JointLab: A VR Framework for Early Robotics Education in Architectural Fabrication

This thesis presents JointLab, an immersive educational platform designed to support early-stage learning in robotics and architectural fabrication. Aimed at students and novices, JointLab leverages Unity, ABB RobotStudio, and the Meta Quest 3’s controller to provide real-time, bi-directional communication with an ABB IRB 120 robot. Set within a digital twin of Carnegie Mellon University’s dFab Lab, JointLab offers an accessible, low-risk environment for learning complex robotics concepts through simulation. Users can operate robotic joints, plan toolpaths, and monitor diagnostic data using intuitive interface panels, including modules for joint control, spatial navigation, live connection via Robot Web Services (RWS), and low-latency control using Externally Guided Motion (EGM). By removing traditional programming obstacles, JointLab sets opportunities for learners without any experience in coding or robotics to engage in spatial understanding and robotic workflow development. A structured user study evaluates the effectiveness of the system in understanding joint manipulation, motion planning intuition, and enabling user confidence with real-world robotics. Many participants with little to no prior experience developed a clearer understanding of spatial understanding and robotic movement. Feedback reinforced JointLab’s value as an engaging educational entry point into robotics, demystifying complex concepts and made robotic systems feel more approachable. This work contributes to computational design, design fabrication, and robotics education by demonstrating how immersive VR environments, paired with simplified layouts and controller interactions, can help make robotics more accessible. JointLab offers a scalable model for bridging experimentation and real-world implementation. A fully documented implementation is publicly available on GitHub to support its further development.