RSS Robot
In the spring semester of my freshman year, I enrolled in Robotics Science and Systems (6.4200). The class provided a strong introduction to algorithmic foundations for robots covering topics including kinematics, computer vision, perception and controls using ROS2 (Robot Operating System). Students were divided into teams and were tasked with programming navigation and path planning algorithms on an autonomous racecar. At the culmination of the class, teams raced against each other in a competition around the Johnson Track and navigating the Stata Basement.
Software
I contributed to programming a wall-following algorithm and assisted in implementing A* path planning and Monte Carlo localization. Through this work, I learned the importance of bridging the gap between simulation and reality, as many parameters required tuning during real-world testing despite appearing to work flawlessly in simulation. I also programmed the camera to detect orange cones and helped implement a safety controller that slowed the robot to a complete stop when obstacles were detected within range.
Results
Our team performed well in each competition. I really enjoyed taking this class as it significantly deepened my knowledge of ROS2. It also was a great learning experience in gaining exposure to localization and path planning algorithms. As part of the course, my team maintained a website to document our progress and put together a final research paper summarizing our learnings and results.