Annual Report 2002 - Örebro universitet

Annual Report
2002
Örebro universitet Örebro University
Institutionen för teknik Department of technology
701 82 Örebro SE-701 82 Örebro, Sweden

Annual Report 2002 1
Contents
SUMMARY 2
PREFACE 4
1. CENTER FOR APPLIED AUTONOMOUS SENSOR SYSTEMS (AASS) 6
1.1 ORGANISATION 9
1.2 MANAGEMENT 11
1.3 FUNDING 12
1.4 UNDERGRADUATE EDUCATION 16
1.5 GRADUATE EDUCATION 17
1.6 CO-OPERATION 21
1.7 DISSIMINATION OF INFORMATION 30
2 RESEARCH LABS 34
2.1 BIOLOGICAL INSPIRED SYSTEMS LAB 34
2.2 MOBILE ROBOTICS LAB 41
2.3 INTELLIGENT CONTROL LAB 48
2.4. LEARNING SYSTEMS LAB 57
3. PUBLICATIONS 65
3.1 BOOKS AND EDITED VOLUMES 65
3.2 CHAPTERS IN EDITED VOLUMES 65
3.3 JOURNALS 65
3.4 CONFERENCE ARTICLES 66
3.5 TECHNICAL REPORTS 68
3.6 PATENTS 68

2 AASS – Center for Applied Autonomous Sensor Systems
Summary
The Applied Autonomous Sensor Systems Center (AASS) organizes all research and graduate
education in the area of Autonomous Systems and also is responsible for the undergraduate
and MSc programs in Computer Engineering (Datateknik) at the Dept. of Technology, at
Örebro University in Örebro, Sweden. The research effort has a pronounced interdisciplinary
character being a fusion of a variety of disciplines from Systems and Control, Measurement
Science and Sensors, Computer Science and Artificial Intelligence, and Operations Research
but, with emphasis on Fuzzy Systems, Intelligent Control and Sensors, Machine Learning,
Planning and Automated Reasoning, and Biologically Inspired Systems.
The Applied Autonomous Sensor Systems Center is organized in four laboratories,
responsible for research and graduate education in their respective areas:
1. Biologically Inspired Systems Laboratory (BIS)
2. Mobile Robotics Laboratory (MR)
3. Intelligent Control Laboratory (IC)
4. Learning Systems Laboratory (LS)
The common research focus for the four research laboratories is the subject of perception and
autonomy in diverse un- and/or semi-structured environments. This subject is pursued using a
common research methodology, which is strictly applied: all technologies under investigation
should answer actual needs that emerge from an application, and their effectiveness is always
validated on the application. Applications considered by AASS include mobile platforms,
e.g., mobile robots, and immobile platforms, e.g., industrial plants and processes.
This report presents AASS organization and management, developments in education,
research and industrial cooperation, activities within the national/international research
community, and various forms of societal interaction that have taken place in year 2002. In
this context the major achievements that characterize year 2002 are as follows:
• During 2002 the number of Phd students has increased from 13 to 22.
• A new research line on Cooperative Robotics started, which has got initial momentum by
four related events: (i) a 5-year grant from CUGS (national graduate school in computer
science), (ii) the employment of two new PhD students, (iii) the organization, jointly with
the Technical University of Lisbon, of the First European Summer School on cooperative
robotics, and (iv) our successful participation at the 2002 international RoboCup
competition.
• Lars Karlsson has been awarded a 3 year grant by Vetenskaprådet (Swedish funding
agency for basic research) to work on planning and plan execution under uncertainty.
• Silvia Coradeschi and Alessandro Saffiotti have been appointed guest editors for a special
issue on "Anchoring" of Robotics and Autonomous Systems, one of the major journals in
this field. This demonstrates the recognition, by part of the international community, of

Annual Report 2002 3
the importance of the subject of "anchoring", invented at AASS, and of the world leading
role played by AASS in this subject. This special issue will appear in 2003.
• Work by Tom Duckett on fast, on-line mapping appeared in a special issue of
Autonomous Robots Journal, one of the most prestigious journals in the field.
• A University of Örebro prize for “bästa nytänkande ex-jobb” for the work on medical
image processing done under the supervision of Tom Duckett.
• AASS started its official activities as a European Marie Curie Training Site on advanced
autonomous robotic systems.
From the top left: Henrik Andreasson (Phd-student), Grzegorz Cielniak (Phd-student), Juan Pedro
Canovas (Phd-student), Pär Buschka (Phd-student), Boyko Iliev (Phd-student), Tom Duckett (Lab
leader), Christer Lindkvist (Phd-student),
Middle row: Alexander Skoglund (Research Engineer and Phd-student), Abdelbaki Bouguerra (Phdstudent),
Udo Frese (Phd-student), Ester Liljedahl (Project Coordinator), Gustav Tolt (Phd-student),
Ola Pettersson (Phd-student), Zbigniew Wasik (Phd-student), Per Sporrong (Research Engineer), Peter
Wide, (Manager AASS), George Fodor (Adj.Prof.), Anani Ananiev (Senior Researcher).
Front row: Linn Robertsson (Phd-student), Malin Lindquist (Phd-student), Lucia Ballerini (Phdstudent),
Li Jun (Phd-student), Letizia Bagnoli (Phd-student), Silvia Coradeschi (Senior Researcher),
Amy Loutfi (Phd-student), Federica Pascucci (Phd-student), Alessandro Saffiotti (Lab leader).
Not present: Dimiter Driankov (Prof. and Research Coordinator AASS), Selim Eskiizmirliler (Lab
leader), Lena Biel (Phd-student), Jan Spännar (Phd-student), Lars Karlsson (Senior Researcher), Kevin
LeBlanc (Phd-student), Robert Johansson (Phd-student), Ivan Kalaykov (Lab leader), Barbro Alvin
(Adm. Ass.), Martin Persson (Phd-student), Lars Jennergren (Phd-student) and Per Munkevik (Phdstudent).

4 AASS – Center for Applied Autonomous Sensor Systems
Preface
The Center for Applied Autonomous Systems was formally established in 1996 at the
Department of Technology with a grant from the KK-Foundation aimed at establishing the
basis for a concentrated effort on interdisciplinary research in the area of Autonomous
Systems with contributions to education. In the period 1996--2002, AASS has developed from
a research group of 2 researchers and 2 PhD students to a full-fledged research center with 4
professors, 2 Associated Prof. (Docent), 5 Phd, 22 Phd students, and 4 technical and
administrative personnel. Now AASS plays an important educational role within the
Department of Technology as well as has a leading position in the 4th strategic research
profile of Örebro University - Human senses, autonomous sensor systems, and industrial
processes. It also has achieved national and international recognition for its research results,
and established well-functioning industrial cooperation. In the context of these achievements,
AASS has, during year 2002, also had a measurable societal impact in a number of ways such
as
1. Industrial relevance: continued cooperation with national/regional industrial companies
such as ABB Automation AB, Bofors Defence AB, Aerotech Telub AB and SMEs like
Amphitech AB.
2. European dimension: co-operation within three EC research networks and a coordinator
for a EU Marie Curie Training Site
3. National dimension: Start-up of a VINNVÄXT VINNOVA initiative “Robotdalen”
together with MdH Mälardalen, KTH/CAS Stockholm and a large number of industrial
and societal partners.
4. Public acceptance: Participation in CEBIT 2002, Hannover, Germany and the Stockholm
Technical Fair.
However, our most important contribution with long-term societal impact is our educational
activity. Improving the content, quality, and degree-level of the undergraduate education in
The Department of Technology in general and in Computer engineering in particular is of
highest priority. Thus undergraduate students with BSc, MSc, and Engineering degrees will
remain our most important “output” - these students are to play important role in the
development of regional and national industrial and business enterprises. At the same time
research and graduate education will be key factors in keeping with up-to-date technological
developments, maintaining and further developing the scientific relevance and quality of
undergraduate education, and last but not least, increasing the attractivity of Örebro
University as an educational center in a regional, national, and international perspective. On
the other hand, our research and graduate education is of great value in itself -by doing
research in the area of Autonomous Systems we contribute to the advancement of
technologies at one of the frontiers of modern interdisciplinary science. A particular
“practical” benefit is the reunification of Measurement, Control, and Computer Sciences into
a disciplinary and technological continuum - a reunification that sadly is lacking today, to the
growing discontent of industry. Thus, industrial cooperation will ensure that these
technological advancements are adopted and their business potential is exploited.

Annual Report 2002 5
Even though the AASS research, educational, and industrial cooperation profile is still
developing there are certain “key” figures that indicate that we are on the right track:
• Number of students in the Computer Engineering undergraduate program 2002:
186 BSc students in all 3 years of education.
• Number of MSc students in Computer Engineering undergraduate program 2002: 49
• Number of PhD students 2002: 22 (4 lic-degrees awarded).
• Number of ex-jobs under AASS supervision 2002: 4st.
• Number of reviewed publications 2002: 44
• External funding 2002: 62% of the total AASS budget.
A continued strong support from our funders and academic/industrial partners shows that we
have all reasons to be optimistic about our further development and contributions to
education, industry, and society.
Peter Wide
Director AASS
Dimiter Driankov
Research Coordinator AASS

6 AASS – Center for Applied Autonomous Sensor Systems
1. Center for Applied Autonomous Sensor Systems (AASS)
The Applied Autonomous Systems Center was formally established in 1996 with a grant from
the KK-Foundation with the purpose of establishing the basis for a concentrated research
effort in the interdisciplinary area of Autonomous Systems with expected contributions to
education and industrial cooperation. In the period 1996 - 2002 AASS has achieved national
and international recognition for its research results, plays an important educational role
within the Department of Technology and in the 4th strategic research profile of Örebro
University - Human senses, autonomous sensor systems, and industrial processes.
Furthermore, AASS has established well-functioning industrial cooperation - a number of
industrial partners have developed a strong interest in AASS and support the research effort in
terms of industrial PhD students, joint research projects, research-relevant equipment
(hardware and software), and expertise (industrial employees acting as part-time AASS
researchers).
The AASS research effort is organized on the basis of the following 4-pronged vision:
1. Recognition by the international and national scientific community not only for AASS's
active contributions to the state-of-the-art in core research areas, but also for its innovative
role in establishing novel research topics that expand the frontiers of the science and
technology for autonomous systems;
2. Significant, long-term industrial interest in utilizing AASS's research results and
educational capacity that will allow improvement of existing industrial technologies and
products as well as, provide possibilities for exploring radically new solutions and future
products;
3. Significant contributions to educational programs at BSc, MSc and PhD-level of
education; and
4. Leading role within the Örebro University's strategic research profile “Human Senses,
autonomous sensor systems, and industrial processes''.
In engineering terms, the fulfillment of the above vision will lead to:
• Providing practitioners with methods for the design of engineering systems at all levels of
control and sensing (low, supervisory and task-level) and whose needs are not met by
existing conventional approaches.
In educational terms, the fulfillment of the above vision will lead to:
• The reunification of Measurement, Control, and Computer Sciences into an
interdisciplinary disciplinary continuum - a reunification which sadly is lacking today
from existing educational programs.

Annual Report 2002 7
Autonomous Systems
Autonomous systems are mobile as well as immobile platforms that employ a vast array of
sensors in order to analyze and/or influence highly dynamic environments.
Examples of such systems include robotic systems, complex process control
systems, flexible manufacturing/inspection systems, automotive systems, and
unmanned underwater/land and air/space vehicles.
In order to achieve an autonomous operation the system has to be supplied
with the ability to: plan and schedule control/sensing actions at varying
levels of detail and functionality; acquire and integrate sensory information
at different levels of granularity and abstraction; learn how to adapt to initially
unknown or changing environments; identify events that threaten the system's
operation and react to these by reconfiguring its control and sensing routines.
The design and analysis of autonomous systems has a pronounced
interdisciplinary character being a fusion of a variety of disciplines from
Systems and Control, Measurement Science and Sensors, Computer Science
and Artificial Intelligence, and Operations Research but, with emphasis on
certain modern developments such as Fuzzy Systems, Intelligent Control and
Sensors, Machine Learning, Planning and Automated Reasoning, and
Biologically-inspired Systems.
There are four major research areas that constitute the basis for autonomous
systems research and development: Perception Science and Technology, Reasoning Science
and Technology, Action Science and Technology, and Learning Science and Technology.
These four research areas and the challenges posed by them are well recognized by the
international research community and underpin the mainstream international research effort in
the area of autonomous systems. Perception provides the means to gather information about
the working environment - information that permits operations such as manufacturing,
process control, navigation, monitoring, and manipulation to be accomplished safely and
robustly. Reasoning is the ”smarts” of an autonomous sensor system - it is the capacity to
reason that provides the connection between perception and action. Action is ability to move
in space and to manipulate objects in a safe and reliable manner - in case of robotic systems,
or maintain certain desired behavior in industrial plants/processes despite disturbances and
limited knowledge of process/plant dynamics. Learning is needed to make technologies from
the areas of Perception, Reasoning, and Action play a practical role in a variety of real-world
dynamic, uncertain and unstructured environments. AASS research reflects these major
research areas, and - what is more important - aims at their integration in order to “produce”
complete autonomous systems solutions. Towards this purpose we have defined a common
demonstrator platform, to which all AASS research projects are contributing. This
demonstrator is seen as central means for the integration and cross-fertilization between the
four laboratories and serves as a convincing demonstration platform towards industry and
funding bodies. The demonstrator of choice, a mobile manipulator, has its potential
application in a number of industrially relevant areas: inspection of industrial sites and
installations, inspection of hazardous waste sites and waste manipulation, product and
material monitoring in manufacturing, and rescue operations. Its industrial relevance is further
reinforced by the present inadequacy of the predominant number of mobile
robots/manipulators in industrial use: very limited autonomy, limited task range, predictability

8 AASS – Center for Applied Autonomous Sensor Systems
of behavior only in well-structured and predictable environments, and
limited sensing capabilities.
AASS is organized around the following interrelated and mutually dependent permanent
activities:
1. Research projects, including technical support pertaining to particular labs or “inter-labs”
projects.
2. The AASS Graduate program, leading to both Licentiate and PhD degrees.
3. The MSc program in Computer Engineering - an educational program at the Department
of Technology reflecting AASS related research.
4. Undergraduate studies.
5. EU Marie Curie Training Program in autonomous robotic systems - duration 2002-2004.

Annual Report 2002 9
1.1 Organisation
AASS is organized in matrix form, as shown in the figure below.
AASS Management
Prof. P. Wide, Director
Prof. D. Driankov, Research Coordinator
E. Liljedahl, Project Coordinator
B. Alvin, Administrative Assistant
Research Laboratories
BIS Lab
S.Eskiizmirliler,
PhD
MR Lab
Prof. A Saffiotti
IC Lab
Doc. I. Kalaykov
Research Projects
Technical Support Group: Per Sporrong and Alexander Skoglund
LS Lab
T. Duckett, PhD
Graduate Program
Director: Lars Karlsson, PhD
MSc Program in Computer Engineering
Program Coordinator: Lars Karlsson, PhD
Undergraduate Studies at Division of Computer Engineering
Director of Studies: Silvia Coradeschi, PhD
EU Marie Curie Training Program in Autonomous Robotics (2002-2004)
Program Coordinator Prof. Dimiter Driankov
The research labs reflect the four main directions in the research area of Autonomous
Systems, i.e., BIS - Perception Science and Technology; ML - Reasoning Science and
Technology; IC - Action Science and Technology; and LS - Learning Systems and
Technology. Thus they represent competence areas within which basic research is conducted.
According to their disciplinary competence, e.g., Computer Science, Computer Engineering,
AI, Systems and Control, Machine Learning etc., they also provide resources - courses,
teachers, and supervisors - for the graduate and undergraduate studies programs and the rest
of the horizontal entries in the above figure:
1. Research projects are defined within labs (lab-specific projects), across labs (inter-lab
projects), and/or cooperation with external partners (external projects). The lab leaders are
responsible for the budget and project work on lab-specific projects and external projects.
The AASS research coordinator is responsible for inter-lab projects. The technical support

10 AASS – Center for Applied Autonomous Sensor Systems
group provides technical assistance in the design and maintenance of robot platforms,
sensor systems and the associated work with these software environments and
programming tools.
2. AASS Graduate program defines the PhD studies curriculum for all graduate students at
AASS independent of their specific research area. The director of the program is
responsible for its structure and contents, as well as the enrollment of graduate students
and supervision of the formal aspects of progress in their PhD studies. The actual PhD
thesis work and its supervision is done within the research labs. AASS graduate students
have the possibility also to be enrolled in the Swedish National Graduate School in
Computer Science (CUGS) where Örebro is one of the participating universities. The
Director of the AASS Graduate program also is a member of the steering committee of
CUGS.
3. MSc program in Computer Engineering is a 4-year studies program, which upon its
completion provides for 160 points of studies and leads to the degree “MSc in
Technology, with profile in Computer Engineering”. The first two years of this program
are common with the 3-year undergraduate program in Computer Engineering at the Dept.
of Technology. The decision whether to join the MSc program is done upon the
completion of the second year of the 3-year undergraduate program. The coordinator of
the MSc program is responsible for the program's structure and content, its monitoring and
“marketing”, enrollment, assignment of teaching staff etc. In its current form the program
contains courses directly related to the disciplinary and research competence at AASS.
4. Undergraduate education in Computer Engineering is a 3-year studies program given
within the Computer Engineering Division of the Dept. of Technology. The director of
AASS also is director of the Computer Engineering Division, and Silvia Coradeschi, a
researcher at AASS, is a director of studies for the Computer Engineering program who is
responsible for both the 3 and 4 years programs. The responsibilities of the director
include program structure and content, teaching staff assignment, enrollment, budget work
etc. AASS research staff and PhD students are natural part of the teaching staff.
5. EU Marie Curie Training Program in Autonomous Robotics is a 4-year, EU-funded
program, aimed at providing support for the training and mobility of PhD students
throughout Europe. This scheme supports short stays by young researchers pursuing
doctoral studies in the area of Autonomous Systems, providing them with the possibility
of undertaking part of their doctoral studies in a country other than their own, and
allowing them the benefit of working within an internationally recognized group - in this
case AASS - in their specialized area of research - a particular AASS laboratory. The
coordinator of the program is responsible for all organizational, management, and budget
work as well as for contacts with the EU Commission, “marketing”, selection and
enrollment of candidates. Actual research work supervision is done within the respective
labs.

