## playerstage-gazebo

 [PlayerStage-Gazebo] Modeling Robocup league in gazebo From: Kamil Muszyński - 2008-07-09 20:49:47 ```Hello, First of all - I'm new to Gazebo:) and I'm trying to make a model of Robocup robot (small-size league), something similar to http://www.youtube.com/watch?v=APUhsu9eUEM This is an example of a game: http://www.youtube.com/watch?v=1JJsBFiXGl0 Robots that I'm trying to create are similar (except that they have to use differential drive). My scheme: http://www.nabble.com/file/p18370471/schemat.jpg To make it clear :) - image on the left - top view of a robot (without wheels on the left and right side) On the right: - blue one - dribbler device - rotating cylinder used to give a ball (orange one) rotation so it will stick to the robot - red box - device used to kick the ball My questions/problems: 1. Dribbler bar is a cylinder, which has to rotate at some speed. I've connected it with the chassis (black shape on the left) with hinge joint : dribbler chassis dribbler 0 1 0 so there is only one connection point at the center of the dribbler cylinder. When dribbler is rotating and robot is doing a turn, dribbler 'needs some time' to follow that turn (rotating round his center along the z axis). Can I connect dribbler to the chassis in a diffrent way (so there will be 2 connection points - one at each end of the dribber) to make it more stable? 2. Ball, once it's kicked, doesn't stop - even though friction coeffs for field and ball are set to be more than 1.0. 3. Kicking device - I've created it using a box attached to a chassis with a slider joint. To kick, i've modified a Differential_Position2d controller so it can apply force to the slider joint. My problem is that to kick a ball with satisfying speed the force should be quite large, so the box (kicking device) can move very fast for a short distance. Because of the step time and the great force used to move a joint box misses (or seemes to penetrate) the ball (so O| turns into |O :) ). I know that is because of the step time - problem disappears if I make it smaller (currently I'm uising 0.001) and the collision beetwen box and ball is detected correctly, but that makes simulation extremly slow. Is there any way to avoid this problem? Thanks for advice (and sorry for the images:) ) Kamil -- View this message in context: http://www.nabble.com/Modeling-Robocup-league-in-gazebo-tp18370471p18370471.html Sent from the playerstage-gazebo mailing list archive at Nabble.com. ```
 Re: [PlayerStage-Gazebo] Modeling Robocup league in gazebo From: Nate Koenig - 2008-07-10 16:21:00 ```Hello, Comments below... On Wed, Jul 9, 2008 at 1:49 PM, Kamil Muszyński wrote: > My questions/problems: > 1. Dribbler bar is a cylinder, which has to rotate at some speed. I've > connected it with the chassis (black shape on the left) with hinge joint : > > dribbler > chassis > dribbler > 0 1 0 > > so there is only one connection point at the center of the dribbler > cylinder. When dribbler is rotating and robot is doing a turn, dribbler > 'needs some time' to follow that turn (rotating round his center along the z > axis). Can I connect dribbler to the chassis in a diffrent way (so there > will be 2 connection points - one at each end of the dribber) to make it > more stable? You can use two hinge joints, one at each end of the cylinder. There is another tag for a joint called , that will allow you place the joint anchor at arbitrary locations. > 2. Ball, once it's kicked, doesn't stop - even though friction coeffs for > field and ball are set to be more than 1.0. Try setting the friction coeffs higher. > 3. Kicking device - I've created it using a box attached to a chassis with a > slider joint. To kick, i've modified a Differential_Position2d controller so > it can apply force to the slider joint. My problem is that to kick a ball > with satisfying speed the force should be quite large, so the box (kicking > device) can move very fast for a short distance. Because of the step time > and the great force used to move a joint box misses (or seemes to penetrate) > the ball (so O| turns into |O :) ). I know that is because of the step > time - problem disappears if I make it smaller (currently I'm uising 0.001) > and the collision beetwen box and ball is detected correctly, but that makes > simulation extremly slow. Is there any way to avoid this problem? You could fake the kick, but writing a custom controller for your robot. When the controller receives the "kick" signal, the robot will apply a force directly to the ball. This will require your robot to know if a ball is in the kicking device. > Thanks for advice (and sorry for the images:) ) > Kamil > > -- > View this message in context: http://www.nabble.com/Modeling-Robocup-league-in-gazebo-tp18370471p18370471.html > Sent from the playerstage-gazebo mailing list archive at Nabble.com. > > > ------------------------------------------------------------------------- > Sponsored by: SourceForge.net Community Choice Awards: VOTE NOW! > Studies have shown that voting for your favorite open source project, > along with a healthy diet, reduces your potential for chronic lameness > and boredom. Vote Now at http://www.sourceforge.net/community/cca08 > _______________________________________________ > Playerstage-gazebo mailing list > Playerstage-gazebo@... > https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo > ```