The manipulation pipeline has been released; THIS DEMO PAGE IS NOW DEPRECATED. For current information on the manipulation pipeline, and information about how to execute pick and place tasks using ROS and the PR2 see the pr2_tabletop_manipulation_apps stack page and the pr2_pick_and_place_demos package.
| Note: This tutorial assumes that you have completed the previous tutorials: Running the ICRA manipulation demo. |
 Please ask about problems and questions regarding this tutorial on answers.ros.org. Don't forget to include in your question the link to this page, the versions of your OS & ROS, and also add appropriate tags. |
ICRA manipulation demo sync point hierarchy
Description: A list of the sync points used to pause the demo when using the demo joystick controller.Tutorial Level:
Level 0
At level 0, the demo runs as if the sync mechanism did not exist:
- no pause (sync) points
- no changes of rviz topics
Level 1
At level 1, the demo will never pause, but rviz topics will be changed automatically during the demo
- no sync points
- rviz topics change automatically. Generally these are:
- almost always shown: collision models
- when detecting objects: narrow stereo
- when detecting objects and taking a static collision map: tilting laser, collision map
- when using the motion planner: motion planner goal (ghosted robot), collision map
- when planning grasp points on unknown clusters: grasp planner
- when attempting to grasp / place: interpolated IK for grasp points / place locations
Level 2
Pauses at the important points in the demo. Same rviz topic changes as level 1.
Sync points are placed as the robot is about to:
- detect table / objects acquire a static collision map
- attempt to grasp an object
call the cluster grasp planner (if grasping unknown object)
- call the motion planner to get to pre-grasp position
go from pre-grasp to grasp, close the gripper, attach the grasped object collision model to the gripper and lift the object (all-in-one). Note that this happens after a good grasp point has been found.
open the gripper and retreat (if grasp has failed)
call the motion planner to move the grasped object out of there (if grasp has succeeded)
- call the motion planner to get the arm to the object pre-place location
go from pre-place to place, open the gripper, detach the grasped object collision model from the gripper and retreat (all-in-one)
Level 3
Adds a few more sync points. Level 3 should generally not be needed.
Additional sync points for:
- each grasp points that is about to be tried
- go from pre-grasp to grasp
- close the gripper and attach the grasped object collision model to the gripper
- lift the object