Annual Report 2002 11
2.2 Management
The management, evaluation, and monitoring of AASS activities is performed by the AASS
Management Body which consists of the AASS Director, AASS Research
Coordinator, lab leaders, and the coordinators of undergraduate/graduate programs.
The major decisions made by the Management Body concern:
1. Budget;
2. Research strategies;
3. Research/Education integration in the Computer Science (Datateknik) program;
4. Long-term research/education planning;
5. Recruitment;
6. Personnel career development;
7. AASS university/department role
The AASS Director who chairs the steering committee is responsible for budget preparation,
recruitment, career development, and interface with the department, the university and with
industry and funding bodies.
The Advisory Board annually reviews AASS research and progress from both a scientific and
industrial perspective and can suggest corrections if necessary. Currently the Advisory Board
consists of:
Hans Skoog, Professor, Strategy and Technology Advisor, ABB Corporate Research,
Västerås, Sweden.
Torbjörn Widmark, PhD, AssiDomän Frövi, Product Development & Technical Service,
Frövi, Sweden
Alexander Lauber, Professor in Measurement Technology, Department of Technology,
Kalmar University, Kalmar, Sweden.
Rainer Palm, Dr. Ing., Dr. Sc.-techn., Corporate Technology Information and
Communications, Siemens AG, Munich, Germany.
Emil M. Petriu, Dr. Eng., P. Eng., Professor, School of Information Technology and
Engineering, University of Ottawa, Canada.
The AASS Research Coordinator is responsible for the formulation and progress of inter-lab
projects, long-term research planning and recruitment needs, and planning and coordination
of research grant applications. He also monitors the AASS Graduate Studies Program and
prepares and organizes progress reports and evaluations concerning AASS activities.
The Lab leaders are responsible for all issues pertaining to lab organization management, and
development. This includes, planning and coordination of lab-specific research projects as
well as lab administration, e.g. budget handling, lab “marketing'' (internal and external), lab
contributions to teaching, etc.

12 AASS – Center for Applied Autonomous Sensor Systems
1.3 Funding
AASS funding comes from a number of sources. The major ones are as follows:
1. KK-Foundation grants
In the year 2002 this grant amounted to a total of 7,2 MSEK. 1,8MSEK was a part of the
KK-Foundation platform II project, while 0,8MSEK are two grants, which two of our
Phd. Students received from the KK-foundation. During the years 2002- 2008 the KK-
Foundation stays for a grant of up to 36 MSEK. This grant, won in a tough competition
with other Swedish research centers and groups, is intended to help achieving
internationally competitive research profile, establishing a long-term development
potential and developing sufficient alternative (other than the KK Foundation) funding
sources. Furthermore the grant is to facilitate active industrial cooperation where the latter
is ensured by requiring 1-to-1 match between the KK-Foundation grant and funding from
industry. This yields a total of up to 72 MSEK for a period of 6 years. This requirement
has had quite a positive effect regarding AASS industrial cooperation - existing one has
been further strengthened and new industrial partners have been found. The matching
funding provided by industry is mainly in the form of, industrial graduate students, parttime
participation of industry personnel in external research projects, and expensive
equipment.
2. Faculty funding
For the year 2002 it amounts to 5,5 MSEK. This funding allows AASS to conduct more
basic research and for maintaining and developing the size and quality of AASS
participation in undergraduate/graduate education.
3. Vetenskapsrådet – Swedish Council of Research grants.
This is a traditional funding source though characterized by its extreme competitiveness.
A VR grant of 1,2 MSEK covering a period of 3 years started in 2001 for research on the
subject of signal-to-symbol conversion in mobile autonomous robots. Another VR grant
was obtained in 2002 for research on the subject of Sensor-based planning for mobile
robotics.
4. Regional funding
During 2002 regional industry has supported AASS with 303 000 SEK for equipment and
375 00 SEK for one PhD-student.
As a measure of having achieved the objectives set by the KK-Foundation AASS has adopted
the ratio between internal (faculty/university) and external (KK-Foundation, industry,
research funding agencies) funding. For year 2002 this ratio was 62% for external funding
(58% in 2001).
The diagram from below summarizes the different funding sources in year 2002.
Others
19%
Facultyfunding
35%
KK-
Foundation
46%

Annual Report 2002 13
The chart below shows the increasing funding levels for the years 1997-2002, which
illustrates the rapid AASS development.
AASS funding growth 1997-2002
20000000
15000000
10000000
5000000
0
1997 1998 1999 2000 2001 2002
The Governor of Örebro County, Gerd Engman, is one of the representants of our regional
funding. Here she is visiting AASS’ lab to get information from two of our PhD students, Malin
Lindquist and Amy Loutfi, about the Electronic Nose.

14 AASS – Center for Applied Autonomous Sensor Systems
1.4 Undergraduate Education
AASS is responsible for the 3-year Computer Engineering Program and the 4-year
MSc Program in Computer Engineering. AASS contributions are in terms of management,
administration, program and course development, and teaching.
Silvia Coradeschi, AASS, is the director of studies for the Computer Engineering Division
Including both the 3 and 4 years programs. Her responsibilities include budget and staff
management for the undergraduate education within the division. Lars Karlsson is the
coordinator for the MSc program. His responsibilities include development, information,
course scheduling, and thesis administration.
1.4.1 Computer engineering program
The Computer Engineering Program is a study program consisting of 120 credit points
(corresponding to three years of studies). The program can be characterized as a fixed study
program with some options during year 3. It leads to a University Diploma in Engineering
(sv. “högskoleingenjörsexamen”), 120 points, and/or a bachelor degree.
Details about the program's study plan can be found at:
www.tech.oru.se/net/Teknat/Huvudsida/Programsidor/Dataprogrammet
The following courses were taught in 2002 by AASS researchers and Phd students:
• Artificial Neural Networks, 5 points. Teacher: Tom Duckett. Assistant: Malin Lindqvist.
Introductory course in artificial neural networks and their applications. Given in spring,
for 3 rd year students.
• Artificial Intelligence, 5 points. Teacher: Silvia Coradeschi. Assistants: Amy Loutfi.
Introductory course in artificial intelligence and LISP. Given in autumn for 2 nd -year
students.
• Real-time programming, 5 points. Teacher: Lars Karlsson. Assistants: Pär Buschka and
Malin Lindquist. Given in autumn for 3 rd -year students. A course in programming and
analysis of real-time and embedded systems.
• Modeling and simulation of dynamic systems, 5 points. Teacher: Selim Eskiizmirliler.
Given in Spring for 3 rd -year students.
• Java Programming, 5 points. Assistant: Zbigniew Wasik. Given in autumn for 3 rd -year
students.
• Computer Architecture, 5 points. Assistant: Pär Buschka and Boyko Iliev. Given in
Autumn for 2 rd -year students.
• Computer Communication, 5 points. Assistant: Amy Loutfi. Given in Spring for 2 rd -year
students.
• Programming methodology, 5 points. Assistant: Grzegorz Cielniak. Given in Spring for
2 rd -year students.
• Databases, 5 points. Assistant: Robert Johansson. Given in Autumn for 3 rd -year students.
• Program development and project managing, 5 points. Assistant: Robert Johansson.
Given in Autumn for 3 rd -year students.

Annual Report 2002 15
BSc Theses
• Three 10-point BSc theses were completed under the supervision of Tom Duckett.
• One 10-point BSc thesis was completed under the supervision of Amy Loutfi.
• One 10-point BSc thesis was completed under the supervision of Alessandro Saffiotti.
• Five 10-point BSc theses were completed under the supervision of Lars Karlsson.
• One 10-point BSc thesis was completed under the supervision of Ivan Kalaykov.
1.4.2 MSc in computer engineering
The MSc Computer Engineering Program is a 4-year study program leading to a Master of
Science with a major in Computer Engineering. The first 2 years of the program are in
common with the 3-year Computer Engineering program, and students choose between the
two programs at the end of the second year. The program belongs to the technological study
sector. The program was started in the autumn of 2001 and the first class is to be graduated in
2002. Currently there are 49 students enrolled in the program.
Details about the program's study plan can be found at:
http://www.tech.oru.se/net/Teknat/Huvudsida/Programsidor/Magister+Data
During the year, AASS members and PhD students have been involved in teaching related to
the program as follows:
• Numerical Methods, 5 points. Teacher: Li Jun. Given in autumn for 3 rd -year students. An
introductory course in numerical methods.
• Pattern Matching, 10 points. Teacher: Gustav Tolt. Given in autumn for 4 rd -year students.
MSc Theses
The following MSc theses in Computer Engineering were completed during year 2002.
10 points
• Eriksson, Fredrik
Utveckling av Windowsapplikation för diagramritning och utskrift
Yrkes- och miljömedicinska kliniken, USÖ.
• Göransson, Therése och Sundgren, Jack
Framtagning för utvärderingsverktyg för datainsamlingssystem
Aerotech Telub AB, Arboga
• Larsson, John och Likidis, Petros
Processkomponent i Java för DocTrade SHS
Statistiska Centralbyrån, Örebro
• Uppgård, Thomas
Utveckling av lagerhanteringsprogrammet LagerMan
Industrisystem AB, Karlskoga

16 AASS – Center for Applied Autonomous Sensor Systems
• Ewerö, Magnus
Utredning av OpenGL i QT och Visual C++
Aerotech Telub AB, Arboga
20 points
• Farrah, Ahmed
Reactive Localisation of an Odour Source by a Learning Mobile Robot
AASS, Örebro universitet. Supervisor: Tom Duckett
• Johansson, Robert
A Graphical Development Platform for Team of Robots
AASS, Örebro universitet. Supervisor: Alessandro Saffiotti
An additional two theses on 10 points each (one at AASS and one at the electronics group at
Örebro university) were started in Spring 2002 but have still not been completed.
1.4.3 Other study programs
AASS members and PhD students were also involved in a number of courses for other study
programs:
-Measurement technology, 5 points. Teacher: Peter Wide. Assistants: Malin Lindquist and
Amy Loutfi. Program: Physics.
-Robot technology, 5 points. Development: Anani Ananiev. Programs: Distance course for the
Automation Engineering program. An introductory course in robot technology.
-Computer-integrated manufacturing systems, 5 points. Teacher: Ivan Kalaykov. Assistant:
Grzegorz Cielniak. Program: Automation Engineering.

Annual Report 2002 17
1.5 Graduate Education
An integrated vital part of research at AASS is its Graduate Program and perhaps the most
important "output" of AASS is highly qualified PhD's and Licentiates whose contributions
provide further strength to Swedish industry and academic community.
AASS is performing interdisciplinary research in the field of autonomous embedded systems,
integrating theories, technologies and tools from sensors and measurement technology,
control, artificial intelligence and computer science. Therefore, the AASS graduate program
aims at giving graduate students a strong interdisciplinary background in these different areas
with emphasis on topics relevant to autonomous systems. It is assumed that the students
already have adequate knowledge in the basics of engineering and/or computer science.
Finally, the program prepares students for a career in academia or industry by including topics
such as pedagogy, management and presentation techniques.
60 study points are required for a PhD. Of these, 25 points (or more) should be from a core
curriculum courses, an additional 25 points should be from other courses or self studies in
research relevant topics, and 10 points can be chosen freely. The Director of the Graduate
program is Lars Karlsson, PhD.
1.5.1 Core curriculum
There are five courses that are given on a regular basis. These courses constitute the AASS
core curriculum. They cover the different fields that are researched in AASS. In order to
ensure that students acquire the interdisciplinary knowledge that characterizes AASS, the core
courses are compulsory. The core courses are as follows.
• AI for embedded systems. 5 points + 3 optional points for advanced studies. Teacher:
Alessandro Saffiotti, Mobile Robotics Lab.
• Soft computing for control. 5 points + 3 optional points for advanced studies. Teacher:
Ivan Kalaykov, Intelligent Control Lab.
• Embodied learning. 5 points + 3 optional points for advanced studies. Teacher: Tom
Duckett, Learning Systems Lab.
• Biology inspired control systems. 5 points + 3 optional points for advanced studies.
Teacher: Selim Eskiizmirliler, Sensors and Measurements Lab.
• Research Methodology for Computer Science and Engineering, 5 points, Teacher: Jan
Gustafsson, Mälardalen University.
Of these courses, AI for mobile systems was given in spring 2002.
1.5.2 Occasional courses
The core courses are complemented by courses that are provided on a more occasional basis -
on request - when a visiting researcher provides for such an opportunity, or when a subject of
interest is covered by a graduate course at another university. During 2002, no such courses
were given.

18 AASS – Center for Applied Autonomous Sensor Systems
1.5.3 Self-study course work
This type of course work is done within supervised self-study circles (smaller courses), or
individual studies.
1.5.4 Cooperation with other graduate programs
As recognition of the research effort at AASS, Örebro University has become one of the
participants in CUGS -- the Swedish National Graduate School in Computer Science. CUGS
is commissioned by the Swedish government and the Swedish Board of Education and Lars
Karlsson is a member of the CUGS steering Committee.
Because of CUGS, AASS students have now the additional possibility to get enrolled as
CUGS students or to follow some courses of interest provided by the CUGS study program.
Currently, two students from the Mobile Robotics Laboratory are enrolled in CUGS,
receiving a total funding of 700' SEK:
• Robert Johansson, with topic cooperative robotics and supervisors Alessandro Saffiotti
and Lars Karlsson
• Kevin LeBlanc, with topic anchoring and supervisors Alessandro Saffiotti and Silvia
Coradeschi.
The scientific scope of CUGS includes central parts of the core computer science
(sv. datalogi) and computer engineering (sv. datorsystem). In this regard, CUGS puts an
emphasis on programming languages, algorithms, software engineering, also including related
areas of autonomous systems, real-time systems, embedded systems, knowledge-based
systems and artificial intelligence. Other members of CUGS are the universities/university
colleges in Linköping (host), Lund, Mälardalen, and Skövde.
Detailed information about CUGS can be found at:
http://www.ida.liu.se: 9080/groups/cugs/FrontPage.
1.5.5 European graduate programs
As a recognition for the quality of its graduate education and research
The European Commission has granted AASS the status of a Marie Curie Training
Site in the area of Autonomous Robotics. The Commission, during the next four years, will
provide the funding necessary for accommodating PhD students from European countries for
periods of stay between 3 and 12 months. During their stay at AASS the guest-students will
participate in projects related to AASS research and also take courses offered within the
AASS Graduate Program. Prof. D. Driankov is the responsible coordinator for the program.
During 2002 we have had 2 Marie Curie Fellows:
• Marcel Oosterom, Delft Inst. of Technology. Topic: Fuzzy gain-scheduled control;
Duration: 3 months.
• Achim Lilienthal, German Aerospace Center, Inst. of Robotics and Mechatronics. Topic:
Simultaneous localization and mapping; Duration: 6 months

