Behalf Of Advait Jain [advait@cc.gatech.edu]
Sent: Friday, July 02, 2010 4:11 PM
To: ros-users@code.ros.org
Subject: Re: [ros-users] Tilting Hokuyo

Hi Sanja,

The hrl_tilting_hokuyo package combines the robotis and hrl_hokuyo
packages to make a tilting hokuyo.

Our specific hardware setup is described here:
http://www.hsi.gatech.edu/hrl-wiki/index.php/Tilting_Hokuyo

hrl_hokuyo is simply a python node that listens to the LaserScan message
from hokuyo_node. It currently does not support a URG but that shouldn't
be hard.

The servo would be the harder part. If you can write a ROS node for your
servo that presents the same interface as the robotis package then you
should be able to use hrl_tilting_hokuyo.

Advait


On Fri, Jul 2, 2010 at 2:28 PM, Vijay Pradeep <vpradeep@willowgarage.com> wrote:
> Hi Sanja,
>
> I haven't used hrl_tilting_hokuyo, so maybe some Georgia Tech folks can
> chime in about that.
>
> I would suggest using the laser_assembler package.  This will expose a
> service call that provides point clouds, given a time interval on which you
> want to assemble laser scans.
>
> Are you adding this to an existing robot running ROS, or are you starting
> from scratch?  If starting from scratch, there are a couple steps you'll
> need to take.  First, to interface with the Hokuyo hardware and publish
> laser scans, you can use hokuyo_node.  The laser_assembler also needs TF
> data for the frame of the laser.  The suggested way to do this is by writing
> a URDF description of your robot, and then use the robot_state_publisher to
> publish the transforms.
>
> You'll also need to write a hardware specific node to control the servo
> system.  This control node should publish sensor_msgs/JointState, since this
> is required by the robot_state_publisher to generate the transforms.  Since
> the controller would also know the current angle of the tilting platform, it
> can decide when each tilting cycle is complete and then make the appropriate
> service call to the laser_assembler to get a point cloud.
>
> Hopefully this makes sense.  Let me know if you run into trouble.
>
> Vijay
>
> On Fri, Jul 2, 2010 at 10:47 AM, Sanja Popovic <sanja@mit.edu> wrote:
>>
>> Hi,
>>
>> I have a Hokuyo URG-04LX and I would like to mount it to a tilt head to
>> get a point cloud. The tilt head I have is a ServoCity DDT500
>> (http://www.servocity.com/html/ddt500_direct_drive_tilt.html) and I also use
>> their servo (http://www.servocity.com/html/servo_controllers.html). Is it
>> possible to use those with ROS with some of the existing packages like
>> hrl_tilting_hokuyo? If not, do you suggest rewriting a package (which one?)
>> or just getting a new head + servo?
>>
>> Thanks,
>> Sanja Popovic
>> _______________________________________________
>> ros-users mailing list
>> ros-users@code.ros.org
>> https://code.ros.org/mailman/listinfo/ros-users
>
>
> _______________________________________________
> ros-users mailing list
> ros-users@code.ros.org
> https://code.ros.org/mailman/listinfo/ros-users
>
>
_______________________________________________
ros-users mailing list
ros-users@code.ros.org
https://code.ros.org/mailman/listinfo/ros-users