Rapidly-Exploring Random Tree Planner with UR5

Objective

The Rapidly-Exploring Random Tree (RRT) algorithm is a widely used motion planning technique in robotics. Its key advantage lies in its ability to rapidly explore the configuration space of the robot by randomly generating new nodes and connecting them to the existing tree structure. The algorithm proceeds by iteratively growing the tree toward the goal configuration, searching for a feasible path while avoiding collisions with obstacles. As part of my class at JHU, ‘Algorithms for Sensor-Based Robotics’, I wrote a goal-biased RRT planner in C++ and tested it in RVIZ and on the real-world UR5 arm.

RVIZ Simulation

Testing Using UR5

*Video coming soon!