Annual Report 2002 19
In 2003 we expect 4 more Fellows to join our Marie Curie Training Site.
More information regarding the Marie Curie Training Programes can be found at:
http://improving.cordis.lu/mc
1.5.6 Graduate students
Biologically Inspired Systems Lab
PhD-student First advisor Second
advisor
Employed at
Biel, Lena Peter Wide Dimiter Bofors
Driankov Defence AB
Lindquist, Peter Wide Selim ÖU
Malin
Eskiizmirliler
Loutfi, Amy Peter Wide Silvia ÖU
Coradeschi
Robertsson, Peter Wide Selim ÖU
Linn
Eskiizmirliler
Spännar , Jan Peter Wide George Fodor Högskolan
Dalarna
Enrolled at
ÖU
ÖU
ÖU
ÖU
ÖU
Mobile Robotics Lab
PhD-student First advisor Second Employed at Enrolled at
advisor
Bouguerra, Alessandro Lars Karlsson ÖU ÖU
Abdelbaki Saffiotti
Buschka, Pär Alessandro Tom Duckett ÖU ÖU
Saffiotti
Johansson, Alessandro Lars Karlsson ÖU ÖU and
Robert Saffiotti
CUGS
LeBlanc, Alessandro Silvia ÖU ÖU and
Kevin Saffiotti Coradeschi
CUGS
Pettersson, Alessandro Lars Karlsson ÖU ÖU
Ola Saffiotti
Wasik,
Zbigniew
Alessandro
Saffiotti
Dimiter
Driankov
ÖU ÖU

20 AASS – Center for Applied Autonomous Sensor Systems
Intelligent Control Lab
PhD-student First advisor Second Employed at Enrolled at
advisor
Iliev, Boyko Ivan Dimiter ÖU ÖU
Kalaykov Driankov
Jennergren, Ivan Anani ÖU ÖU
Lars Kalaykov Ananiev
Lindkvist, George Fodor Peter Wide ÖU ÖU
Christer
Persson, Dimiter Ivan AeroTech ÖU
Martin Driankov Kalaykov Telub AB
Tolt, Gustav Ivan Dimiter ÖU ÖU
Kalaykov Driankov
Skoglund,
Alexander
*(25%)
Ivan
Kalaykov
Tom Duckett ÖU ÖU
Bagnoli,
Letizia*
Ivan
Kalaykov
Learning Systems Lab
Anani
Ananiev
PhD-student First advisor Second Employed at Enrolled at
advisor
Andreasson, Alessandro Tom Duckett ÖU ÖU
Henrik Saffiotti
Cielniak, Peter Wide Tom Duckett ÖU ÖU
Grzegorz
Jun, Li Peter Wide Tom Duckett ÖU ÖU
Skoglund, Ivan Tom Duckett ÖU ÖU
Alexander * Kalaykov
(25%)
Munkevik,
Per
Peter Wide Tom Duckett ÖU ÖU
* Officially registered as PhD students in 2003
1.5.7 PhD theses
No PhD theses were completed during 2002.
1.5.8 Licentiate theses
During 2002 the following licentiate theses were completed:
• Boyko, Iliev. Minimum-time Sliding mode control of robot manipulators.
• Spännar, Jan. Grey box modelling for temperature estimation.
• Persson, Martin. A simulation environment for visual servoing.
• Biel, Lena. Modelling of perceptual systems - a sensor fusion model with active
perception.
ÖU
ÖU

Annual Report 2002 21
1.6 Co-operation
The AASS research effort builds upon two types of co-operation: industrial and academic.
The latter one has both national and international dimension.
1.6.1 Industrial cooperation
• ABB Robotics, Västerås - industrial robotics; development of new specific constructions
for industrial robots; contract for co-operation (KK);
• ABB Automation Products AB, Dept. AMC, Västerås - advanced industrial control
systems, development; test and implementation of model-based fuzzy control techniques;
feasibility study on vision-based quality control of cold rolling mills production; contract
for co-operation (KK);
• FFV Aerotech Telub AB, Örebro - visual servoing techniques for airborne platforms;
contract for co-operation (KK);
• HARDI Electronics AB, Lund - design of electronic circuits; joint FPGA-based design
and implementation of high-speed image and signal processing; contract for co-operation
(KK) being discussed.
• Bofors Defence AB, Karlskoga, Sweden - perceptual sensor fusion architectures and
applications; one joint PhD student; contract for co-operation (KK)
• Amphitech AB, Askersund, Sweden - on-line quality measurement of drinking water.
• Elumotion Ltd, Bath, UK - development of an artificial sensory hand.
• Cerealia R&D, Järna, Sweden - automation in food manufacturing.
• NIROS, Karlskoga, Sweden - using autonomous systems for rescue operations: a
preliminary study.
1.6.2 National Cooperation
1.6.2.1 Scientific cooperation
- KTH, Center for Autonomous Systems: joint organization of the 2nd Swedish Workshop
on Autonomous Robotics
- Joint research project on medical image processing with University Hospital, Örebro, on
detection and diagnosis of abnormalities in human breast tissue.
- Joint application to the VINNOVA VINVÄXT Program with MdH, KTH, and 20 other
industrial companies: winner of the first phase of the competition.

22 AASS – Center for Applied Autonomous Sensor Systems
- WITAS, Univ. of Linköping, Dept. of Computer Science - development of Takagi-Sugeno
fuzzy model based observers and controllers for unmanned aerial vechicles.
- SIK, Göteborg - joint project on Future Food Factory (FFF), one joint PhD student;
several visits to/from SIK during 2002;
- Computer vision laboratory, Linköping University, Dept. of Electrical Engineering -
exchange of visits, identification of future joint research project.
1.6.2.2 National networks and societies
- Coordinator of Team Sweden (team of 3 univ: Lund Univ, Blekinge Institute of
Technology, Orebro Univ)
- Member of CUGS, participating with two PhD students
- Member of SLSS (Swedish Learning Systems Society). Co-organiser of the joint
SAIS/SLSS meeting in 2003.
- Member of the Swedish IEEE chapter on Signal Processing.
1.6.3 International Cooperation
1.6.3.1 EC Networks/Programes
- The European Community Network of Excellence on Planning (PLANET);
- The European Community Network of Excellence on Robotics Research (EURON) --
Cooperative Robotics coordinator; and
- The European Community Thematic Network on Walking and Climbing Robots
(CLAWAR).
- Marie Curie Training Site (coordinator)
1.6.3.2 Participation in scientific events
- Joint organization with the Technical University of Lisbon of the first European Summer
School on Cooperative Robotics (Lisbon, Portugal, Sept 2-7, 2002). Chairs: I. Ribeiro
and A. Saffiotti. Attended by 42 students and sponsored by the EURON EC network.
- Organization of the IROS Workshop on cooperative robotics (Lausanne, CH, Oct 1,
2002). Chair: A. Saffiotti. Attended by 50 people and Sponsored by the EURON EC
network.
- Joint organization, with the Ecole Nat. Sup. de Telecomunications (Paris, France), of an
invited session on "Uncertainty in spatial representations" at the Int. Conf. on Information

Annual Report 2002 23
Processing and Management of Uncertainty (Annecy, France, Jul 1-5, 2002). Chairs: I.
Bloch and A. Saffiotti. Attended by 30 people.
- Participation in the above Summer School by two PhD students (R. Johansson and K.
Leblanc).
- Participation in the Int. Conf. on Intelligent Robotic Systems (IROS, Lausanne, CH, Oct
1-4, 2002) by A. Saffiotti, P. Buschka and Z. Wasik. Presentation of two papers at that
conference.
- Participation in the Int. Robocup Symposium (Fukuoka, Japan, Jun 24-25, 2002) by A.
Saffiotti.
- Participation in the Int. Conf. on Pattern Recognition (Quebec City, Canada, Aug 11-15)
by Z. Wasik. Presentation of one paper at the conference.
- Participation in the European Summer School on Visual Servoing (Beniccasim, Spain,
September 16-20) by Z. Wasik.
- 2nd Swedish workshop on Autonomous Robotics (3 papers presented)
- 2nd European Medical and Biological Conference (1 paper presented)
- AFSS International Conference on Fuzzy Systems, Calcutta, India, February 3-6, 2002 -
G. Tolt presented a paper.
- IEEE International Conference on Fuzzy Systems FUZZ-IEEE'2002, May 12-17, 2002,
Honolulu, Hawaii, USA - B. Iliev presented a paper.
- NAFIPS-FLINT 2002 Conference of the North American Fuzzy Information Processing
Society, New Orleans, USA, June 27-29 - I. Kalaykov presented a paper.
- Third International NAISO Symposium on Engineering of Intelligent Systems, Malaga,
Spain, September 24-27 - I. Kalaykov presented a paper.
- 33rd International Symposium on Robotics (ISR'2002), Stockholm, Sweden, October 7-
11, 2002 - A. Ananiev, B. Iliev, I. Kalaykov, L. Jennergren presented a paper.
- Third IEEE International Workshop on Robot Motion and Control, RoMoCo'02,
November 9-11, Bukowy Dworek, Poland - B. Iliev presented a paper.
- Third SPIE International Conf. on Optomechatronics Systems III, Stuttgart, Germany,
November 12-14 - I. Kalaykov presented a paper.
- TOK02, Turkish Automatic Control Congress, Ankara, 9-11 September 2002
- CLAWAR meeting, Paris, 24 September, 2002

24 AASS – Center for Applied Autonomous Sensor Systems
- CLAWAR 2002 Conference, Climbing and Walking Robots, Paris, 25-27 September,
2002
- Concerted Action of CNRS on Neuronal Plasticity and Computation, Annual meeting,
Paris, 28-29 November 2002
- SPIE 2002, Sensor fusion: Architectures algorithms and applications IV, Orlando, USA –
Session Chairman P.Wide.
- IEEE Virtual Instrumentation and Measurement Systems. Anchorage, Alaska, USA, 2002
– paper presented by A.Loutfi.
- IPMU 2002 Information Processing and Management of Uncertainty in knowledge based
systems, Annecy, France, July 2002 – paper presented by P. Wide.
- Artificial Intelligence Research Symposium. Lyon, France 2002 – paper presented by
A.Loutfi.
- Sinnena, Konsten och Vetenskapen Del 2, Örebro, 16-17 January 2002 – paper presented
by A. Loutfi.
- IMTC02 – 19th IEEE Instrumentation and Measurement Technology Conference, Alaska,
USA, May 2002 – paper presented by A. Loutfi.
- HAVE 2002 – IEEE International workshop on haptic virtual environments and their
applications, Ottawa, Canada, Nov 2002 – paper presented by P. Wide.
1.6.3.3 Scientific cooperation
Visiting researchers/students to AASS
- Humberto Martinez (Univ of Murcia, Spain) from May 29 to June 28, 2002. Joint research
on cooperative robotics in the framework of RoboCup.
- Achim Lilienthal, June-Dec 2002. Visiting Ph.D. student from Dept. of Computer
Science, Tuebingen University. Work on navigation by olfaction project, Marie Curie
fellow.
- Assoc. Prof. Dmitry Kriwetz, Rostov State University, Research Institute on
Neurocybernetics, Rostov on Don, Russia. Work on force/torque sensory feedback in
robotic systems.
- Prof. Gancho Vachkov, Kagawa University, Japan. Work on joint publication and seminar
at AASS.
- Dr. Rainer Palm, Siemens Corporate Research & Development, Munchen, Germany.
Seminar and joint work on robot arm control.

Annual Report 2002 25
- Jörgen Widmark, Msc student in Linköping University. Work on facial expression
association to the e-nose output.
AASS visits to other sites
- A. Saffiotti visiting Univ of Murcia, Spain, on March 13-18, 2002, for joint research
project on cooperative robotics.
- A. Saffiotti visiting Univ of Palermo, Italy, from Apr 29 to May 3, 2002, for joint research
project on anchoring.
- A. Saffiotti visiting the Ecole Norm. Sup. de Telecommunications (Paris, France) on July
8-12, 2002, for joint work on fuzzy mathematical morphology for robotic applications.
- A. Saffiotti visiting Technical Univ of Lisbon, Portugal, on Aug 26-30, 2002, for
organization of summer school on cooperative robotics.
- A. Saffiotti visiting Univ of Torino, Italy, on Sep 23-27 for scientific exchange.
- A. Saffiotti visiting Univ of Murcia, Spain, on Nov 4-29, 2002, for joint research project
on architectures for autonomous robots.
- Grzegorz Cielniak, Oct. 2002- Jan. 2003 AASS Ph.D. student visited the AI Lab of
Freiburg University. Work on human identification and tracking project as a Marie Curie
fellow.
- CeBIT '2002 International exhibition on IT, Hanover, Germany. Exhibiting fast vision and
image processing system prototype - G. Tolt, I. Kalaykov, P. Wide;
- NTNU, Trondheim, Norway, Department of Production and Quality Engineering - May
2002. PhD thesis defense and identification of joint research topics - I. Kalaykov;
- Delft University of Technology, Department of Electrical Engineering, Control
Laboratory - December 2002. PhD thesis defence, invited talk - I. Kalaykov;
Joint work/projects
- Joint work on "Anchoring" with the Univ of Palermo, Italy.
- Joint work on "cooperative robotics" with the Univ of Murcia, Spain.
- Joint work on "fuzzy mathematical morphology for robotic applications" with the Ecole
Norm. Sup. de Telecommunications of Paris, France.
- "Thinking Cap": joint research project on architectures for autonomous robots with the
Univ of Murcia, Spain.

26 AASS – Center for Applied Autonomous Sensor Systems
- Fast, online mapping by mobile robots with Dr. Jonathan Shapiro and Dr. Stephen
Marsland, Artificial Intelligence Group, Manchester University, UK.
- Robot navigation by olfaction with Achim Lilienthal (supervisor: Prof. Andreas Zell),
Dept. of Computer Science, Tuebingen University, Germany.
- Human identification and tracking with Prof. Wolfram Burgard, AI Lab, Freiburg
University, Germany.
- Segmentation of liver images with Prof. Leonard Bochi, University of Florence, Italy.
- Lightweight robot arms with Bulgarian Academy of Sciences, Institute of Mechanics, Dr.
Detelina Ignatova;
- Minimum-time SMC of robot manipulators with Technical University Sofia, Department
of Control and Systems Engineering, Dr. Kamen Perev;
- Model-based fuzzy control with Siemens Corporate Research and Development,
Germany, Dr. Rainer Palm
- Reconfigurable Control Architectures with Western Michigan University, USA, George
Fodor.
- Sensory-Motor Information Fusion with ENST, Ecole Nationale Superieure des
Telecommunications, Paris, Dept. of Image & Signal Processing, CNRS URA 820,
C.Darlot.
- EMG based Analysis of human forearm movements with University of Bourgogne,
INSERM, Lab of Movement Analysis, T. Pozzo, Ch. Papaxanthis.
- Cerebellar like motor control of an anthropomorphic robot arm with INSA, Institut
National des Sciences Appliquees, Toulouse, Dept. of Electrical & Computer
Engineering, B.Tondu .
- Sensory Information Fusion, Study on Motion Sickness in High Speed Tilting Trains with
University of Pierre Mendès, Grenoble, CNRS, UMR 5105, Experimental Psycolgy Lab.,
J. Neimer, T. Ohlmann.
- Motion Control of a 6 d.o.f. robot arm (Bridgeston arm) actuated by McKibben artificial
muscles with Bosphorus University, Istanbul, Dept. of Electrical & Electronics
Engineering, Biomedical Engineering Institute, Y. Istefanopulos.
- Design and development of a five fingers artificial hand with Bath University, Dept. of
Mechanical Engineering, Graham Whiteley.
- Biologically inspired motion control of an artificial hand with Paris 6, University of Pierre
et Marie Curie, INSERM-U483 Laboratory, M. Maier.
- Electro Active Polymer Actuators and Devices with NASA-JPL, Dr. Yoseph Bar Cohen

Annual Report 2002 27
- Temperature estimation with Högskolan Dalarna, Björn Sohlberg
- Virtual Social Intelligent Agents with Linköping University, Mark Olilla
- Intelligent Systems with Politecnico di Milano, Department of Electronics and
Information, Prof. Vincenzo Piuri, Milano, Italy
- Sensory Information Fusion with School of Information Technology and Engineering
(SITE) University of Ottawa Prof. Emil Petriu, Dr. Eng., P.Eng., Canada
- Robot measurement with Prof. Mel Siegel, Robotic Institute, Carnegie Mellon University,
USA.
1.6.4 Services to the research community
1.6.4.1 Membership in professional organizations/societies/management-bodies
- The Board of Trustees of the RoboCup International Federation
- The Advisory Board of the International Conference on AI (ICAI-2001)
- The European Coordinating Committee for Artificial Intelligence (ECCAI)
- The Scandinavian RoboCup Committee
- Membership in IEEE Computer Soc.; the IEEE Robotics and Automation Soc.; the IEEE
System, Man and Cybernetics Soc.; the American Assoc. for Artificial Intelligence; the
Intelligent Autonomous Systems Soc.; IEEE CSS/SP societies; SPIE and EUSFLAT;
AAAI (American National Association for Artificial Intelligence); SLSS member
(Swedish Learning Systems Society); EvoNet - The Network of Excellence in
Evolutionary Computing; IAPR - International Association for Pattern Recognition; IEEE
- Institute of Electrical and Electronics Engineers; SSAB - Swedish Society for
Automated Image Analysis.
- The Swedish Learning Systems Group
- The Board of SWIRA, Swedish Industrial Robot Association.
- The Örebro University board.
- The Faculty board for Natural Science, Technology and Medicine, Örebro University
- IEEE Technical Committee on Intelligent Control, Section: Fuzzy Control
- Editorial Board International Journal of Uncertainty, Fuzziness, and Knowledge-Based
Systems, World Scientific Publ.

