Description
Self-Driving Cars with Duckietown is an innovative course offered on edX, providing a practical introduction to robotics and AI through the construction and programming of a scale-model self-driving car. This course is designed for learners interested in understanding how autonomous vehicles operate. Participants start with a box of parts and work towards creating a Duckiebot that can navigate autonomously in a scaled-down city environment known as Duckietown.
Read More
The course uses state-of-the-art approaches, software tools, and real hardware to offer an engaging hands-on learning experience. It explores real-world solutions to the challenges of autonomy, translating theoretical concepts into algorithms and deploying them in both simulation and on hardware. The course employs modern software architectures built with Python, Robot Operating System (ROS), and Docker, highlighting the strengths of classical architectures and modern machine learning-based approaches.
Presented by professors and scientists passionate about robotics and accessible education, the course utilizes the Duckietown robotic ecosystem, an open-source platform created at MIT and used by over 200 universities worldwide. Learners can choose between a simulation track and a hardware track, the latter involving the assembly of a Duckiebot and deployment in a Duckietown city track.
The course is supported by ETH Zurich, the University of Montreal, and the Toyota Technological Institute at Chicago. It covers essential topics such as robot subsystems, computer vision-based navigation, reinforcement learning, and the setup of an efficient software environment for robotics. The course is designed for learners with basic knowledge of Linux, Python, and Git, and requires a computer with a native Ubuntu installation.
Participants will gain skills in programming Duckiebots to navigate model city road lanes, recognize and avoid obstacles, and make autonomous driving decisions using both traditional and modern approaches. The course also offers opportunities to submit robot agents to public challenges, testing skills against peers.
- CategoryCorporate Training & E-Learning
- Pricing ModelFree
Self-Driving Cars with Duckietown's Core Features
Hands-on robotics and AI learning
Build and program a scale-model self-driving car
Use state-of-the-art autonomy techniques
Deploy solutions in simulation and hardware
Learn with Python, ROS, and Docker
Explore classical and modern AI approaches
Supported by ETH Zurich and other institutions
Open-source Duckietown robotic ecosystem
How to use Self-Driving Cars with Duckietown?
Assemble: Build your Duckiebot from parts
Program: Use Python and ROS to code autonomy
Deploy: Test in simulation and real-world environments
Challenge: Submit and compete in public challenges
Self-Driving Cars with Duckietown's Use Cases
- Autonomous Vehicle Development
- AI in Robotics
- Hands-on Learning
- Simulation and Hardware Deployment
- Computer Vision Applications
