Robot Motion Planning

What it is about

You learn methods for planning robot movements in different information scenarios: from planning procedures with complete environment knowledge to local, partially-known environments to probabilistic sampling methods for high-dimensional spaces. You will also learn data fusion and filtering procedures (Kalman filter, particle filter) with which maps are created and Simultaneous Localization and Mapping (SLAM) implemented, so you can plan collision-free trajectories.

What you will be able to do

• Understanding direct planning methods for low-dimensional problems
• Application of probabilistic planning approaches for high-dimensional spaces
• Design of planning systems based on a-priori information and sensor data
• Construction of maps and SLAM using Kalman filters and particle filters
• Planning efficient, collision-free trajectories under uncertainty

What the module consists of

• LectureProvides the theoretical foundations of planning and data fusion
• Exercise/TutorialSolving problems independently in depth; solutions are presented during the exercise session
• Guest lectures/ApplicationsShow practical examples and industry perspectives on the lecture topics

Teaching method

• LectureIntroduction to the content and concepts
• Interactive DiscussionDeepening and clarification of open questions
• Recorded Lectures (Self-Study)Enable independent review and repetition
• Tutorials for self-studyEncourage autonomous application of material; preparation for presentations in the exercises