28 AASS – Center for Applied Autonomous Sensor Systems
- Advisory Board of the Electrical and Computer Science Department, Western Michigan
University, USA.
- IEEE Instrumentation & Measurement society – Chairman Technical Committee TC 27.
1.6.4.2 Organization of scientific events
- Organization of the IROS Workshop on cooperative robotics (Lausanne, CH, Oct 1,
2002). Chair: A. Saffiotti. Attended by 50 people and Sponsored by the EURON EC
network.
- Joint organization, with the Ecole Nat. Sup. de Telecomunications (Paris, France), of an
invited session on "Uncertainty in spatial representations" at the Int. Conf. on Information
Processing and Management of Uncertainty (Annecy, France, Jul 1-5, 2002). Chairs:
Bloch and A. Saffiotti. Attended by 30 people.
- Joint organization, with KTH, Stockholm, Sweden, of the 2nd Swedish Workshop on
Autonomous Robotics (Stockholm, Oct 10-11, 2002). Chairs: Saffiotti, T. Duckett, H.
Christensen. Attended by 40 people.
- NAISO Conference on Neuro-Fuzzy Systems, February 2002, Havana, Cuba - member of
IPC I. Kalaykov; NAISO Conference on Engineering of Intelligent Systems, September
2002, Malaga, Spain – member of IPC Ivan Kalaykov.
- SPIE, Sensor fusion: Architectures algorithms and Applications IV, Orlando, USA, April
2002 – member of IPC P.Wide.
- IMTC02 – 19 th IEEE Instrumentation and Measurement Technology Conference, Alaska,
USA, May 2002 – member of TPC P.Wide.
1.6.4.3 Organization/participation of educational events
- Joint organization with the Technical University of Lisbon of the first European Summer
School on Cooperative Robotics (Lisbon, Portugal, Sept 2-7, 2002). Chairs: I. Ribeiro
and A. Saffiotti. Attended by 42 students and sponsored by the EURON EC network.
- Participation in the above Summer School by two PhD students (R. Johansson and K.
Leblanc).
- Participation in the European Summer School on Visual Servoing (Beniccasim, Spain,
September 16-20) by Z. Wasik.
1.6.4.4 Invited talks
- A. Saffiotti on "Team Sweden" at the Spanish Workshop on Physical Agents (Murcia,
Spain, March 14-15, 2002).

Annual Report 2002 29
- A. Saffiotti on "Perceptual Anchoring" at the Univ of Palermo, Italy, on April 29, 2002.
- A. Saffiotti on "Trends and fashion of autonomous robots" at the Univ of Torino, Italy, on
Sept 23, 2002.
- A. Saffiotti on "Perceptual Anchoring" at the Univ of Torino, Italy, on Sept 24, 2002.
- A. Saffiotti on "Perceptual Anchoring" at the Univ of Padova, Italy, on Oct 7, 2002.
- Ivan kalaykov on “Fast fuzzy hardware - it's so simple!’’ - plenary talk at NAISO Conf.
on Engineering of Intelligent Systems, September 2002, Malaga, Spain;
- Ivan Kalaykov on “Minimum-time SMC of robot manipulators’’ - invited talk at the
Control Laboratory, Delft University of Technology, December, 2002;
- S. Eskiizmirliler on “Biologically Inspired Control systems” – Invited talk at TOK02-
Turkish Automatic Control Congress, Middle East Technical University. Ankara, 9-11
September, 2002
1.6.4.5 Evaluation activities
- Hirpa L. Gelgele - Hybrid Intelligent Systems in Manufacturing Optimization; PhD thesis
examiner at NTNU, Trondheim, Norway, Ivan Kalaykov
- Stanimir St. Mollov - Fuzzy Control of Multiple-Input Multiple-Output Processes, Delft
University of Technology, Delft, Netherlands. PhD thesis examiner Ivan Kalaykov.
- Teodor Sundies - Data Processing Methods for Smart Instruments based on Chemical
Sensor Arrays; Phd thesis examiner at Sistemes d’Instrumentaciõ i Comunicacions,
Department d’Electrõnica, Universitat de Barcelona, Spain.
1.6.4.6 Reviewing activities
- International journals 2002: Artificial Intelligence; Robotics and Autonomous Systems;
IEEE Transactions on System Man and Cybernetics; IEEE Transactions on Robotics and
Automation; Robotics and Autonomous Systems; Autonomous Robots; Journal of
Robotic Systems; IEEE Transaction on Industrial Electronics; IEEE Transaction on
Instrumentation and Measurement; IEEE Transactions on Fuzzy systems; Automatica;
Int. J. Fuzzy Sets and Systems; J. Applied Intelligence; IEEE Trans. on Automatic
Control; Int. J. of Adaptive Control and Signal Processing; Journal of food engineering;
Institute of Physics Publishing;
- International conferences 2002: IEEE Int Conf on Fuzzy Systems; Int RoboCup
Symposium; European Conf on Artificial Intelligence; Int Conf on Information Processing
and the Management of Uncertainty; Int Conf on Uncertainty in Artificial Intelligence; Int
Conf on Intelligent Robotic Systems; Second Swedish Workshop on Autonomous
Robotics; Neuro-Fuzzy and NAISO Engineering of Intelligent Systems; IEEE_FUZZ; 5 th

30 AASS – Center for Applied Autonomous Sensor Systems
international conference on information fusion, FUSION 2002; SPIE IEEE
Instrumentation and Measurement Conference.
1.7 Dissimination of Information
The dissimination of information pertaining to AASS activities has the overriding goal to
establish a firm and clear identity and acceptance of the AASS research profile, its industrial
relevance, and societal contributions.
To achieve the above goal our dissimination of information activities are towards the
following three target audiences:
A. Industry, public authorities, and research funding bodies;
B. Research community and university students; and
C. General public.
Further information about our goals and philosophy regarding dissemination of information
you can find in our Information Policy for 2003.
1.7.1 Activities during March 2002 – February 2003
Activity: Time: Performance: Target
group:
Fairs:
CEBIT 2002 02-03-02 Monter A,B and C
Tekniska Mässan,
Stockholm, Sweden
8-12/10-02 Monter A,B and C
Utbildningsmässan,
Örebro, Sweden
6-7/11-02 Presentation of AASS, by Henrik
Andreasson
B and C
EC Conference,
Brussels, Belgium
11-13/11-
02
Represented in CLAWAR- Networks
monter
A,B and C
Visits:
10 visits of different
representatives from
industry
Presented the laboratories and the AASS’
research
A
Nordisk Trådteknisk
Förening
Tullängsskolan,
teachers
23-08-02 Presented labs for 86 pers. A
06-11-02 Presented labs and research at AASS B and C
Visit from VIPE, 16-11-02 Presentation of AASS A

Annual Report 2002 31
Vanne, France
French attaché for 24/02-03 Presentation of AASS A
Culture and Science
Media:
Teknik
Vetenskap
och
Nr2/2002 “Avfärda aldrig det enkla! A,B and C
www.KKS.se 16/10-02 “Teknikhus med rum för framtiden” –
Article from “open house” and inauguration
of new building. 500 visitors
www.KKS.se 15/10-02 “På Örebro universitet formas framtiden”
- Article about AASS.
A, B and C
A and B
Radio P4, “På
jobbet”
16/1-03 Interview of Phd-student Ola Pettersson
about AASS
C
Nerikes Allehanda 25/9-02 “Robotar skruvas ihop i nytt teknikhus".
4 th graders build robots together with
researchers and look at demos in lab.
TV4 -Nyheter 24/9-02 Showed robot arms, interviewed Phdstudent
Lars Jennergren and children
building robots.
Örebro Kuriren 09/11-02 Article and picture from the price award
ceremony with Atlas Copco Rock Drills
C
C
C
Nerikes Allehanda 08/11-02 “Teknik studenter prisas” C
Other
Presentations:
Sinnena och Konsten Jan. –02
Phd.-student Amy Loutfi presents her
research work and AASS
B and C
RoboCup June –02 Competition with robot dogs playing
football
B and C
Inauguration and
Open house, Örebro
University
24/9-02 Inauguration and “open house” with demos
in AASS lab, 86 4 th graders building robots.
500 other visitors.
A, B and C
Price
ceremony
award
8/11-02 Atlas Copco Rock Drills gave the award to
two of our ex-job students in Learning lab.
A, B and C
Presentation
students
for
26/11-02 Presentation of AASS and possible ex-jobs
for master students.
B

32 AASS – Center for Applied Autonomous Sensor Systems
“Inspirationsdag” 28/11-02 Industry, research funding bodies and
public authorities were invited to see the
result of earlier contributions as well as get
inspired to continued contribution.
“Landshövding” Gerd Engman was one of
many interested visitors in our lab.
B and C
Räddningsverket
Nora
i
Dec. –02 Presented the outdoor robot and AASS A
Referensgruppsmöte 5-6/12-02 Discussion about AASS and future
Sinnena och Konsten Jan. –03
Phd.-student Amy Loutfi presents her
research work and AASS
B and C
1.7.2 Industry, public authorities, and research funding bodies
Well-established means for disseminating information about AASS research results and their
industrial potential amongst these parts of the Swedish industry that can directly/indirectly
benefit from it are:
- Joint projects: See section 1.6.
- Industrial PhD students: See section 1.5.6
- AASS Seminars: See section 4.2
- Industrial fairs and exhibitions: See table above 1.7.1
- Industrial associations and interest groups
Swira, Cerealia, IRS (Intelligent Rescure Systems), SIK, OIC (Orebro Innovation Center).
Funding agencies are informed about AASS activities via project proposals, activity-reports
such as the current one, and by using some of the dissemination means already described.
1.7.3. Research Community and university students
Disseminating information about AASS research profile and its educational role and potential
is done via the traditional dissemination channels, such as:
- Publications: 44
- Conference participation: 29
- Organization of scientific events: 6
- AASS guest-researchers, guest-students, and visits: 18
- National/international networks: 4
- Joint research projects: 25

Annual Report 2002 33
1.7.4 General Public
The general public and various civil authorities are informed via the web, public lectures, and
media presence such as newspaper articles, TV and radio programs. We also arranged several
“open houses” with demos for the general public as well as public seminars that we have
regularly see 4.2
1.7.5 Seminars
1.7.5.1 AASS International Seminar Series
During 2002 the event didn’t take place.
1.7.5.2 Regular AASS Seminars
Schedule for April 2002 - January 2003
Apr 8 Ola Pettersson Steps towards model-free execution monitoring
Apr 15 Par Buschka A virtual sensor for room detection
Apr 22 Xjob presentations
Ahmed Mahamod Farah: Active localization of an
odour source Linus Göransson: Dead reckoning
using omni-directional vision
Apr 29
AASS coordinators and lab
leaders
Future directions for AASS
May 6 Boyko Iliev
Variable structure control using Takagi-Sugeno
system as a sliding surface
May 13 Xjob presentation
Robert Johansson: Graphical development
platform for RoboCup
June 3 Xjob presentations
Daniel & Mihaljo - People Tracking
Karin & Catharina - Medical Image Processing
Nov 4 Jörgen Widmark
VR social agent expressions: driven by real
electonic sensors
Nov 11 Lucia Ballerini Image analysis for food quality measurements
Nov 18 Zbigniew Wasik
Reports: visual servoing summer school and ICPR
conference
Dec 2
Par Buschka, Alessandro
Saffiotti, Zbigniew Wasik
Report: IROS 2002 Conference
Dec 9
Kevin
LeBlanc
Visual Tracking Algorithms
Jan 7 Henrik Andreasson Laser ranger for underwater vehicles
Jan 20 Lars Jennergren
Assembly of spherical roller bearings (MSc thesis)
plus goals of his Ph.D project "Flexible food
production"

34 AASS – Center for Applied Autonomous Sensor Systems
2. Research Labs
At the present time AASS consists of four research laboratories that have joined their
expertise to cover and integrate the broad spectrum of scientific disciplines needed for the
design and implementation of industrially relevant autonomous systems.
This would be practically impossible for any individual laboratory, but is achievable by a
joint effort, which is organized in the form of AASS. This joint effort is aimed not only at
expanding the frontiers of the science of autonomous sensor systems, but also gradually
incorporating it in the graduate and undergraduate curriculum, and producing physical
demonstrators of autonomous sensor systems with industrial and educational relevance.
The common scientific denominator for the four research laboratories is the subject of
"perception and autonomy in diverse un- and/or semi-structured environments''. We
pursue this subject using a common research methodology, which is strictly applied: all
technologies that we investigate should answer actual needs that emerge from an application,
and their effectiveness is always validated on the application.
2.1 Biologically inspired systems lab
The lab’s research addresses the development of human-like (or more generally biological
systems like) sensors/actuators and biologically inspired methods for sensori-motoric
information processing, fusion and perception. With respect to perception, the objective is to
incorporate in the design of intelligent sensors biologically inspired features (human-like) of
perception (granularity, uncertainty, imprecision and vagueness), and make a reference to
anatomical localization and biological functions. Regarding sensori-motoric motion control,
our concern is the investigation and development of totally neurobiologically inspired,
adaptive and intelligent control schemes for visually guided reach, grasp and dexterous
manipulation by means of an anthropomorphic hand actuated by artificial muscles. This
research direction takes also into account the study of arm-hand coordinated movements,
which could be driven by a combination of different types of artificial muscles and/or muscle
like actuators.
Achieving the above objectives would provide on one hand more reliable, robust and accurate
quality testing and assessment devices for purely industrial applications including the food
industry and intelligent rescue systems; on the other hand, it contributes to the improvement
of our current understanding about the behavior and functioning of biological control systems
namely the central nervous system. This is essential for designing artifacts that help humans
to avoid repeated unskillful and dangerous tasks.

Annual Report 2002 35
2.1.1 Focus
The research effort at BISL is organized around four major research directions:
1. Design and development of accurate, compact sensors which are able to perform humanlike
smell and taste sensing in a variety of real world environments that have not been
specifically modified to accommodate sensors;
2. Investigation of biologically inspired feature extraction, sensory information fusion and
active perception methods that make reference to anatomical localization and biological
functions;
3. Building a heterogeneous sensory platform that is to be employed as an assessment and
quality-testing device in the food industry. It's to be equipped with sensors dedicated to
mimic five primary sensory actions and would perform human-like perception and
classification tasks; and
4. Design and development of a five fingers anthropomorphic hand actuated by artificial
muscles and equipped with force and position sensors in order to investigate and
implement dexterous manipulation.
2.1.2 Research Projects
Project: Design and development of electronic tongues and
associated data processing systems.
Project leader: Prof. Peter Wide
Project staff: Malin Lindquist (PhD student), Selim Eskiizmirliler (PhD)
Funding: KK foundation
Cooperation: Amphitec AB
Synopsis:
Human like sensors become more and more important for real time,
nondestructive inspection, quality test and assessment in industrial
applications. In this project we work on designing two integrated,
robust, reliable, low mass/power, electronic tongues each having different physical and
electronic characteristics w.r.t. two different industrial applications. The first tongue is
designed to be used in a underwater vehicle in order to provide real time water quality
monitoring in a drinking-water reservoir. Therefore, the initial design, which consisted of four
different metalic working electrodes and a glass tube (as reference electrode) has been altered.
The current version is made of a stainless steel tube acting as a reference electrode, including
two gold and platinum electrodes (as working electrodes) in its inner part in order to let the
water flow run through the tube.
The second tongue is an improved version of the initial design with the aim of adapting it to
the structural and functional constraints of the new artificial sensory head. Study and

36 AASS – Center for Applied Autonomous Sensor Systems
implementation of human-like cognitive abilities in terms of linguistic descriptions of the
taste is also envisaged.
Results in 2002:
- Finishing the design and development of the new tongues to be used in an underwater
vehicle and with the new 2G artificial sensory head.
Future developments:
- Licentiate thesis in November 2003 by Malin Lindquist on Electronic tongue as a
quality and taste assessment device.
- Two conference papers
Project: Design and development of an electronic nose and associated data processing
systems.
Project leader: Prof. Peter Wide
Project staff: Amy Loutfi (PhD student), Silvia Coradeschi (PhD),
Andreas Säterås (Exjobber)
Funding: KK foundation
Cooperation: Mobile Robotics Lab.
Synopsis:
The objective is to ground the names of odours to the sensor data from an electronic nose.
This means that instead of representing odour classification numerically, we want to use
linguistic descriptions. This is a worthwhile problem to consider for several reasons. Firstly,
using language to describe categories of measured sensor data is a desirable result in artificial
sensing systems. The motivations for this range from an introspection into our own way of
developing language, the formation of linguistic communication between robot and human, to
the need for industrial applications to accommodate users who often do not possess an expert
knowledge of the system. Secondly, using an electronic nose adds a new dimension to the
problem of grounding symbols to sensor data. This is due to the fact that humans odours are
often described using linguistic qualifiers and consequently, there are few widely accepted
standards for which we can systematically obtain a categorization. The work considers the
correspondence between symbols or linguistic labels and categories of sensor data.
Specifically, in what way can the integration of the linguistic labels with the sensor data occur
without compromising the perceptual differences of the electronic nose. This means that we
do not want to match the perceptions of an electronic nose exactly to that of a human user but
rather, enable the electronic nose to use the lexicon provided by a human user to describes its
own perceptions of odours by relying on its sensory information. The results from this project
are intended to be used both on the electronic head as well in other platforms which deal with
anchoring systems. This project is a joint effort between the Biologically Inspired Systems
Lab and the Mobile Robotics Lab.
Results in 2002:
- Finishing the design and development of infrared e-nose
- Four conference papers
- 2 Ex-jobs: Maria Larsson, Johanna Rehn
Future developments:
- One Conference paper

