Duckietown is an open research project in the domain of autonomous robotics. It consists of two parts—robots (“duckiebots”) and a town ("duckietown") where these robots are driving around. All components of real traffic environment like traffic lights, signs, roads with markup, and some obstacles are modeled in the town. Each mobile robot is built on an autonomous chassis and it is controlled by a Raspberry Pi computer with ROS. All software is autonomous. A simple RGB-camera is used for obtaining information about the environment. Despite the technical simplicity, this project allows us to do the the following:
- move in accordance with road markup using computer vision;
- avoid obstacles;
- perform localization and mapping;
- perform local and global path planning;
- interact with other robots.
Duckietown combines simplicity and high potential for learning and research.
Initially, Duckietown was created at the Massachusetts Institute of Technology (MIT) as an educational and research workbench for autonomous robotics algorithms, and now it is spreading around the world.