Dr. rer. nat. Bernd Kleinjohann


Dr. rer. nat. Bernd Kleinjohann

Bernd Kleinjohann is research assistant in the research group Design of Distributed Embedded Systems.

Address
C-Lab
Universität Paderborn
Fürstenallee 11
33102 Paderborn
e-Mail:bernd@c-lab.de
telephone:+49 5251 60-6101
facsimile:+49 5251 60-6066
room:FU.214






CV

Bernd Kleinjohann received his diploma in computer science from University of Dortmund in 1985 and his Ph.D. from University of Paderborn in 1994. Since 1985 he is with C-LAB a joint R&D institute of University of Paderborn and Siemens. He was/is project leader of several national research projects and responsible for many international research projects at C-LAB. His current position is Vice Director of C-LAB and head of the Cooperative Systems Group working in the fields of intelligent mobile systems, embedded real-time systems and distributed interactive systems. He acted as Professor for Technical Computer Science during summer term 2003 and winter term 2003/2004.

Professional scientific activities

Since 2006 Dr. Kleinjohann is Vice Chair of the IFIP Working Group on Embedded Systems (WG10.2). He has chaired the IFIP Working conference on Distributed and Embedded Systems (DIPES) since 2000. He co-authored more than 100 publications and was advisor of more than 50 diploma, bachelor and master theses.

Scientific profile and research projects

ESLAS- A modular approach for Evolving Societies in Learning Autonomous Systems

As part of the DFG’s Organic Computing Priority Program (SPP 1183) the project ESLAS (Evolving Societies in Learning Autonomous Systems) is conducted under Dr. Kleinjohann's supervision. The goal of this project is to investigate new approaches to learning in groups of heterogeneous robots. The ESLAS architecture is able to learn at the different levels of abstraction: 1) an active strategy-learning module that uses reinforcement learning and 2) a dynamically adapting skill module that proactively explores the robot’s own action capabilities and thereby provides actions to the strategy module. A robot uses already known skills and strategies to understand the behavior of other robots and eventually imitate them. This sporadic imitation is often the only possibility to learn from other robots in a group. Different goals to be pursued by the robot can be specified by means of an intuitive motivation system.

Collaborative Research Center 614 "Self-optimizing concepts and structures in mechanical engineering"

The DFG Collaborative Research Center 614, Self-optimizing concepts and structures in mechanical engineering, takes a different angle on organic computing. Ongoing work in the two project parts A2 and C3 is investigating among others the development of a hybrid planning architecture, with the ability to handle continuous-valued system properties. Simulations of system behavior enable relatively precise - even though abstract - predictions of the performance of the real system. For instance they allow to predict the behavior of the system in terms of user-defined goals, energy saving etc.. Since it is difficult or impossible to obtain exact models for all properties of the system and its environment, we use approximation methods to make the simulation as realistic as possible. One such method is fuzzy identification. For example, for modeling the air gap of a linear motor, we use a fuzzy system, that also takes account of environmental parameters such as the weather.

SOGRO – Rapid support for disasters with large numbers of casualties

The BMBF funded project SOGRO (Sofortrettung bei Großunfall mit Massenanfall von Verletzten) has as major target to shorten the initial, often chaotic phase, before the injured are transported to hospital. To obtain an overview of the situation, the project uses aerial video. As part of this work, University of Paderborn (C-LAB) is developing two components: an intelligent distributed image processing system, that combines information obtained from cameras with visible spectrum and infrared spectrum, and a coordination and guidance component, capable of controlling a swarm of drones developed by Stuttgart University. Further partners are DRK Frankfurt (consortium coordinator), Frankfurt Airport (application), Siemens (RFID-based triage and emergency management communications), Andres Industries (PDA hardware) and Freiburg University (legal, social and economic issues).

E-Mobil – Simulation supported design for electric vehicles

This project is funded within the North Rhine Westfalian Progres initiative. Its goals is the development of a simulation based design method and environment, that enables early tests and experiments with virtual (simulated) artifacts from the electromobilty aresa instead of real vehicles or vehicle parts as for instance batteries, hereby enabling the development of control and optimization strategies across different domains, like mechanics, electronics or software. Project partners are dSPACE GmbH (project coordinato), DMEcS GmbH & Co. KG, Infineon Technologies AG, Warstein) and the University of Paderborn with LEA and C-LAB. The focus of Dr. Kleinjohann’s group at C-LAB is the development of a design method incorporating cross-domain simulation and optimization and the development of simulation acceleration methods within the envisioned design environement.

Intelligent embedded systems and robotics

In this area Dr. Kleinjohann's working group at C-LAB has developed two classes of robots. MEXI is a robotic head, which uses facial expressions and natural language to recognize human emotions and to express its own artificial emotions. MEXI allows to study behavior control methods using socio-biologically based models driven by a robot’s emotions and needs. The second class of robots developed in C-LAB are the Paderkickers. The Paderkickers are a team of seven robot soccer players. In 2010, as in previous years, the Paderkickers participated in the RoboCup German Open. The robot soccer application area allows the C-LAB team to investigate issues in computer vision, communication networks, team coordination and team development, amongst others. The Paderkickers also act as a useful support of Dr. Kleinjohann's teaching activities. Large parts of software and hardware have been developed and enhanced in several so called project groups, where about twelve students conduct applied research and development over two semesters.