Re: [ros-users] human approach

Top Page
Attachments:
Message as email
+ (text/plain)
+ (text/html)
Delete this message
Reply to this message
Author: User discussions
Date:  
To: ros-users
Subject: Re: [ros-users] human approach

Thanks Eitan, this helps a lot.
Yes, I do care about obstacles.
Will I have to preserve an additional costmap for my
TrajectoryPlannerRos object or should I read the costmaps from move_base?
Nikos

On 01/11/2011 06:31 PM, Eitan Marder-Eppstein wrote:
> Nikos,
>
> If you don't care whether or not the robot avoids obstacles, you can
> just send velocity commands directly to the robot. If you do care
> about avoiding obstacles, and you want control of what velocity
> commands get sent to the base, you can instantiate a
> TrajetoryPlannerROS object within your code and use the
> "checkTrajectory" function
> (http://www.ros.org/doc/api/base_local_planner/html/classbase__local__planner_1_1TrajectoryPlannerROS.html#a9155bebb3d073ae4b19b8d0da87478a6)
> to decide whether or not that velocity command is legal. Then, you can
> either send it on to the base, or decide to stop.
>
> Hope this helps,
>
> Eitan
>
> On Tue, Jan 11, 2011 at 4:32 AM, Nikos Mitsou <
> <mailto:nmitsou@lsr.ei.tum.de>> wrote:
>
>     Hello everyone,
>     My mobile robot needs to interact with humans...
>     In order to do that it has to approach a person in a user-friendly
>     way.
>     This means that the robot should follow a very specific trajectory
>     that
>     considers social aspects,
>     for example: the robot should approach a person face to face, the
>     trajectory should be smooth with specific velocities etc.

>
>     We can calculate the desire trajectory for free space but we don't
>     know
>     how to make the robot follow it.
>     One solution we are considering is to send multiple goals to move
>     base.

>
>     Another problem here is that a goal might happen to be inside or very
>     close to an obstacle and thus the low lever planner will probably
>     design
>     an undesirable trajectory.

>
>     Do you have any suggestion towards this problem?
>     Thanks in advance
>     Nikos

>
>     _______________________________________________
>     ros-users mailing list
>      <mailto:ros-users@code.ros.org>
>     https://code.ros.org/mailman/listinfo/ros-users

>
>
>
> _______________________________________________
> ros-users mailing list
>
> https://code.ros.org/mailman/listinfo/ros-users