After successful completion of the course, students are able to implement fundamental methods of robotics such as mobile robot navigation indoors with laser sensors as well as using cameras for object recognition and localisation. The application to test is RoboHockey, where students program a small robot to find a push a puck.
Mobile Robotics:- Overview about motors and sensors and their modelling for robotics- Robotic locomotion- Coordinate transformations- Planning, navigation- Behaviour- Mapping, SLAM (Simultaneous Localization and Mapping)
Computer Vision:- Basics, image representation, color spaces- Filter operations- Feature extraction
The exercise part consists of a "Robohockey" competition with its final at the end of the course, where always 2 mobile robots compete against each other and have to score points (recognize the field, score goals...). Over the course of the semester, teams with 3-4 students each program a mobile robot, such that - using laser and camera sensors - it can automatically recognize the field, the opponent, the goals and the pucks and autonomously navigate and play. The development is done in a 3D simulator first, then code is deployed to the real robot. It is possible, to assemble a team before the first lecture.
Details about the exercise can be found here. A native linux installation (preferably Ubuntu 18.04) and programming skills in C++ or Python are absolutely mandatory!
75% counts the implementation work for RoboHockey graded individually but done in teams.
25% counts the final exam over the content of the lecture.
Limited number of students, attendance in the first lecture is mandatory
Programming experience in C++ or PythonLinux
