Skip to main content

Module 6.2: MoveIt 2 Basics for Motion Planning

Motion planning is a fundamental aspect of robotic manipulation, ensuring that a robot can move its limbs or end-effector from a starting configuration to a target configuration without colliding with itself, obstacles in the environment, or even the objects it intends to interact with. MoveIt 2 is the most widely used software suite for mobile manipulation in ROS 2, providing a comprehensive set of tools for motion planning, manipulation, and control.

What is MoveIt 2?​

Definition: MoveIt 2

MoveIt 2 is an open-source robotics manipulation platform for ROS 2. It integrates state-of-the-art motion planning algorithms, kinematics, collision checking, and control capabilities, allowing developers to easily enable complex manipulation tasks for a wide range of robotic arms and mobile manipulators.

Key Features of MoveIt 2:

  • Motion Planning: Implements various sampling-based (e.g., OMPL, RRT-Connect, PRM) and optimization-based motion planners.
  • Kinematics: Integrates with KDL, OpenRave, or custom kinematics solvers for Forward and Inverse Kinematics.
  • Collision Detection: Uses FCL (Flexible Collision Library) for efficient collision checking between robot links and environmental objects.
  • Robot State Monitoring: Keeps track of the robot's current joint states and transforms.
  • Manipulation Primitives: Provides tools for common manipulation tasks like grasping, placing, and collision avoidance.
  • Visualization: Seamlessly integrates with RViz for real-time visualization of robot motion and planned trajectories.

Core Concepts in MoveIt 2​

Planning Scene​

The Planning Scene is MoveIt 2's internal representation of the robot and its environment. It includes:

  • The robot's URDF/SRDF model.
  • The current joint states of the robot.
  • Attached collision objects (objects held by the robot).
  • Known static and dynamic obstacles in the environment.

The Planning Scene is constantly updated with sensor data (e.g., from depth cameras) to maintain an accurate representation of the real world.

Motion Planning Request​

To plan a motion, you typically specify:

  • Start State: The robot's current configuration.
  • Goal State: A desired end-effector pose or joint configuration.
  • Constraints: Optional requirements like keeping the end-effector level, avoiding certain regions, or maintaining specific joint limits.

MoveIt 2 then uses its planners to find a collision-free trajectory that satisfies these criteria.

Basic Workflow for Motion Planning (Python Example)​

Here's a conceptual Python workflow using moveit_commander (a Python interface for MoveIt 2) to plan a simple motion:

import rclpy
from moveit_commander import PlanningSceneInterface, MoveGroupCommander, RobotCommander

def main():
    rclpy.init()
    node = rclpy.create_node('moveit_simple_example')

    robot = RobotCommander(node=node)
    scene = PlanningSceneInterface(node=node)
    group = MoveGroupCommander("arm", node=node) # Replace "arm" with your robot's planning group

    # 1. Define a target pose for the end-effector
    pose_goal = geometry_msgs.msg.Pose()
    pose_goal.orientation.w = 1.0
    pose_goal.position.x = 0.4
    pose_goal.position.y = 0.1
    pose_goal.position.z = 0.4

    group.set_pose_target(pose_goal)

    # 2. Plan the motion
    plan = group.plan()

    if plan[0]: # Check if a plan was found
        node.get_logger().info("Motion plan found, executing...")
        group.execute(plan[1], wait=True)
    else:
        node.get_logger().warn("Could not plan motion to target pose.")

    group.stop()
    group.clear_pose_targets()

    rclpy.shutdown()

if __name__ == '__main__':
    main()
Installation

To use MoveIt 2, you typically install it as part of your ROS 2 distribution (e.g., sudo apt install ros-humble-moveit for humble) and then configure it for your specific robot using the MoveIt Setup Assistant.

MoveIt 2 is an indispensable tool for developing sophisticated robotic manipulation applications, providing a robust and flexible framework for motion planning and control.