Perception-Driven Manipulation Introduction


The purpose of these exercises is to implement a ROS node that drives a robot through a series of moves and actions in order to complete a pick and place task. In addition, they will serve as an example of how to integrate a variety of software capabilities (perception, controller drivers, I/O, inverse kinematics, path planning, collision avoidance, etc) into a ROS-based industrial application.


  • Understand the components and structure of a real or simulated robot application.
  • Learn how to command robot moves using Moveit!.
  • Learn how to move the arm to a joint or Cartesian position.
  • Leverage perception capabilities including AR tag recognition and PCL.
  • Plan collision-free paths for a pick and place task.
  • Control robot peripherals such as a gripper.