Developing a lightweight neural network model that can recognize handwritten digits locally on the duckiebot.
The duckiebots tries to earn their driver's license
Using ML to recognize single digits
Implementing a behaviour that tries to follow a robot without crashing into it.
We used OpenCV to extract features from the duckiebot's camera to do lane following, april tag detection, and localization using sensor fusion.
Using odometry and a PID control to navigate around duckietown.
Our first lab assignment. Here we assembled our own duckiebot and play around with duckietown's ROS framework for the duckiebots.