Annual Report 2002 37
- Licentiate thesis in May 2003 by Amy Loutfi on Communicating perceptions by
grounding language in artificial sensor systems.
Project: Design and development of an artificial sensory head
Project leader: T. Selim Eskiizmirliler (PhD)
Project staff: Linn Robertsson (PhD student)
Lena Biel (PhD student)
3 Xjobbers: Elisabeth Wikström & Mikael Sivebratt, Jorgen Widmark
Anani Ananiev (PhD) Prof. Peter Wide.
Funding: KK-Foundation, Bofors Defence AB
Cooperation : SIK, Intelligent Control Lab.
Synopsis:
The objective of this project is to build an artificial sensory head equipped with sensors
dedicated to five primary human like senses and which is to be used as a quality testing and
assessment device in the food industry. However the whole system must also be considered as
a common application platform for any sensory and/or data fusion based research activities.
Our main goals are twofold: to make this second version (2G) of the electronic head more
suitable for industrial applications; and to bring its processing abilities closer to its biological
counterpart w.r.t. sensation and perception . In this way, we have first equipped the new
robust electro-mechanical structure, designed in collaboration with the Intelligent Control
Lab, with the new e-tongue and e-nose. All processes including the control schemes,
measuring, feature extraction and perception phases are being incorporated to be driven by a
common graphical user interface. Moreover, two new Sensory Information Fusion methods
(SeFMAP and hybrid ANN based sensory information fusion) will be applied and the results
will be compared with already existing conventional and biologically inspired ones.
Furthermore, the sensing and clustering/classification results will also be associated with
facial expressions by means of 3D facial animation software from Cambridge Research
Laboratory. We are also planning to replace the existing wire-type muscle models used by the
animation software with the biologically plausible ones for the aim of transforming the
system to a simulator suitable for biologically inspired sensori-motoric control studies of
facial movements. A project that aims to automate the testing procedure for the mixture of
chocolate and wrapping paper tastes is done in collaboration with the AssiDomän Corporate
R&D, Frövi and is considered to be a first industrial application of the new 2G version.
Results in 2002:
- Two conference papers
- Licentiate thesis byLena Biel on Modeling of Perceptual Systems – A sensor fusion model
with active perception
Future developments:
- 1 Marie Curie Fellow, Iasen Hristozov will work on wavelet based feature extraction and
Sensory Information Fusion
- Association of facial expressions to the artificial sensory head assessment.
- First industrial application in collaboration with the AssiDomän Corporate R&D, Frövi.
- Two conference papers
- One journal paper

38 AASS – Center for Applied Autonomous Sensor Systems
Project: Design and development of an artificial hand
Project leader: T. Selim Eskiizmirliler (PhD)
Project staff: Linn Robertsson (PhD student)
Funding: KK + Faculty
Cooperation: Elumotion-England, ENST-Paris, INSERM 483–Paris 6, Bath University-
England, NASA-JPL, and Umeå University-Sweden.
Synopsis:
The project belongs mainly to the Biologically Inspired Sensori-Motoric Information Fusion
research area. The objective is to build a 5 fingers anthropomorphic left hand with 15 joints
(5DIP+5PIP+5MCP) actuated by electrically active polymer (EAP) muscles and equipped
with force (tactile) and joint angle sensors (i.e. somatosensory and proprioceptive sensors).
Our main objective is to implement and demonstrate principles of a totally neurobiologically
inspired, adaptive and intelligent sensori-motoric control schemes for visually guided reach
and grasp. These will incorporate biological constraints in terms of architecture and behavior
as defined by available experimental neurobiological data on the cellular, cell-population and
psycophysical level. The work will be based on the interaction of four fundamental elements
of movement control: 1) Task related anticipatory information concerning goal selection
(Motion planning), 2) Visual information, 3) Proprioceptive and somatosensory information,
and 4) Execution related information. The hand is being designed to be used by the already
existing 6 degrees of freedom light weight robot arm of Intelligent Control Lab and, a new
robot forearm actuated by 2 McKibben (designed by Shadow Co.) muscles. A study of the
coordinated left-right hand motion control in collaboration with the JPL lab of NASA is also
envisaged.
Results in 2002:
- Three journal papers
- Contract with Elumotion for manufacturing and instrumenting a five fingers
anthropomorphic hand.
Future developments:
- 1 Marie Curie Fellow, Olivier Manette, Neurobiologically inspired control of dexterous
manipulation
- One conference paper
- One journal article submission.
- Application for a research grant to the Vetenskapsrådet based on an international
collaboration between 5 research institutions.
- Application for a research grant to the Human Frontier Science Program in an
international collaboration with 5 research institutions.
- Probable involvement in an EU/FP6 project.

Annual Report 2002 39
2.1.3 Staff
Prof. Peter Wide received his MSc. Eng. and Licentiate-of-Technology degrees in 1984 and
1987 from the Dept. of Measurement Technology, University of Linköping, Sweden. In
parallel to his academic career he has worked in different electronic industries for more than
15 years as test engineer, development manager, and technical consultant. He received his
PhD in Measurement Science and Technology from Linköping University in 1996. Peter
Wide has been a visiting researcher at the Siemens Corporate R & D Fuzzy Research
Laboratory in Munich, Germany in 1994 and at the School of Information Technology and
Engineering (SITE), University of Ottawa, Canada, in 1997. He is the founder of the Center
for Applied Autonomous Sensor System, where he is now the director. He has authored and
co-authored more than 100 conference/journal papers. He serves in the program committees
of several international conferences and as a reviewer for IEEE-journals. He is an active
member of the IEEE Instrumentation and Measurement (IM) Society. His professional
interests are mainly related to the fields of sensors, feature extraction, sensor data fusion and
quality measurements.
T. Selim Eskiizmirliler received his BS degree in Electrical Engineering from the Yildiz
Technical University, Istanbul and MS degree in Electrical and Electronic engineering from
the Middle East Technical University, Ankara, Turkey in 1989 and 1993, respectively. His
MS thesis dealt with the automated classification of chromosome images based on artificial
neural networks. From 1993 to 1995 he worked at the Biomedical Engineering Research Unit
of Hacettepe University, Faculty of Medicine, Ankara, on expert biomedical diagnostic
systems. He received PhD degree in Signal & Image Processing from the Ecole Nationale
Superieure des Telecommunications, Paris, France in 2000. His PhD thesis dealt with
modeling the sensory-motor information in the cerebellar pathways and its applications to the
prediction of motion sickness in high speed tilting trains and to the control of a robot limb
actuated by artificial muscles. He has 17 conference/journal publications and is currently a
researcher at the Center for Applied Autonomous Sensory Systems Örebro University,
Sweden. His research interests concern biologically inspired control systems, artificial neural
networks, artificial muscles, sensory-motor information fusion, digital signal and image
processing.
Lena Biel was born in Karlskoga, Sweden on February 12, 1969. She was awarded the degree
of Master of Science, Computer Science and Engineering with the specialization
Communication Technique, at Luleå University of Technology in 1993. After the degree, she
has been employed at Bofors Defence in Karlskoga. In 1998 she started the PhD studies at
Örebro University in the area of sensor fusion. She received her licentiate-of-technology
degree from AASS, Örebro Univ. in 2002.
Malin Lindquist was born in Örebro, Sweden on November 5, 1975. She received her MSc
degree in automation engineering in 2001 from Örebro University.
She has been a graduate student at the Center for Applied Autonomous Sensor Systems,
Örebro University, Sweden since 2000.
Her current research interests are in biologically inspired sensors and specifically electronic
tongues.

40 AASS – Center for Applied Autonomous Sensor Systems
Amy Loutfi was born in Saint John, New Brunswick Canada. She received her Bachelor of
Science in Electrical Engineering at the University of New Brunswick in 2001. Since fall
1996 he has been a graduate student at the Center for Applied Autonomous Sensor Systems,
Örebro University, Örebro, Sweden. Her current research interests include grounding of
language to artificial systems and sensor measurement technology.
Jan Spännar was born in Smedjebacken, Sweden on March 23,1964. He recevied a Master
degree in Applied Physics and Electrical Engineering at Linköping Institute of Technology,
Sweden in 1989. Since fall 1999 he has been a graduate student at the Center for Applied
Autonomous Sensor Systems, Örebro University. He received his licentiate-of-technology
degree from AASS, Örebro Univ. in 2002. His current research interests include modeling
and identification with applications to process measurement.
Linn Robertsson was born in Örebro in 1979. She was awarded the degree of Master of
Science in Engineering with the specialization on aerospace technology at Luleå University of
Technology in 2001. Since fall 2002 she has been a graduate student at the Center of Applied
Autonomous Sensor Systems, Örebro University, Örebro, Sweden. Her current research
interests are Biologically inspired Sensory Information Fusion Systems and feature extraction
methods.

Annual Report 2002 41
2.2 Mobile Robotics Lab
Our objective is to advance the state-of-the-art in the synthesis of intelligent physical agents
capable of autonomous operation in natural environments. The world "natural" denotes real
world environments that have not been specifically modified to accommodate the agents.
These systems must incorporate motor and perceptual processes to interface with the physical
world, and abstract cognitive processes to reason about the world and the available options.
2.2.1 Focus
The focus of the research performed in this laboratory is on the crucial problem of the
integration of cognitive processes into a physically embedded reasoning system. In particular,
we address three specific, important facets of the integration problem:
1. The integration between abstract reasoning about actions and goals, and the performance
of physical action;
2. The integration between models of the world used at different levels of abstractions; and
3. The integration between the symbols used by the reasoning processes to denote physical
objects, and the perceptual data corresponding to these objects.
The methodology adopted is based on the tight integration between interdisciplinary
investigation, formal analysis, and practical experimentation. The experiments are to be
conducted on different tasks, including: (i) observation, navigation, and manipulation in
complex real-world environments; (ii) robot-human cooperation, where the sensori-motoric
and cognitive abilities are distributed between the robot and the human; and (iii) teams of
cooperating robots, where these abilities are distributed among robots.
2.2.2 Research Projects
Project: Hybrid Maps
Project leader: Pär Buschka
Project staff: Alessandro Saffiotti, Tom Duckett (from
the Learning Systems Lab).
Funding: Faculty. Pär Buschka is funded by the KK
Foundation.
Cooperation: AASS Learning Systems Lab; ENST (Ecole
National Superieure de Telecommunications)
Paris, France.
Synopsis:
Autonomous mobile robots need to use spatial information about the environment in order to
effectively plan and execute navigation tasks. This information can be represented at different
level of abstractions and of locality. Each level is adequate for some sub-tasks, but not for

42 AASS – Center for Applied Autonomous Sensor Systems
others. In this research line, we study hybrid maps, representation of the space that integrates
different levels of locality and abstraction. We focus on four levels: Perceptual - the space
accessible to the robot's perception Geometric: the geometry of the current workspace;
Topological - the topological structure of the global environment; Semantical - the semantic
structures in the space.
Main results in 2002:
- Paper published in the Robotics and Autonomous Systems journal.
- Paper presented at the Int Conference on Intelligent Robotic Systems (IROS) in Lausanne,
CH, Oct 2002.
- Paper presented at the Swedish Workshop on Autonomous Robotics (SWAR) in
Stockholm, Sweden, Oct 2002.
- Paper presented at the Int ICSC-NAISO Congress on Neuro-Fuzzy Technologies in La
Havana, Cuba, Jan 2002.
- Paper presented at the Int Conf on Information Processing and the Management of
Uncertainty (IPMU) in Annecy, France, July 2002.
Future developments:
- Pär Buschka will defend his licentiate thesis on this subject in the first half of 2003.
Project: Real-time deliberation
Project leader: Ola Pettersson
Project staff: Lars Karlsson and Alessandro Saffiotti
Funding: KK Foundation and Faculty funding.
Cooperation: Blekinge Institute of Technology, Sweden.
Synopsis:
Intelligent mobile robots need the ability to integrate robust navigation facilities with higher
level reasoning. This research line aims at combining results and techniques from the areas of
robot navigation and of intelligent agency. In particular, we investigate techniques that allow
the robot to decide when to go on with execution, and when to stop and ponder about a new
situation. We focus especially on the use of model-free diagnostic tools to detect execution
failures and thus trigger a re-deliberation process.
Main results in 2002:
- Paper presented at the Int RoboCup Symposium in Fukuoka, Japan, June 2002.
- Paper presented at the Swedish Workshop on Autonomous Robotics (SWAR) in
Stockholm, Sweden, Oct 2002.
- Paper submitted to the Int Conf on Advanced Robotics (ICAR 2003).
Future developments:
- Ola Pettersson is planned to present a PhD thesis on this subject by the first half of 2003.

Annual Report 2002 43
Project: Generating and Executing Plans under Uncertainty
Project leader: Lars Karlsson
Project staff: Abdelbaki Bouguerra
Funding: KK Foundation, Faculty funding, EU funding via the PLANET Network.
Cooperation: PLANET, European Network of Excellence in AI Planning.
Synopsis:
An intelligent mobile robot both needs to react quickly to changing conditions and to
formulate and implement plans to achieve its more long-term objectives. From the robot's
perspective, there are a number of sources of uncertainty that complicate its planning and
acting, including other agents and dynamical processes, incomplete a priori information,
restricted and unreliable sensing capabilities, and the uncertainty inherent the robot's own
actions. Fuzzy logic is an interesting tool for taking these sources of uncertainty into account,
as well as for estimating the quality and success of plans. Within this context, we
investigate issues such as how to represent and reason about actions, sensing, and complex
plans, how to integrate deliberation and acting, and how to detect when a plan is not going
well or when better opportunities arise.
Main results in 2002:
- A grant of about 1,8 MSEK has been awarded by Vetenskapsrådet to this project for the
period 2003-2005.
- Paper published in the edited volume "Plan-Based Control of Robotic Agents" (Springer,
Germany, 2002).
- Msc thesis by Tommaso Schiavinotto (visiting student at AASS in 2001) defended at the
University of Pisa, Italy.
Future developments:
- Work will focus on the project objective defined in the application funded by
Vetenskaprådet.
- A new PhD student (Abdelbaki Bouguerra) has been enrolled and will work on this
project.
Project: Anchoring Symbols to Sensor Data
Project leader: Silvia Coradeschi
Project staff: Alessandro Saffiotti, Kevin LeBlanc
Funding: KK Foundation, Faculty funding, Vetenskapsrådet
Cooperation: University of Palermo, Italy
Synopsis:
Intelligent agents embedded in physical environments need the ability to connect, or anchor,
the symbols used to perform abstract reasoning to the physical entities which these symbols
refer to. Anchoring must deal with indexical and objective references, definite and indefinite
identifiers, and temporary impossibility to percept physical entities. Furthermore it needs to

44 AASS – Center for Applied Autonomous Sensor Systems
rely on sensor data which is inherently affected by uncertainty, and to deal with ambiguities.
This research line is devoted to the definition and the in-depth study of the problem of
anchoring, and of its use in autonomous agents.
Main results in 2002:
- Paper published in the edited volume "Plan-Based Control of Robotic Agents" (Springer,
Germany, 2002).
- A special issue of the Robotics and Autonomous Systems journal on "Anchoring Symbols
to Sensor Data" had been edited by S. Coradeschi and Saffiotti. Expected publication date:
May 2003.
- Initiated a scientific cooperation with the University of Palermo, Italy, by a one week visit
of A. Saffiotti to Palermo.
Future developments:
- Amy Loutfi will defend her licentiate thesis on perceptual anchoring of odour signals by
the first half of 2003.
- A new PhD student (Kevin LeBlanc) has been enrolled and will work on cooperative
anchoring.
Project: Robot Team Coordination
Project leader: Alessandro Saffiotti
Project staff: Robert Johansson, Kevin LeBlanc, Zbigniew Wasik
Funding: CUGS, EU via the EURON Network, and KK Foundation.
Cooperation: EURON, the European Research Network; Instituto Superior Tecnico (IST) of
Lisboa, Portugal; University of Murcia, Spain; and Team Sweden (Örebro
University, Blekinge Institute of Technology, Lund University).
Synopsis:
The coordination of the actions of a team of collaborating autonomous robot presents
problems that are considerably more complex than those typically considered when planning
and regulating the motion of a single robot. In this research line, we investigate the extension
of the desirability function approach, which we use for the control of a single
physical agent, to the case of a team of cooperating agents.
Main results in 2002:
- Participation of Team Sweden to RoboCup
2002 in Fukuoka, Japan, in June 2002.
- Two papers published in the edited volume
"RoboCup 2001: Robot Soccer World Cup
V" (Springer, Germany, 2002).
- One paper presented at the Int Conf on
Pattern Recognition (ICPR) in Quebec City, Quebec, Canada, August 2002.
- Initiated cooperation with the University of Murcia, Spain.
- Organized a workshop on cooperative robotics at the Int Conf on Intelligent Robotic
Systems (IROS) in Lausanne, Switzerland, October 2002.
- Co-organized a European Summer School on Cooperative Robotics together with the
Instituto Superior Tecnico (IST) of Lisboa, Portugal.

