WPI’s Unified Robotics III course covers the basic principles surrounding conventional, serial robotic manipulators. For the final project for this course, I worked with a 4-Degree of Freedom RRRR OpenManipulator-X arm with a simple servo actuated gripper as the end-effector. 4 Dynamixel motors positioned at each joint allowed for both joint and task space control, enabling the use of both forward, inverse, and differential kinematics. Additionally, the use of a webcam, a black and white checkered board, and Matlab’s camera calibration and image processing tool boxes allowed for complex computer vision tasks including image acquisition, RGB masking, median filtering, centroid mapping, and object depth localization.

Photo of OpenManipulator-X arm

Sketch of arm used to derive the DH axis and forward kinematics

Sketch used to derive the arms inverse kinematics using the geometric approach

Using these tools we made the arm detect, pick, and sort colored balls on the checkerboard. In this project I learned to solve forward, inverse, and differential kinematics. I gained a deeper understanding of programming in Matlab, and learned how to use simple computer vision.