This year at UTS TechFest's Robotics Showcase, my group (Nikhil Kumar, Edan Anonuevo) and I were able to present our project we've been working on for our Robotics Studio 2 Class this semester: Pablo - The Selfie Drawing Robot!
Put simply, our project allows the user to take a selfie using the GUI and watch as our robot arm traces the outline of their face on card with a marker, and my team and I are very proud to win the second place highly commended robotics project award for the showcase!
We chose to write this project primarily in C++ for its superior processing speed, with a java based web front-end/GUI and ROS2 communication protocols to control the Universal Robots UR3e arm. We utilised a number of libraries such as RemBG, OpenCV and DLib for facial recognition and outlining, as well as several TSP/optimisation algorithms we devised to simplify the drawing process.
I was personally in charge of creating the path planning and optimisation algorithms, to maximise the image fidelity whilst drawing the image in the shortest time possible. I also created the custom 3D printed end effector, incorporating a spring-dampened pen holder as well as the canvas mount.
This project has been such a valuable learning experience for myself, and a highlight of my course so far! I'm so thankful for my team members who worked so hard to get this project to a polished state by the showcase date, seeing people lining up to watch the robot draw their face was so rewarding. I'd also like to thank Tony Le and Tan Huynh for the guidance and feedback on how to make this project ready for TechFest.