Annual Report 2002 45
Future developments:
- Team Sweden will participate in RoboCup 2003 (Italy) again under the guidance of
AASS.
- A new PhD student (Robert Johansson) has been enrolled and will work on this project.
Project: Fuzzy Behavior-Based Control of a Mobile Manipulator
Project leader: Zbigniew Wasik
Project staff: Alessandro Saffiotti
Funding: KK Foundation and Faculty funding.
Cooperation: Not yet.
Synopsis:
In this research line, we investigate new solutions to the mobile manipulation problem. The
investigation proceeds along two lines: (i) to develop a behavior-based approach to real-time
visual servoing of a mobile manipulator; and (ii) to extend the "Thinking Cap" architecture to
cope with the simultaneous control of the manipulator and of the mobile platform on which
the arm is mounted. The project will be carried out on a custom-built robot arm installed on a
Nomad 200 platform. The target tasks will include the ability to open doors, and to grasp and
move objects in the environment.
Main results in 2002:
- Paper presented at Int Conf on Intelligent Robotic Systems (IROS) in Lausanne,
Switzerland, October 2002.
Future developments:
- Zbigniew Wazik will complete his PhD thesis on behavior-based manipulation by the first
half of 2004.

46 AASS – Center for Applied Autonomous Sensor Systems
2.2.3 Staff
Alessandro Saffiotti, PhD, is a professor of Computer Science at the Center for Applied
Autonomous Sensors Systems of Örebro University, Sweden, where he heads the Mobile
Robotics Lab. He has previously been a researcher with the University of Pisa (Italy), SRI
International (USA), and the Universite Libre de Bruxelles (Belgium). His research interests
encompass autonomous robotics, soft computing, and non-standard logics for common-sense
reasoning. He has published more than 70 research papers in international journals and
conferences and co-edited the book ``Fuzzy logic techniques for autonomous vehicle
navigation'' (Springer, 2001). He was a chair of the 1999, 2001 and 2003 IJCAI workshops
on ``Reasoning with Uncertainty in Robotics'', of the AAAI 2001 fall symposium on
``Anchoring Symbols to Sensor Data'', and of the 2002 IROS workshop on ``Cooperative
Robotics''. In 2002 he co-organized the First European Summer School on cooperative
robotics. He is the leader of Team Sweden, a national effort in robot soccer comprising about
15 people from four universities. He is a member of IEEE, AAAI, and IAS.
Silvia Coradeschi is an assistant professor at the Center for Applied Autonomous Sensor
Systems at Örebro University. She has received a master degree in Philosophy at the
university of Florence, a master degree in Computer Science at the university of Pisa, and a
PhD in Computer science at Linköping University. She is a member of the board of trustees
of the RoboCup Federation and was general chair of the Third Robot World Cup Soccer
Games and Conferences (RoboCup-99). She is also a member of the Advisory Board for the
Seventeenth International Joint Conference on Artificial Intelligence (IJCAI-01) and member
of the board of the European Coordinating Committee for Artificial Intelligence. She is
director of the undergraduate studies for the Computer Engineering Division at Örebro
University and she is vice-head of the Technology Department. She has published 18
conference/journal papers and her main research interest is in establishing the connection
(anchoring) between the symbols used to perform abstract reasoning and the physical entities,
which these symbols refer to. She also works in multi-agent systems and cooperative robotics.
Lars Karlsson is an assistant professor at the Center for Applied Autonomous Sensor
Systems at Örebro University. He received a MSc. degree in Computer Science at Linköping
University in 1993, and a PhD in 1999. He was a co-chair of the Third Robot World Cup
Soccer Games and Conferences (RoboCup-99). He is the coordinator for the MSc program in
Computer Engineering at Örebro University, and he is the director for graduate education in
AASS. He has 18 conference/journal publications and his research interests include planning
under uncertainty, reasoning about actions and integration between reasoning and
action/perception in autonomous systems.
Pär Buschka was born in Gothenburg, Sweden on April 22, 1966. He received the MSc
degree in computer science and engineering in 1996 from Chalmers University of Technology
in Gothenburg, Sweden.
Since autumn 1999 he has been a graduate student at the Center for Applied Autonomous
Sensor Systems, Örebro University, Örebro, Sweden. His current research interests include
mobile robot localization and navigation using hybrid maps.
Ola Pettersson was born in Falkenberg, Sweden on March 7, 1972. He received the
BSc degree in electrical engineering in 1994, and the MSc degree in electrical engineering in
1995 (specialization in photonics and image analysis), both from Halmstad University,

Annual Report 2002 47
Halmstad, Sweden. In 2000 he received a Licentiate of Technology in computer science from
Linköping University, Linköping, Sweden.
Since fall 1996 he has been a graduate student at the Center for Applied Autonomous Sensor
Systems, Örebro University, Örebro, Sweden. His current research interests include mobile
robot action control, planning, and execution monitoring.
Zbigniew Wasik was born in Nowy Dwor Mazowiecki, Poland on October 13, 1973. He
received both the BSc degree and the MSc degree in electrical engineering in 1998
(specialization in Automation and Robotics) from Wroclaw Univesrity of Technology,
Wroclaw, Poland. Since fall 1998 he has been a graduate student at the Center for Applied
Autonomous Sensor Systems, Örebro University, Örebro, Sweden. His current research
includes a behavior-based approach to real-time visual servoing of a mobile manipulator.
Kevin LeBlanc was born in Moncton, New Brunswick, Canada, on November 11, 1977. He
received his B.Sc. in Electrical Engineering with a Computer Engineering Option at the
University of New Brunswick in May 2001. In the fall of 2002 he began graduate studies at
the Center for Applied Autonomous Sensor Systems, Orebro University, Orebro, Sweden.
His research falls within the cooperative perception area; his primary focus deals with
cooperative perceptual anchoring.
Robert Johansson was born in Jämshög, Sweden on September 17, 1978. He received the
MSc degree in Computer Technology in 2002 from Örebro University, Örebro, Sweden.
Since late spring 2002 he has been a graduate student at the Center for Applied Autonomous
Sensor Systems, Orebro University, Orebro, Sweden. His current research is within the field
of cooperative robotics.
Abdelbaki Bouguerra was born in M'sila, Algeria on October 10, 1974. He received an
Engineering degree in Computer Science in 1997 (in Computer Systems) from The National
Institute of Computer Science, Algiers, Algeria, and a Master degree in Computer Science in
2001 (specialization Industrial Computer Science) from the National Military Polytechnics
School, Algiers, Algeria. He joind Schlumberger-Geoquest as a data management Engineer
in the same year where he worked for approximately one year, then he joined the department
of computer science at Linköping University as a visiting scholar for six months. Since
Janauary 2003, he has been a graduate student at the Center for Applied Autonomous Sensor
Systems, Orebro University, Orebro, Sweden. His current research includes planning and
execution for mobile robots.

48 AASS – Center for Applied Autonomous Sensor Systems
2.3 Intelligent control lab
Our objective is to develop intelligent control methodologies where techniques from soft
computing, knowledge-based control, and conventional control are synergistically integrated
for the purpose of: (i) low-level robust and adaptive control of non-linear, time-varying and
uncertain systems; (ii) supervisory-level control for gain-scheduling, monitoring, adaptation,
and reconfiguration; (iii) task-level control of goal-oriented system behavior and
management, and supervision of activities at lower levels.
2.3.1 Focus
The focus of research is on the integration of soft computing, conventional control, and
sensing technologies into intelligent systems for high-speed motion control. Included here are
also advanced aspects of the design of dedicated mechatronic systems with embedded
intelligence, e.g., unique light-weight robot arms operating in a stand-alone mode, or mounted
on mobile robot platforms.
The most important directions of our research are:
1. Integrating visual-sensing and the control of a mechatronic system into perception-based
control system (visual-servoing) able to perform human-like motions and operations;
2. Development of advanced control architectures and hardware for the implementation of
control routines based on soft computing;
3. Development of advanced vision- and image-processing algorithms incorporating highspeed
digital hardware and soft computing; and
4. Integration of all of the above into a complete perception-based motion control system
able to operate at high speed at all levels of intelligent control: low-, supervisory, and
task-level control.
The current research projects are grouped according to the following three major fields: (A)
Advanced mechatronic systems; (B) Advanced control; and (C) Vision, image and signal
processing.

Annual Report 2002 49
2.3.2 Research projects
(A) Advanced mechatronic systems
Project title: Lightweight Robot Arms
Project leader: Anani Ananiev
Project staff: Docent Ivan Kalaykov, Boyko
Iliev
Funding: KK Foundation and Faculty
Cooperation: Bulgarian Academy of Sciences,
Institute of Mechanics
Synopsis:
The objective is to create a lightweight robot arm with an open and re-configurable
architecture for installing on top of mobile platforms (for example the Nomad 200 robot). The
primary need for this arm came from the necessity of implementing mobile manipulation for
the needs of the research in the Mobile Robotics Lab. The secondary goal encountered during
the work on the project was the possibility to extend the application to other types of robotic
tasks with more industrial emphasis, like food industry. Major contributions are large
payload-to-weight ratio, space-saving design, flexible mounting, extremely low maintenance
requirements, advanced open architecture controller.
Two different models are designed and prototyped. The bigger arm corresponds in
dimensions to the specific construction of the Nomad 200 mobile platform, while the smaller
arm can be attached to various constructions, including also desktop installations. The smaller
arm can be used also as a test platform for cooperative robot manipulation. The project is not
directed to produce a PhD-thesis, but is a tool that can be used in other PhD projects.
Results in 2002:
- The project ended in 2002 with the bigger arm moving on top of Nomad 200 robot and a
desktop version of the smaller arm.
- The arms were demonstrated at Tekniska mässan, Stockholm, October 2002 where they
attracted significant interest. One WIPO international patent is already obtained and
published in 2001.
Future developments:
- Further improvements aiming at achieving better performance have to be made as
expected under the future VINNOVA-Robotdalen initiative.
- Visual feedback and force-sensor feedback is currently under development and will be
demonstrated during 2003.

50 AASS – Center for Applied Autonomous Sensor Systems
Project title: Flexible Food Cell
Project leader: Docent Ivan Kalaykov
Project staff: Anani Ananiev, Lars Jennergren
Funding: KK Foundation, Faculty and ABB
Cooperation: SIK, Göteborg
Synopsis:
The project started in April 2002 and the objective is to create an intelligent assembly cell for
assembling ready-to-eat meals. The production has to be customer oriented, which leads to
shorter and more irregular production series with an increased number of different products.
During the entire project, the hygienic aspects and mechanical and thermal restrictions
associated with the food assembly process are considered. The work is strongly related to the
food factory of the future, which is planned to be ready for inauguration in June 2003, and the
results will hopefully improve its capacity. The following major contributions are expected:
hygienic and robust manipulation of specific food groups, a flexible cell layout that is easy to
reconfigure for new product variants, a demonstration cell that can assemble a set of prespecified
products.
Results in 2002:
- The general concept of the flexible food cell specifications and requirements for various
types of food grippers/manipulators are defined. Primary design of some food
grippers/manipulators is scheduled for 2003.
Future developments:
- As the project is in its initial phase, all planned and expected results should be considered
in the next few years.
- Work on a licentiate (and later PhD) thesis has been started by Lars Jennergren.
(B) Advanced control
Project title: Minimum-Time SMC of Robot Manipulators
Project leader: Docent Ivan Kalaykov
Project staff: Boyko Iliev
Funding: KK Foundation and Faculty
Cooperation: Technical University Sofia
Synopsis:
The objective is to develop techniques that provide near time-optimal control of robot
manipulators. These are needed for obtaining predicted time response of the robot
manipulator when the robot motion must be synchronized with other motions, for example
when the robot should intercept a moving object. The necessity of this project come both from
scientific challenges and application needs. Major contributions are: achieving minimum-time
performance in the presence of uncertainty, using the theory of sliding modes; quasiminimum
time control algorithms for Point-to-Point motion of serial robot manipulators;
design and implementation of control systems for two lightweight robot arms designed at
AASS.

Annual Report 2002 51
Results in 2002:
- Two publications and a test platform for verification and evaluation of the developed
methods and algorithms. The tests will be performed firstly on a 1 DOF manipulator and
then on the 6 DOF lightweight robot arm (Pandi –II) using the available control
prototyping system DS 1103 from dSpace Inc.
Future developments:
- Demonstrations - different control algorithms implemented on test platforms – 1 DOF
arm, and Pandi-II, quasi minimum-time control methods on Pandi – II.
- A PhD thesis is planned to be presented in November-December 2003.
Project title: Model-based fuzzy control
Project leader: Prof. Dimiter Driankov
Project staff: Pontus Bergsten, Bourhane Kadmiry
Funding: KK Foundation and Faculty
Cooperation: WITAS, Univ. of Linköping, Dept. of Computer Science;
ABB Automation Products AB, Dept. AMC;
Siemens Corporate Research and Development, Germany.
Synopsis:
The focus is on analysis and design issues for observers and controllers for Takagi-Sugeno
(TS) fuzzy systems. Many physical systems are nonlinear in nature and using the well-known
linear techniques for such systems may result in bad performance, and even instability. On the
other hand, analysis and design of observers and controllers for general nonlinear systems
tend to be a quite involved procedure. It turns out, however, that a TS fuzzy system is able to
represent or approximate a large class of nonlinear systems. Developing methods for
observation and control for TS systems is thus worthwhile.
Results in 2002:
- Lic-thesis: A lic-thesis by Bourhane Kadmiry (WITAS) on the use of fuzzy controllers,
derived from Takagi-Sugeno systems, for the design and analysis of unmanned
helicopter's flight control system.
Future developments:
- The subject of our future research effort is the use of fuzzy controllers, derived from
Takagi-Sugeno systems, for the purpose of visual-servoing based control for unmanned
aircraft and control of cold steel mills.

52 AASS – Center for Applied Autonomous Sensor Systems
Project title: Visual-servoing based simulated flight
Project leader: Prof. Dimiter Driankov
Project staff: Martin Persson, Bourhane Kadmiry,
Pontus Bergsten
Funding: WITAS, KK Foundation, Faculty and FFV
Aerotech Telub
Cooperation: FFV AerotechTelub, Örebro;
WITAS, Univ. of Linköping, Dept. of
Computer Science.
Synopsis:
In order to allow for autonomous behavior of UAVs (Unmanned Aerial Vehicle) in
surveillance and reconnaissance missions, it is of interest to embed a vision sensor as an
integral part of the control loop. This technique is called visual-servoing and is mainly found
in robotics applications while very few studies of visual-servoing for airborne platforms can
be found in the literature. To investigate the performance and possibilities of various visualservoing
techniques applied to acquisition and tracking in the case of UAVs, we develop
SEAVS, a 3D-simulation environment for aerial visual-servoing. SEAVS visualizes the UAV
simulation in a 3D window together with terrain, map features and orthographic photos. This
gives a realistic environment, which also is used to create images simulating the onboard
video camera. The object of the research is to define how visual servoing techniques can be
applied to airborne platforms in acquisition and tracking.
Results in 2002:
- A lic-thesis, reporting the SEAVS simulation environment and preliminary results on the
use of visual-servoing defended by M. Persson.
- First version of SEAVS was completed.
Future developments:
- A PhD thesis on visual-servoing based flight control for unmanned helicopter is to be
defended by B. Kadmiry in 2003.
Project title: Reconfigurable Control Architectures
Project leader: Adj. Prof. George Fodor
Project staff: Christer Lindqvist, Nur Yilmazturk (ABB), Pontus Bergsten (ABB)
Funding: ABB Automation Technology Products and Faculty
Cooperation: Western Michigan University, USA
Synopsis:
Reconfigurable distributed real-time control systems consist of a large number of loosely
connected processing units that can be hosts for software agents acting in real-time. These
units act over a wide geographical area. Such systems have increased importance for the next
generation of industrial control systems. In a dynamic architecture it is the role of a so-called
Architecture Broker to distribute software agents dynamically among processing units, such
that the formal specification of the Architecture Broker is the key knowledge needed for the
design of such systems. The scope of the research is a theoretical approach for the analysis
and design of advanced Architecture Brokers for real-time distributed architectures. Results

Annual Report 2002 53
are expected in the following areas: verification and supervision of global real-time properties
for systems with distributed components in uncertain environments; specification of
continuous control algorithms that are robust to variations of the system’s granularity;
integration of continuous and discrete algorithms into distributed reconfigurable architectures.
Results in 2002:
- Work on a lic-thesis by Christer Lindqvist was
commenced under the supervision of George
Fodor.
Future development:
- Licentiate thesis by Christer Lindqvist in 2003.
(C) Vision, image and signal processing
Project title: Fast fuzzy logic based image processing
Project leader: Docent Ivan Kalaykov
Project staff: Gustav Tolt
Funding: KK Foundation and Faculty
Cooperation: not yet
Synopsis:
The objective is to develop techniques for high-speed image processing able to provide over
50-60 frames per second for the low and medium level of image processing. This
performance is needed for real-time image processing systems such as fast vision systems in
visual-servoing tasks when it is necessary to control high-speed motion. This project is based
completely on the academic challenge to outperform the normal speed industrial vision
systems, which generally limit the frequency band of robot motion control. However, the fast
image processing algorithms can be used in various applications where the processing time is
important factor. Major contributions are: unified approach to low and medium level image
processing based on the fuzzy similarity principle, high performance algorithms for low and
medium level image processing both in terms of speed and quality of the processing, set of
hardware structures for implementing the typical image processing algorithms.
Results in 2002:
- A library of fuzzy similarity and homogeneity based algorithms for low level image
processing are developed and tested.
- First prototype system developed and exhibited at CeBIT, Germany.
Future developments:
- Working prototype system and experimental tests in real-time image processing in 2003.
- Licentiate thesis by Gustav Tolt to be presented in March 2003. PhD thesis planned for
2004.

