Sawyer’s Travels was one of the final projects of Northwestern University’s ME 495: Embedded Systems in Robotics in Fall 2017. The goal of this project was to write a ROS program for Sawyer to navigate a ball through a labyrinth using vision feedback.
The ROS program written in this project included three major parts, computer vision, path planning, and robot control. At first, computer vision part of the program took care of extracting maze, ball tracking and final goal recognition. With all the vision information, then path planning program could solve for a global path plan to reach goal and dynamically solve for local path plans when the ball drifts off the global path plan. Finally, robot control program implemented PID controller on both joint velocity control and joint position control of the robot to manipulate the end effector to control the movement of the ball.
For more details, please view the README and source code of this project on Github.
Credit: Mazes were generated by Maze Generator.