FB3 AG BKB
			Deutsch English
	Prof. Dr.-Ing. Dipl.-Inform. Udo Frese
	Curriculum Vitae From October 1993 until October 1997 I studied computer science at University of Paderborn. The focus of my studies was on parallel computing and in my Diploma thesis I developed an efficient parallel algorithm for detecting intersection of aligned rectangles. After graduation I changed to robotics and started working as a Ph.D. student at the German Aerospace Center in Oberpfaffenhofen in their mobile robotics lab. My main topic has been the problem of simultaneous localization and mapping (SLAM) which requires the robot to incrementally build a map while moving and determine its own position in the map. From January 2003 to March 2003 I have been guest Ph.D. student at Center for Applied Autonomous Sensor Systems at University of Orebro, Sweden, again working on SLAM. In May 2004 I received the Ph.D. degree from University of Erlangen. My thesis is titled "An O(log n) algorithm for simultaneous localization and mapping of mobile robots in indoor environments". Since January 2004 I am working in the SFB/TR 8 "Spatial Cognition" (project [AmbiguousSpace], A1-[RoboMap]) at University of Bremen. Since June 2008 I am assistant Professor for real time computer vision at University of Bremen. I am also associated with the safe and secure cognitive systems lab of the DFKI in Bremen.

Inside Data-Association (Workshop at RSS 2008, Zürich)

Publications

Teaching & available Thesis topics (German)

Research

I think algorithms are most exciting. - GUI, design, software integration - That's all nice, like having your car equipped with air conditioning or power windows. But what really drives your car is the motor and what really drives your software are the algorithms working underneath. Now the reason, why robotic algorithms are even more fun than ordinary ones is that 'drives your robot' can be taking literally because with robotics abstract algorithms can create real physical behavior.

SLAM

My main research interest is autonomous map making, the so called simultaneous localization and mapping problem. SLAM, i.e., the Simultaneous Localization and Mapping problem, aims at a fully autonomous answer to the question ``Where am I?'' by providing an autonomously built map. It demands from the robot to derive a map from its perceptions while moving through the environment and to simultaneously determine its own position in that map. The problem has been under investigation since the late eighties with first contributions dating back to the work of Chatila, Smith, Self and Cheeseman and Durrant-Whyte. The name simultaneous localization and mapping was later coined by Durrant-Whyte.

Inside Data-Association (Workshop at RSS 2008, Zürich)

Least-Square based SLAM and the Treemap Algorithm

Treemap as a Generic Open Source Backend for Different SLAM Variants

Closing a Million-Landmarks Loop

Treemap: An O(log n) Algorithm for Simultaneous Localization and Mapping

A Multilevel Relaxation Algorithm for Simultaneous Localization and Mapping

A Discussion of Simultaneous Localization and Mapping

Rescue Robotics

[AmbiguousSpace] 3-D Perception and Map Building in Unstructured, Highly Ambiguous Environments

"Sport Robotics"

Apart from the above mentioned more serious and application oriented studies, I am ultimately fascinated by the idea of robots taking part in sport activities. The probably most well known example is the robot soccer competition RoboCup. Personally during my time at the German Aerospace Center I was involved in a different ball game.

A Vision for 2050

A Robotic Ballcatcher

Links

Code

• openslam.org Many open source implementations of different SLAM algorithms, including a grid based FastSLAM algorithm by C. Stachniss and G. Grisetti.
Tutorials
• Thrun, Burgard, Fox: Probabilistic Robotics (highly recommended)
• SLAM Summer School 2002, KTH, Stockholm (Excellent material containing presentations, list of publications and lab exercises)
• W.L. Briggs: A Multigrid Tutorial Introductory tutorial on multigrid methods (excellent)
• Dieter Fox, Wolfram Burgard, Sebastian Thrun: Probabilistic Techniques for Mobile Robots, Tutorial ECAI-2002 (Highly recommended)
• Sebastian Thrun, Robotics Mapping a Survey
• CVonline: Web based compendium of vision topics
• HIPR2: Java based interactive Vision tutorial

People working on Simulataneous Localization and Mapping

• Cyrill Stachniss, University of Freiburg
• Wolfram Burgard, University of Freiburg
• José A. Castellanos, University of Zaragoza, Spain
• Tom Duckett; Department of Technology, University of Oerebro, Sweden
• Gamini Dissanayake, University of Sydney, Australia
• Hugh Durrant-Whyte, University of Sydney, Australia
• Kurt Konolige, SRI International, USA
• John Leonard, MIT, USA
• José Neira, University of Zaragoza, Spain
• Sebastian Thrun, Artificial Intelligence Lab, Stanford University, USA
• Shinichi Yuta, University of Tsukuba, Japan
• Jens Steffen Gutmann, University of Freiburg, Germany
• Mark Paskin, Stanford University, USA

Private:

• Building a paper bridge across the lake in my parent's garden (10m). (German)

Coordinates:

E-mail:	ufrese@informatik.uni-bremen.de
Address: (Postal)	Dr.-Ing. Udo Frese Senior Researcher Safe and Secure Cognitive Systems DFKI-Lab Bremen Robert-Hooke-Str. 5 D-28359 Bremen	Address: (office @ Universitaet Bremen)	Udo Frese Cartesium 1.057 FB 3 Mathematik - Informatik Universitaet Bremen Enrique Schmidt Str. 5
Telephone:	+49 421 218-64207
Fax:	+49 421 218-9864207

     
Author: Dr. Udo Frese     AG BKB 

Last updated: August 1, 2008