54 AASS – Center for Applied Autonomous Sensor Systems
Project title: Fast Hardware for Fuzzy Logic Signal and Image Processing
Project leader: Docent Ivan Kalaykov
Project staff: Gustav Tolt
Project funding: KK Foundation and Faculty
Cooperation: Hardi Electronics AB, Lund (started end of 2002)
Synopsis
The project has several objectives the common denominator of which is achieving high-speed
image processing. The first objective is to build a high-speed fuzzy logic processor able to
operate at speeds more than 50 MFLIPS (Mega Fuzzy Logic Inferences Per Second) . Here
major contributions are: new concept of Fired-Rules-Hyper-Cube (FRHC) concept created as
a base for the hardware design; layered parallel and pipelined architecture with predictable
behavior in time; high-speed performance that can be optimized depending on the progress of
the semiconductor technology. The second objective is to design and develop high-speed
vision system in hardware able to provide over 50-60 frames per second for typical image
processing algorithms. It is needed for real-time control of high-speed motion. Major
contributions are: high-speed vision sensor for robot motion control; embedded high-speed
image processing algorithms using fuzzy logic; programmable architecture
Results in 2002:
- One WIPO international patent is already obtained.
- Basic structures for image processing are developed and simulated.
Future developments:
- Implementing the main results achieved so far on advanced FPGA-based boards and
completing IP-cores for FPGA-based design of complex processing systems. Working
prototype system and experimental tests will be done in 2003. Part of this work is to be
included in the PhD thesis of Gustav Tolt.

Annual Report 2002 55
2.3.3 Staff
Docent Ivan Kalaykov received his Dipl. Eng. in Electrical Engineering, Ph.D. in Control
Engineering from the Technical University of Sofia, Bulgaria, 1972 and 1985 respectively.
He has been an Associate Professor at: Dept. of Control and Systems Engineering, Technical
University of Sofia, since 1987; Dept. of Physics, Tromsoe University, Norway, 1998-1999;
and the Dept. of Technology, Örebro University since 2000, where he is now the head of the
Intelligent Control Lab. He has been a visiting researcher at the Control Laboratory of Dept.
of Electrical Engineering, Delft University of Technology, Netherlands and Institute for
Peripheral Microelectronics, University of Kassel, Germany. He has been also Chairman and
Vice-chairman of the Bulgarian Union of Automation and Informatics (UAI - the professional
society of control engineers). He served as IPC-member of many scientific events and has
more than 60 conference/journal publications. His research interests include adaptive control
systems, intelligent control systems, fuzzy logic control, fuzzy logic hardware, and visual
servoing for robot manipulation. He is a member of IEEE, SPIE and EUSFLAT.
Prof. George Fodor received his M.Sc. degree in Electrical Engineering from the
Polytechnic Institute of Cluj, Romania, in 1979, and the Ph.D. degree in Computer Sciences
from the Linköping University, Sweden, in 1995. He is currently the head of the Systems
Development Dept. for Force Measurement and Systems at ABB Automation Products,
Sweden. He is an Adjunct Professor at Western Michigan University, Kalamazoo, USA, from
1997 and member of the Advisory Board of the Electrical and Computer Engineering Faculty
at the same university from 1998. He is an associated professor at Örebro University, Sweden,
from 2000. His research interests include discrete control, control of complex industrial
systems, discrete fault detection and isolation, ontological control, software architectures. He
is the author of one book and some 30 conference and journal articles.
Anani Ananiev is an Assistant Professor at the Center for Applied Autonomous Sensor
Systems at Örebro University. He received his Dipl.Eng. in Mechanical Engineering and
Ph.D. in Robotics and Automation from the Technical University of Sofia in 1987 and 1992
respectively. He is an associate professor (on leave) at the Dept. of Robotics and
Mechatronics, Institute of Mechanics, Bulgarian Academy of Science. He has
been a visiting researcher at the Robotics Laboratory of the Wuppertal University , Germany,
1994 -1998; and Dept. of Robotics at the Institute Superior Technico, Lisbon, 1995-1997. He
has 65 conference/journal publications, 27 patents and has received 2 bronze medals at
INPEX IX (US Invention New Product Trade Show) in 1993. His research interests include
robotics, supper fast robot arms, light-weight robot arms.
Prof. Dimiter Driankov received the BSc Math degree, the MSc degree in Computer Science
from the University of Sofia, Faculty of Mathematics, Sofia, Bulgaria, and the PhD degree in
Computer Science from the Univ. of Linköping, Linköping, Sweden in 1973, 1975, and 1988
repectively. From 1985 to 2000 he was with the Dept. of Comp. Sci.,Univ. of Linköping
where he was heading the Autonomous Systems Lab (1996-1999). At present he is a visiting
researcher with the Division for Artificial Intelligence and Integrated Computer Systems at
this department and a member of the Wallenberg Lab for Information Technology and
Autonomous Systems. Since 2001 he is a professor at the Technology Dept., Örebro Univ.,
Örebro, Sweden acting as a research coordinator for the Applied Autonomous Sensor Systems
Center. During the period 1990--1997 he has spent a total of three years as Visiting Scientist
at Siemens Corporate R & D, Munich, Germany. He is a member of the IEEE Technical

56 AASS – Center for Applied Autonomous Sensor Systems
Committee on Intelligent Control and on the editorial board of the Int. Journal of Uncertainty,
Fuzziness and Knowledge-based Systems. He has co-authored and authored more than 40
conference, journal articles, and book chapters, 2 books, and 3 edited volumes in the areas of
fuzzy logic and fuzzy control. He is a coordinator for a EC Marie Curie Training Site in the
area of autonomous robotic systems. His current research interests and activities are in the
areas of model-based fuzzy control, computational theories of perception, and perceptionbased
control of autonomous vehicles.
Boyko Iliev received his MSc in Control Systems Engineering of the Technical University,
Sofia, Bulgaria in 1996. During the period 1996-1998 he worked as a research assistant at the
Central Laboratory for Parallel Processing, Bulgarian Academy of Sciences. Since 1998 he is
a graduate student at the Center for Applied Autonomous Sensor Systems, Örebro University,
Sweden. His current research interests include fuzzy and sliding mode control and
biologically inspired control. Main application area: development and control of robotic
systems.
Christer Lindkvist was born in Örebro, Sweden, on December 23, 1954. He received his
BSc degree in 1988 from Örebro University. From July 1, 1989 he works as teacher in the
Department of Technology, Örebro University. In the fall of 2001 he became a graduate
student with AASS, Örebro University. His current research interests include distributed
component-based systems.
Martin Persson received MSc degree in Applied Physics and Electronic Engineering from
the University of Linköping, Sweden, in 1994. He has been employed by Aerotech Telub
since 1994 as system engineer working mainly with guidance, control and simulation models
for dispenser weapon systems and stand-off lissiles. He is since 1999 an industrial PhD
student with AASS at Örebro University. His research interests are visual servoing in aerial
applications.
Gustav Tolt was born in Örebro, Sweden, on January 14, 1975. He received MSc degree in
engineering physics (specialization in signal processing and image analysis) from Chalmers
University of Technology, Gothenburg, Sweden. Since February 2000 he is a graduate student
at AASS, Örebro University. His research interests include high-speed image processing and
computer vision.
Lars Jennergren received his MSc in Mechanical engineering at the Technical University of
Luleå, Sweden in 2002. His main subject was production engineering, this was also the area
of his thesis work conducted at SKF, Gothenburg, Sweden. Since april 2002 he is a graduate
student within a project conducted in co-operation between SIK, Swedish institute for food
and biotechnology, and AASS, Örebro University, Sweden. His research includes
requirements and building of a flexible assembly cell for ready-to-eat meals including topics
such as production strategies, flexible cell layout, grasping and manipulation of fragile items,
such as food.

Annual Report 2002 57
2.4. Learning systems lab
Our objective is to advance the state-of-the-art in the theory and practice of learning and
adaptation in autonomous sensor systems. This will be achieved by developing strongly
autonomous agents with the capacity to learn from experience and thus to adapt to highly
dynamic and uncertain environments and working scenarios.
These situations are extremely difficult to model by traditional methods of numerical analysis
due to the highly unpredictable nature of both the agent's sensory perceptions and the effects
of motor actions, and the presence of other agents such as humans. We therefore focus on
learning and adaptation through interaction of the agent with the user and environment.
The techniques that we apply include supervised and unsupervised machine learning
algorithms such as artificial neural networks, genetic algorithms, self-organisation, etc.
2.4.1 Focus
The research focus of the Learning Systems Lab is the development of robots and
autonomous systems, which can improve performance by learning from their own sensory
input. At present, the research activities of the lab are organized around two major themes:
1. Learning from machine-human interaction. This includes learning which facilitates
cooperation between humans and machines, e.g., recognition, tracking and identification
of humans by an autonomous robot; knowledge acquisition from a human expert, e.g.,
diagnosis of medical images; and learning of robotic skills from demonstrations by a
human operator.
2. Learning sub-symbolic representations and behaviours for robotic systems. This includes
the acquisition of internal models such as cognitive maps through active exploration of the
environment, and learning of sensor-motor competences that map sensor readings directly
onto motor actions, resulting in complex behaviour grounded in the sensor-motor
experience of the agent.
The research methodology of the Learning Systems Lab is based on developing theoretically
sound solutions to real world problems, and places strong emphasis on a tight coupling
between theory and empirical methods; including the use of performance metrics, quantitative
comparisons of actual systems, and statistical tests of significance.

58 AASS – Center for Applied Autonomous Sensor Systems
2.4.2 Research Projects
Project: Hierarchical Learning of Robot Behaviours
Project leader: Tom Duckett
Project staff: Li Jun
Funding: KK and Faculty funding.
Cooperation: not yet
Synopsis:
This project will develop a generic architecture for learning sensor-motor
behaviours in robotic systems. The robot will develop its own internal
representations through a process of incremental on-line learning. One
application of this technology would be programming by demonstration,
where the human operator simply shows the robot what to do (e.g., ”water
my plants”, ”clean my bathroom”, etc.). The behaviours acquired will be
complex, non-linear and non-achievable by a purely reactive system, so
the learning system must incorporate representations that encode the
temporal and spatial context of the task and environment. A particular
challenge will be to carry out all of the learning in real-time. The project
investigates the implementation of a hierarchy of fast, on-line learning
algorithms that are integrated by layered learning. The techniques
investigated include, but are not limited to, embedded machine learning
algorithms such as self-organisation, recurrent artificial neural networks
and reinforcement learning.
Results in 2002:
- Analysis and design of a neural network architecture for behaviour learning incorporating
unsupervised and supervised learning. An experimental analysis has been conducted for
the application of mobile robot control.
- Paper published in Swedish Workshop on Autonomous Robots, 2002.
Future developments:
- The next experiments will investigate alternative applications and platforms, e.g.,
feedback control of a mobile manipulator, and alternative sensory inputs, e.g., vision.
- Adaptation of the acquired behaviours through on-line learning
- Licentiate thesis: expected 2003 (Li Jun)
- PhD-thesis: expected 2005 (Li Jun)

Annual Report 2002 59
Project: Human-Robot Cooperation
Project leader: Tom Duckett
Project staff: Grzegorz Cielniak
Funding: KK and Faculty funding.
Cooperation: research visit by Grzegorz Cielniak to the AI Lab, University of Freiburg,
Germany as a Marie Curie Fellow (Oct 2002 – Jan 2003), with Prof. Wolfram
Burgard.
Synopsis:
The ability to interact with people is an important requirement for robots, which operate in
populated environments. In tasks such as cleaning, housekeeping, rehabilitation, entertainment,
inspection and surveillance, so-called service robots need to communicate and
cooperate with people. To enable this interaction, the robot needs to know how many people
there are in the neighbourhood, their position, and who they are (the three fundamental
problems of recognition, tracking and identification). This project investigates embedded
machine learning algorithms for robot-human cooperation. In current work, we are integrating
different methods for recognition, tracking and identification of people by a mobile robot
using vision and laser range-finder sensors. Possible themes for future research include
recognition, interpretation and prediction of human behaviour; and development of natural
language or gesture-based interfaces for communicating tasks to robots.
Results in 2002:
- Experiments on human identification using vision.
- Development of a learning system for recognizing, tracking and identifying several
humans in an office environment using Abstract Hidden Markov Models (AHMMs),
paper submitted to IJCAI’03.
Future developments:
- Scaling up of the AHMM system, and improved techniques for human identification, e.g.,
incorporating face recognition, gait recognition, etc.
- Development of a complete application “the Robotic Security Guard”, which should be
capable of recognizing AASS members and tracking their movements, while raising the
alarm if unauthorized persons enter the environment.
- Licentiate thesis: expected 2003 (Grzegorz Cielniak)
- PhD-thesis: expected 2005 (Grzegorz Cielniak)

60 AASS – Center for Applied Autonomous Sensor Systems
Project: Simultaneous Localization and Map Building
Project leader: Tom Duckett
Project staff: Tom Duckett
Funding: KK and Faculty funding.
Cooperation: Artificial Intelligence Group, Manchester University, UK, with Dr. Stephen
Marsland and Dr. Jonathan Shapiro.
Synopsis:
This project addresses the problem of navigation by mobile robots operating in large, real
world environments, which have not been modified for the purpose of robot navigation. Maps
are essential for mobile robots navigating in complex environments, being needed for selflocalisation,
path planning and interaction with humans. To navigate in unknown
environments, a self-governing robot is faced with a fundamental dilemma: to explore and
learn maps of uncharted territory, the robot needs to know its location, but in order to know
its location, the robot needs a map. Most existing solutions to the problem of simultaneous
localisation and mapping (SLAM) work by decoupling the localisation and mapping
processes, assuming that the data association problem (i.e., landmark identification) is already
solved when measurements are updated into the map. However, this assumption is bound to
fail eventually for complex environments. This project investigates robust SLAM algorithms
which take into account both the measurement uncertainty (e.g., due to sensor noise) and
identification uncertainty (e.g., due to perceptual aliasing, environmental dynamics, sensor
noise, etc.).
Results in 2002:
- Article on map learning published in Autonomous Robots Journal, based on work carried
out in cooperation with Manchester University.
- Development of a new approach for SLAM based on genetic algorithms, paper submitted
to IEEE ICRA’03.
Future developments:
- Planned research visit by Udo Frese, a Ph.D. student from DLR (German Aerospace
Center), Germany as a Marie Curie fellow in January-March 2003.
- Development of new SLAM algorithms based on techniques for statistical analysis and
global optimization.
An example of a learned map

Annual Report 2002 61
Project: Robot Navigation by Smell
Project leader: Tom Duckett
Project staff: Achim Lilienthal (visiting Ph.D. student from Tuebingen University, Germany,
as a Marie Curie fellow)
Funding: Faculty and EU via the Marie Curie Training Program.
Cooperation: Tuebingen University, Germany
Synopsis:
Smell is perhaps the least studied sense in robotic applications. Gas sensor systems known as
”electronic noses” have been widely used under laboratory conditions, e.g., for food analysis,
but so far there have been few applications on mobile platforms in real world environments,
e.g., for use by intelligent service robots. There are a number of reasons why this should be
so, including the complexity of the sample handling process and environmental influences
such as air turbulence. This project investigates learning mechanisms for navigation using
olfaction. Several different approaches are considered, including environment modelling (e.g.,
map building), reactive sensor-motor behaviours, or reimplementation of the navigation
strategies used by biological systems. The learning techniques investigated include, but are
not limited to, embedded machine learning algorithms such as artificial neural networks,
genetic algorithms and self-organisation.
Results in 2002:
- Paper published at Swedish Workshop on Autonomous Robots on previous research work
reported in the activity report of 2001.
- Development of a new experimental platform, including an absolute positioning system
for tracking the mobile robot and a new design of the mobile nose.
- Development of new odour tracking systems and experimental analysis, results to be
reported in 2003.
Future developments:
- Papers on the work of the Marie Curie fellow to be submitted in 2003.

