The Bielefeld team comprises members from the Research Institute for Cognition and Robotics (CoR-Lab) and the Excellence Center in Cognitive Interaction Technology (CITEC) from the groups in Cognitive Robotics and Learning (CogRob, Prof. J. Steil, CoR-Lab), in Cognitive Systems Engineering (CSE, Dr. S. Wrede, CoR-Lab), and in Neuroinformatics (NI, Prof. Ritter, CITEC).
These groups have strong experience with interactive robotics, redundant and compliant robots, visual recognition, grasping and manual intelligence, and software engineering. They are in command of several Kuka-LWR and an OmniRob. The groups have participated in various robotics oriented projects (e.g. EU-FP7 AMARSi with coordinator Steil, FP7-ECHORD, FP7-HUMAVIPS) and they are involved in many projects in the German leading edge cluster in Intelligent Technical Systems, where they develop methods for machine learning (project “Selbstoptimierung”), human-machine interaction, robotics and system integration (project “Mensch-Maschine Interaktion”, in collaboration with industrial partners e.g. Miele, Böllhoff, Kannegiesser). Finally, CoR-Lab develops novel production technology by connecting ICT for manufacturing, automation technology and robotics (projects “FlexIMon” and “FlexIMiR”, together with industrial partner Harting).
The team has ample expertise in the different fields involved in the challenge.
- shown how to operate a Kuka LWR in constraint environments (Kinesthetic teaching using assisted gravity compensation for model-free trajectory generation in confined spaces. J.J. Steil et al. In: Echord - European Clearing House for Open Robotic Development, Springer Tracts in Robotics, vol. 94, pp. 107-127, 2013)
- shown how to learn inverse kinematics and provide automatic path planning (Assisted Gravity Compensation to Cope with the Complexity of Kinesthetic Teaching on Redundant Robots.
C. Emmerich et al (ICRA), pp. 4307 - 4313, 2013; D. Seidel et, Model-Free Path Planning for Redundant Robots Using Sparse Data from Kinesthetic Teaching, IROS 2014, to appear)
- devised model-based software engineering tools including a domain specific software language for automatic code generation for robotic motion control systems (e.g. Nordmann et al, "A Domain-Specific Language and Simulation Architecture for Motor Skill Exploration", Workshop on Software Development and Integration in Robotics (SDIR), Karlsruhe, 2013, www.amarsi-project.eu/deliverables: D7_5)
- developed methods for model-free scene interpretation in robotic contexts (Real-Time 3D Segmentation of Cluttered Scenes for Robot Grasping. Ückermann A, Haschke R, Ritter H (2012) IEEE Humanoids 2012)
These skills are integrated by means of a event-based open-source middleware (Robotics Service Bus, RSB, https://code.cor-lab.org/projects/rsb) that also provides bridges into ROS as used in EuRoc.