The goal of this project is to develop an autonomous driving stack for Duckiebot hardware platform using ROS2. The final system should enable the robot to navigate a city-like environment and follow road markings.
The robot should be capable of:
- Following road markings autonomously
- Navigating through intersections
- Localize itself in the environment using April Tags placed at intersections
- Follow a predefined (possibly looped) route
- Emergency stop in front of large object
- Understand and follow traffic light colors
Report
To achieve these goals, students had to understand the ROS2 framework, learn how to work with robot’s sensors and actuators, and implement a set of nodes that can perform the target actions. Working with robots is not trivial, starting with the launch of the project. Robots operate in the real world with a great deal of uncertainty, and quite often a situation arises where a program that was just working perfectly starts to throw errors. The students coped perfectly with all the challenges. As a result, students implemented an application that can do each of the target actions and combine all of this in one ROS2 node that sequentially calls the rest nodes. See a demo and a project repository.
