This interdisciplinary block seminar deals with methods of modelling, generating and controlling movements in humans and robot systems. Students get an insight into this interdisciplinary field and learn the basics of biological motion, biomechanical simulation, robotics, and machine learning. In the introduction, motion generation as effect of muscle contraction is discussed. It will be shown how movement patterns can be identified and categorized based on the observation of human movements and how these patterns can be reproduced on a humanoid robot. Finally, methods for the learning of movement primitives from human demonstration will be presented and their application for the generation of motion for humanoid robots will be explained.

Learning Obejectives:

The student knows procedures for modelling human motion, as well as possibilities for its processing and analysis. He/she knows methods for learning motion primitives and mapping human motion to robots that have different kinematics and dynamics and can apply them in new contexts.

Einrichtung: KIT IAR-H²T Prof. T. Asfour, Universität Stuttgart 

Weitere Beteiligte: Interdisziplinäre Veranstaltung in Kooperation mit:

• Universität Stuttgart, Modellierung und Simulation im Sport
• Hertie Institute for Clinical Brain Research (HIH), Centre for Integrative Neuroscience (CIN)

Prüfungen: Die Erfolgskontrolle erfolgt durch Ausarbeiten eines Wiki-Moduls sowie der Präsentation der selbigen als Erfolgskontrolle anderer Art nach § 4 Abs. 2 Nr. 3 der SPO.

Modul für Master Sportwissenschaften, Maschinenbau, Mechatronik und Informationstechnik, Elektrotechnik und Informationstechnik

Arbeitsaufwand: 90 h

Zugangsvoraussetzungen: Empfehlung:

Die Fähigkeit zum Erstellen von Programmen und die Beherrschung einer Programmiersprache wie z.B. Matlab, Python oder C++.