<> '''EXPERIMENTAL''' <> == Overview == === Behaviors currently available === * `move_floor` - moves the base to a place on the floor * `table_approach` - approaches a table for tabletop manipulation * `grasp_object` - grasps object using [[pr2_grasping_behaviors]] overhead grasp * `place_object` - places object using [[pr2_grasping_behaviors]] overhead place === Running the Demo === To get this demo successfully working on your robot, you must remap most of the image and pointcloud topics so that they reference your preferred perception. The interface assumes your point_cloud and image generally overlap. The following lines of code run the demo on Georgia Tech's PR2: {{{ roslaunch pixel_2_3d run_srv.launch roslaunch gt_pr2 gt_map_nav.launch # launches both a global map (you must provide your own) and a navigation stack roslaunch hrl_pr2_lib openni_kinect_polled.launch # launches our Kinect nodes roslaunch hrl_clickable_behaviors clickable_world_demo.launch # The majority of the demo code is launched from here. }}}