Current page: Members->List of Members->Christof Eberst
 

Christof Eberst
Profactor Produktionsforschungs GmbH
Membership Number: 42
Address: Wehrgrabengasse 1-5, A-4400 Steyr, AUSTRIA
Email: christof.eberst@profactor.at
Phone: +43-7252-884-230
Fax: +43-7252-884-244
URL: http://www.profactor.at

Biographical Sketch
Christof Eberst received his P.h.D in Electrical Engineering for his work "Incorporation of Recognition Strategies in Sensory Exploration" from the Technische Universität München, Munich in 2000. In 1993 he developed software for robot-perception at Siemens ZFE. 1993-1994 he performed research on human vision at Dep. of Psychology at the University of Houston. In 1999 Dr. Eberst was delegated to the Center for Autonomous Robots at the Royal Institute of Technology at Stockholm, Sweden where he developed vision-based behaviors for control of motion. Dr. Eberst lead and participated in several research projects on computer-vision, telepresence/telerobotics, vision-based manipulation and vision-based autonomous robots. Since 2000, he is senior researcher of PROFACTOR in the group of sensor-based robotics, since 2001 he is head of this group. His current research interest is in active computer-vision for control of robots and particular in integration of learning, perception, interpretation, feedback and action for dependable vision systems and cognitve vision-based robots.

Profactor Produktionsforschungs GmbH (Group for Sensor-based Robotics)
The group for sensor-based robotics at Profactor works on development of 2D and 3D vision systems for control of robots, efficient human-machine interfaces and multi-sensor/multi-robot systems, stereo-vision, reconstruction and map-building, visual inspection and quality control systems. The group has long-term experience in mobile vision-based manipulation and sensor-based control of redundant manipulators, and sensor-based robots for manipulation of unknown parts or parts with variances. Profactor is participating in the CV-project ActIPret. In the field of cognitive vision the main areas of interest are:

1. Frameworks for Cognitive Vision systems

2. learning

- distributed learning

- distributed decision

- learning of vision-based behaviors and cooperative control of multi-roboter systems

- contract management and control in stationary multi-robot systems

- feature learning

3. recognition

- recognition of objects

Profactor targets to investigate autonomous cooperative multi-sensor / multi-robot systems for manufacturing and human-machine interfaces with cognitive capabilities. Our main focus of research is the evolving behavior and interaction between sensor-systems, robots, behaviors and services in multi-robot systems. The system is controlled by multiple concurrent operating vision- and control-services, each integrating limited cognitive capabilities. For this topic we focus on communication and interaction structures, the role of feedback in decision making and computer vision, and the incorporation of information from reasoning. Also the interaction of distributed modules with limited decision and learning capabilities based on local information to obtain globally (sub-) optimal solutions is investigated. Recently we have started to use and evaluate CV methods in multi-sensor obstacle avoidance, robotic behaviors, attentive and investigate vision behaviors, human-machine interaction and for the control of heterogeneous robots.


Site generated on Friday, 06 January 2006