Curriculum Vitae

Rudolf Lioutikov

Research Interests
Robot learningThe human environment is highly dynamical. Manually programming robots for such an environment is infeasible, hence robots that are supposed to assist in our daily life need the ability to adapt to changing environments.
Human-Robot InteractionRobots which are supposed to actively participate in our daily life, through assistance or service, will work in close proximity to or even collaborate with humans. These robots need to be safe around humans. Furthermore they need to react to humans and even anticipate their actions.
Motion Segmentation & Skill LibrariesTraditional programming of robots is expensive and occasionally infeasible. Learning from Demonstration offers a promising alternative, however demontrating each skill independetedly is very tme consuming and might not be very intuitive. The demontration of an entire task however is very natural. Motion Segmentation and allows the robot to extract the necessary skills into a library of skills.
Current Position
Since 2013Ph.D. student at Intelligent Autonous Systems Group
 Technische Universität Darmstadt, Germany
 Topic: Semi-autonomous learning in human-robot interactive manipulation tasks
 Supervisor: Prof. Dr. J. Peters.
Educational Background
2008-2011Master of Science in Computer Science
 Technische Universität Darmstadt, Germany.
 Thesis: "Learning time-dependent feedback policies with model-based policy search" (2013).
 Supervisors: Gerhard Neumann, Jan Peters
2006-2010Bachelor of Science in Computer Science
 Technische Universität Darmstadt, Germany.
 Thesis: "Porting ROS to Windows" (2008) .
 Supervisor: Thomas Lens.

(:authorsearch lioutikov :)

2014International Conference on Intelligent Robots and Systems (IROS) 2014