/rrt-motion-planning

Rapidly-Exploring Random Tree (RRT) algorithm for sampling-based motion planning to control a UR5 robotic arm.

Primary LanguagePythonMIT LicenseMIT

RRT Motion Planning for UR5 Robotic Arm

Objective

This project focused on implementing the Rapidly-Exploring Random Tree (RRT) algorithm for sampling-based motion planning to control a UR5 robotic arm. Utilizing the MoveIt! library in a ROS2 environment, the goal was to develop a robust motion planning solution capable of navigating the robotic arm through various environments, avoiding obstacles, and reaching specified end-effector poses.

Video 1 Video 2
YouTube Video 1 YouTube Video 2

Key Features

  1. RRT Algorithm Implementation:

    • RRT algorithm to compute joint trajectories, ensuring the arm's end-effector could reach desired positions while avoiding obstacles.

  2. Collision Avoidance:

    • Collision detection and avoidance using data from an /obstacles topic, ensuring all planned paths were free from collisions.

  3. Inverse Kinematics (IK):

    • IK functions to convert end-effector goals into joint space configurations, essential for effective motion planning.

  4. Path Optimization:

    • Path shortcutting and resampling techniques to refine the planned trajectories, improving execution efficiency and ensuring smooth movement.

Project Structure

  • rrt_motion_planning.py: Main script for implementing RRT-based motion planning for the UR5 robotic arm.