Integrating a Low-Cost Stereoscopic Vision System Into a Six Degree of Freedom Robotic Arm for Engineering Education

Date of Award


Degree Name

M.S. in Mechanical Engineering


Department of Mechanical and Aerospace Engineering and Renewable and Clean Energy


Advisor: Timothy Reissman


In order to implement robotic systems for engineering education, an optimumcombination of low cost, open source modules and actual industrial systems should beused to give students hands-on experiences with the tools they may use in a future career.The purpose of this work is to expand the options within the former by integrating anopen source 3D camera vision system at an added cost of {dollar}200 into an already low-cost,{dollar}2,400, open source robotics platform. Using the Annin Robotics AR3 robotic arm, andcode first developed by Patrik Reizinger, a stereoscopic vision module was integrated andtested. The overall absolute average error of the camera system was found to be 0.41inches in the X direction, 0.44 inches in the Y direction, and 1.23 inches in the Zdirection. Using a 2.5 kg pull force magnetic gripper, the desired tolerance of +/- 0.5inches in the X and Y direction, and +/- 0.25 inches in the Z, it was found that the X andY directions are within the tolerance, while the Z requires some improvement. While theerror may be beyond our present magnet's tolerance, it is believed that with a higherpowered electromagnet, this system is suitable for use in teaching vision assisted pick andplace robotic operations. It is suggested for future improvement to focus on improvingthe accuracy of the vision system. The combined low-cost system fills a gap in costefficient engineering educational systems openly available on the current market.



Rights Statement

Copyright 2020, author