Technische Universität München (TUM)

Technische Universität München is ranked the top academic and research institution in Germany consistently in several independent rankings in recent years. It provides an excellent environment by substantial funding from the Bavarian state government. One of the missions of the University is to boost interdisciplinary research between engineering, medicine, and humanities.

The Intelligent Autonomous Systems (IAS) group at the Department of Informatics employs about 20 doctoral students in the areas of perception, interpretation, and analysis of intentional activities and plan-based and action-aware control of robotic agents. It is one of the core research groups in the German cluster of excellence Cognition for Technical Systems, where it is responsible for one of the demonstrators – the Assistive Kitchen. The IAS group is in close contact and exchanges researchers with leading research groups all over the world (Willow Garage, CMU, Intel Research Pittsburgh, Georgia Tech, LAAS-CNRS, etc).

The research focus of IAS is cognition-enabled mobile manipulation with a particular emphasis on bridging the gap between AI and robotics control in the context of integrated robot systems. To this end, IAS targets its the main research threads at knowledge-enabled robot perception, knowledge representation and reasoning for robots, Bayesian cognitive robotics, and plan-based control of robotic agents and the tight integration of these component technologies. IAS has a proven record of successfully demonstrating their research results in the context of complete robotic systems, including autonomous robot soccer and, most recently, robots preparing meals.

Most research results of the IAS group are provided as open-source software, primarily in the ROS software library. For their research activities and the dissemination of these activities as open-source software the group has won an iCub and a PR2 robot. The software contributed to the PR2 beta program (CRAM -- Cognitive Robot Abstract Machine) will also form the basis for the integration of symbolic and sub-symbolic system components in the RoboHow project.

Inside RoboHow, TUM coordinates the project and takes the lead in the work packages “Knowledge representation and processing” and “Plan-based control”. TUM will also lead the integration efforts on the robot platforms TUM-James and TUM-Rosie, which will be used for the RoboHow milestones.