62 AASS – Center for Applied Autonomous Sensor Systems
Project: Medical Image Processing
Project leader: Tom Duckett
Project staff: Lucia Ballerini
Funding: KK and Faculty funding.
Cooperation: University Hospital, Örebro, with Prof. Lennart Franzen.
Synopsis:
Breast cancer is the most common form of cancer among women. Analysing the tissue sample
is a slow procedure, and sometimes it is difficult to separate benign changes from malignant
changes. This project investigates the automatic analysis of images taken from breast tissue
samples with a digital microscope, by learning from the diagnoses given by a pathologist. The
aim of the project is first to discriminate abnormal tissue from normal tissue, and then to
classify both benign conditions (e.g., sclerosing adenosis, fibroadenosis) and malignant
conditions (e.g., ductal cancer, lobular cancer) using machine learning techniques for pattern
recognition. The eventual goal of the project is to develop a fully automatic system including
autonomous control of the digital microscope.
Results in 2002:
- Prototyping of pattern recognition systems for determining abnormal from normal tissue
and for classification of abnormalities. Preliminary results were very encouraging.
- Award of the “most new thinking student project” at Örebro University for the work of
the two C-level exjobb students Catharina Carlsson and Karin Bergstrand in 2002.
Future developments:
- Further D-level exjobb project in 2003.
- Improved methods for feature extraction and pattern recognition.
- Attention focussing by selecting target areas of the tissue samples with the digital
microscope.

Annual Report 2002 63
2.4.3 Staff
Tom Duckett is an assistant professor at Dept. of Technology, Örebro University and the
head of the AASS Learning Systems Lab. He received the B.Sc.(Hons) degree in Computer
and Management Sciences from Warwick University (UK) in 1991. After working in
industry as an Analyst/Programmer for several years, he was awarded the M.Sc. degree with
distinction in Knowledge Based Systems by Heriot-Watt University (UK) in 1995. His M.Sc.
dissertation was carried out as an ERASMUS exchange student at Karlsruhe University (DE).
He received his Ph.D. on concurrent map building and self-localization for mobile robot
navigation from Manchester University (UK) in 2000, including a one month research visit to
Bremen University (DE) on landmark selection for mobile robot navigation. He has authored
4 journal articles and 15 full paper refereed conference papers, and is a regular reviewer for
several international conferences and journals. He also has 7 years of teaching experience at
university level. His professional interests are in autonomous robotics, machine learning and
neural networks, artificial intelligence, and navigation systems.
Lucia Ballerini is a guest researcher at the Dept. of Technology, Örebro University in the
AASS Learning Systems Lab since October 2002. She graduated from the University of
Florence (Italy) in Electronic Engineering in 1993 and received the Ph.D. in Bioengineering
in 1998 working at the Department of Electronic Engineering (Florence, Italy) on a project to
develop an automatic system to aid medical diagnosis. She partecipated in several projects for
medical image analysis, sponsored by MURST. From 1999 to 2002 she was a post-doc at the
Centre for Image Analysis, Uppsala (Sweden) working on the project "Image analysis
methods for food quality measurements" sponsored by MISTRA. She has published more
than 30 papers in international journals and refereed conferences. She is a member of ADI -
Italian Ph.D. Association (board member and ``foreign affairs'' delegate of the association);
EvoNet - The Network of Excellence in Evolutionary Computing; IAPR - International
Association for Pattern Recognition; IEEE - Institute of Electrical and Electronics Engineers;
SSAB - Swedish Society for Automated Image Analysis. Her research interests are image
processing and analysis, genetic algorithms and active contour models.
Li Jun is a Ph.D. student at the Dept. of Technology, Örebro University in the AASS
Learning Systems Lab. He received the B.S. and M.S. degree from Gui Zhou University of
Technology, Gui Zhou, China, in 1982 and 1985, respectively.
His current research interest includes theory and application of artificial neural network
learning systems, specifically, the learning algorithms for autonomous mobile robot systems.
The objective is to provide adaptive on-line learning mechanism and algorithms inspired by
biological learning theory to study the mobile robot behaviors such as goal finding, obstacle
avoiding, target approaching, and wall following etc.
Grzegorz Cielniak is a Ph.D. student at the Dept. of Technology, Örebro University in the
AASS Learning Systems Lab. He was born in Krapkowice, Poland in 1975. He received his
MSc degree in robotics in 2000 at the Wroclaw University of Technology, Wroclaw, Poland.
Since January 2001 he has been a graduate student at AASS. His current research interests
include artificial intelligence, robot learning and vision systems.

64 AASS – Center for Applied Autonomous Sensor Systems
Henrik Andreasson is a Ph.D. Student at the Dept. of Technology, Örebro University in the
AASS Learning Systems Lab since October 2002. He received his M.Sc. degree in
Mechatronics at the Royal Institute of Technology, Stockholm, Sweden in 2001. His master
thesis were done at the Autonomous System Laboratory at the University of Hawaii, USA.
Since October 2002 he has been a graduate student at AASS. His current research interests are
robotics vision and learning. He is currently working on defining his Ph.D. topic and research
directions for 2003.
Per Munkevik is a joint Ph.D. student between the AASS Learning Systems Lab at the Dept.
of Technology, Örebro University, and SIK, Gothenburg since November 2002. He received
the M.Sc. degree in computer science in 2002 at the Luleå University of Technology. He is
currently working on defining his Ph.D. topic and research directions for 2003 within the SIK
project “the Food Factory of the Future”.

Annual Report 2002 65
3. Publications
3.1 Books and edited volumes
1. A. Birk, S. Coradeschi and S. Tadokoro, Eds. RoboCup
2001: Robot Soccer World Cup V. Springer, Germany,
2002.
3.2 Chapters in edited volumes
1. Saffiotti, A. Bjorklund, S. Johansson, and Z. Wasik. Team
Sweden. In: A. Birk, S. Coradeschi and S. Tadokoro, eds,
RoboCup 2001: Robot Soccer World Cup V (Springer,
Germany, 2002) pp. 725-729.
2. S. Johansson and A. Saffiotti. Using the Electric Field Approach in the RoboCup
Domain. In: A. Birk, S. Coradeschi and S. Tadokoro, eds, RoboCup 2001: Robot Soccer
World Cup V (Springer, DE, 2002) pp. 399-404.
3. S. Coradeschi and A. Saffiotti. Perceptual Anchoring: a key concept for plan execution in
embedded systems. In: M Beetz et al (eds) Advances in Plan-Based Control of Robotic
Agents. LNAI 2466 (Springer, DE, 2002) pp. 89-105.
4. L. Karlsson and Tommaso Schiavinotto. Progressive planning for mobile robots: a
progress report. In: M Beetz et al (eds) Advances in Plan-Based Control of Robotic
Agents. LNAI 2466 (Springer, DE, 2002) pp. 106-122.
3.3 Journals
1. P. Bergsten, R. Palm, D. Driankov. Observers for Takagi-Sugeno fuzzy systems. In
Systems, Man and Cybernetics, Part B, IEEE Transactions on, v.32, no. 1, Feb. 2002, p.
114 -121.
2. E. Fabrizi and A. Saffiotti. Augmenting Topology-Based Maps with Geometric
Information. Robotics and Autonomous Systems 40(2):91-97, 2002.
3. Tom Duckett and Stephen Marsland and Jonathan Shapiro. Fast, On-line Learning of
Globally Consistent Maps. Autonomous Robots Vol. 12, No. 3, pp. 287-300, 2002.
4. Eskiizmirliler S., Forestier N., Tondu B., Darlot C., "A model of the cerebellar pathways
applied to the control of a single joint robot arm actuated by Mckibben artificial muscles",
Biological Cybernetics, 86, 379-394, 2002.
5. J.Spännar, P.Wide, Estimation of a Nonmeasurable Variable by Using Grey Box
Modeling, Journal of IEEE Instrumentation and Measurement, December 2002.

66 AASS – Center for Applied Autonomous Sensor Systems
6. G.A. Fodor, L. Jonsson. New Software Architecture for flatness Control, Aluminium
World (Journal), Issue 2, Nr. 1, 2002
3.4 Conference articles
1. M. Persson, D. Driankov. SEAVS - a 3D simulation environment for aerial visual
servoing applications. In Proc. 1st Int. NAISO Congress on Autonomous Intelligent
Systems, Geelong, Australia, 12-15 Febr. 2002.
2. M. Persson. Visual servoing based tracking for an UAV in a 3D simulation environment.
In Proc. AEROSENSE, SPIE's 16th Annual Int. Symp. On Aerospace/Defense Sensing,
Simulation and Control, Orlando, USA, 1-5 April, 2002.
3. Tolt, G. and I. Kalaykov. Fuzzy-Similarity-Based Image Noise Cancellation. In: N. R. Pal,
M. Sugeno (Eds.): Advances in Soft Computing – AFSS 2002, Proc. of 2002 AFSS
International Conference on Fuzzy Systems, Calcutta, India, February 3-6, 2002, Lecture
Notes in Computer Science 2275, Springer 2002, p. 408-413.
4. Kalaykov, I. and G. Tolt. Fast Fuzzy Signal and Image Processing Hardware. Proc. of the
NAFIPS-FLINT 2002 Conference of the North American Fuzzy Information Processing
Society, New Orleans, USA, June 2002 27-29, pp.7-12.
5. Kalaykov. I. Fast fuzzy hardware - it's so simple! Plenary speech at 3rd International
NAISO Symposium on Engineering of Intelligent Systems, Malaga, Spain, September
2002, pp.24 - 27.
6. A. Ananiev, D. Ignatova. An analysis of the Technological Contact at the Robotized
grinding of parts with complex shapes. In Proceedings of the 33 rd ISR (International
Symposium on Robotics). October 7 - 11, 2002
7. Ananiev, A., I. Kalaykov and B. Iliev. An Approach to the design of light-weight
reconfigurable robot arm for a mobile robot. Proc. International Symposium on Robotics,
ISR 2002, Stockholm, Sweden, Oct 7-11, 2002.
8. Iliev, B. and I. Kalaykov (2002) Improved sliding mode robot control – a fuzzy approach,
Preprints of the 3rd IEEE International Workshop on Robot Motion and Control,
RoMoCo'02, November 9-11, Bukowy Dworek, Poland.
9. Boyko Iliev, Minimum-time sliding mode control of robot manipulators, Licentiate
Disseration, Oerebro Studies in Technology 2, Universitetsbiblioteket, ISSN 1650-8580,
ISBN 91-7668-304-4, Oerebro 2002.
10. B. Iliev and I. Hristozov. Variable structure control using Takagi-Sugeno fuzzy system as
a sliding surface" In Proc. IEEE International Conference on Fuzzy Systems FUZZ-
IEEE'2002, May 12-17, 2002, Honolulu, USA

Annual Report 2002 67
11. Tyrsa, V., L. Burtseva, V. Tyrsa, I. Kalaykov and A. Ananiev Three-Dimensional Vision
System. Proc. of SPIE International Conf. On Optomechatronics Systems III, Stuttgart,
Germany, November 2002, 12-14, vol. 902, pp. 645-651.
12. Tolt, G. (2002) Image noise reduction based on fuzzy similarity, Proc. Of Promote IT
2002 Conference organized by KK-foundation, Skövde, Sweden, no.2, pp. 257-262.
13. A. Loutfi, S. Coradeschi, and P. Wide. An introductory concept to human like odor
descriptions by grounding linguistic symbols. Proc. of the IEEE Instrumentation and
Measurement Technology Conference. Anchorage, Alaska, USA, 2002.
14. A. Loutfi and S. Coradeschi. Relying on an electronic nose for odor localization. Proc of
the IEEE Virtual Instrumentation and Measurement Systems Conference. Anchorage,
Alaska, USA, 2002.
15. A. Loutfi and S. Coradeschi. Symbolically describing the sensory perceptions from an
artificial nose. Proc. of the Artificial Intelligence Research Symposium. Lyon, FR, 2002.
16. O. Pettersson, L. Karlsson, and A. Saffiotti. Steps towards model-free execution
monitoring on mobile robots. Proc. of the 2nd Swedish Workshop on Autonomous
Robots (SWAR). Stockholm, Sweden, 2002, pp. 45-52.
17. Z. Wasik and A. Saffiotti. A Fuzzy Behavior-Based Control System for Manipulation.
Proc. of the IEEE/RSJ Intl. Conf. on Intelligent Robots and Systems (IROS). Lausanne,
CH, 2002.
18. P. Buschka and A. Saffiotti. A Virtual Sensor for Room Detection. Proc. of the IEEE/RSJ
Intl. Conf. on Intelligent Robots and Systems (IROS). Lausanne, CH, 2002, pp. 637-642.
19. I. Bloch and A. Saffiotti. Why Robots Should Use Fuzzy Mathematical Morphology.
Proc. of the 1st Int. ICSC-NAISO Congress on Neuro-Fuzzy Technologies. La Havana,
Cuba, 2002.
20. I. Bloch and A. Saffiotti. On the Representation of Fuzzy Spatial Relations in Robot
Maps. Proc. of the 9th Int. Conf. on Information Processing and the Management of
Uncertainty (IPMU). Annecy, France, 2002,
21. I. Bloch and A. Saffiotti. Les robots et la morphologie mathematique floue. Rencontres
Francophones sur la Logique Floue et ses Applications (LFA). Montpellier, France, 2002,
pp. 145-152.
22. Z. Wasik and A. Saffiotti. Robust Color Segmentation for the RoboCup Domain. Proc.
of the Int. Conf. on Pattern Recognition (ICPR). Quebec City, Quebec, CA, 2002.
23. Li Jun and Tom Duckett, Learning Robot Behaviours with Self-Organizing Maps and
Radial Basis Function Networks, Proc. SWAR'02, Second Swedish Workshop on
Autonomous Robotics, Stockholm, Sweden, October 10-11, 2002.

68 AASS – Center for Applied Autonomous Sensor Systems
24. Ahmed Mohamod Farah and Tom Duckett, Reactive Localisation of an Odour Source by
a Learning Mobile Robot, Proc. SWAR'02, Second Swedish Workshop on Autonomous
Robotics, Stockholm, Sweden, October 10-11, 2002.
25. Lucia Ballerini and Leonardo Bocchi, Segmentation of Liver Images by Texture and
Genetic Snakes, Proc. 2nd European Medical and Biological Conference, Vienna, Austria,
December 4-8, 2002.
26. A. Loutfi, P.Wide. Symbolic estimation of food odors using fuzzy techniques. In
Information Processing and Management of Uncertainty. Annecy, France, 2002.
27. L. Biel, P. Wide. An Intelligent Model Approach for Combination of Sensor Information.
In Have 2002 - IEEE International Workshop on Haptic Virtual Environments and their
Applications. Nov 2002.
28. A. Loutfi, "In what way can autonomous sensor systems duplicate human senses. An
example of the electronic nose.", Sinnena, Konsten och vetenskapen del 2, 16-17 January
2003, Örebro, Sweden.
29. J.L. Grantner, G.A. Fodor. Fuzzy Automation for Intelligent Hybrid Control Systems,
Proc, of the 2002 World Congress on Computational Intelligence, FUZZ-IEEE 2002,
Hawaii, May 12-17, 2002
30. G.A. Fodor. Fault Detection, Isolation and Recovery (FDIR) in Distributed Agent
Systems – A Domain Independent Approach. Proc. and Invited Talk, Net. Object Days
Conference, Erfurt, 2002
31. L. Biel, P. Wide. Active Perception in a Sensor Fusion Model, Proc, of SPIE – Sensor
Fusion: Architectures, Algorithms, and Applications VI. Orlando, USA, 3-5 April 2002.
3.5 Technical Reports
1. Peter Wide, Malin Lindquist, Kvalitetskontroll av dricksvatten med hjälp av elektronisk
tunga, 1428 dagar med U2000, 42-43, 2002.
3.6 Patents
1. Kalaykov, I., P. Wide and D. Driankov. Method and apparatus for generating degrees of
membership in fuzzy inference. World Intellectual Property Organization (WIPO) patent
WO-02/052501, priority date 27.12.2000.(Published 2002)
2. Patent application 02 01 782-0, Förfarande och arrangemang relaterade till en degbildande
sekvens.