Intelligent Systems: Safety, Reliability and Maintainability Issues

356 Pages·1993·23.47 MB·English

by  A. Meystel (auth.)

Intelligent Systems:
Safety, Reliability
and Maintainability Issues
NArD ASI Series
Advanced Science Institutes Series
A series presenting the results of activities sponsored by the NA TO Science
Committee, which aims at the dissemination of advanced scientific and
technological knowledge, with a view to strengthening links between
scientific communities.
The Series is published byan international board of publishers in
conjunction with the NATO Scientific Aftairs Division
A Life Sciences Plenum Publishing Corporation
B Physics London and New York
C Mathematical and Kluwer Academic Publishers
Physical Sciences Dordrecht, Boston and London
D Behavioural and
Social Sciences
E Applied Sciences
F Computer and Springer-Verlag
Systems Sciences Berlin Heidelberg New York
G Ecological Sciences London Paris Tokyo Hong Kong
H Cell Biology Barcelona Budapest
I Global Environmental
Change
NATO-PCO DATABASE
The electronic index to the NATO ASI Series provides full bibliographical
references (with keywords and/or abstracts) to more than 30000
contributions from international scientists published in aii sections of the
NA TO ASI Series. Access to the NA TO-PCO DAT ABASE compiled by the
NATO Publication Coordination Oftice is possible in two ways:
-via online FILE 128 (NATO-PCO DATABASE) hosted by ESRIN,
Via Galileo Galilei, 1-00044 Frascati, Italy.
-via CD-ROM "NATO Science & Technology Disk" with user-friendly
retrieval software in English, French and German (© 'NTV GmbH and
DATAWARE Technologies Inc. 1992).
The CD-ROM can be ordered through any member of the Board of
Publishers ar through NATO-PCO, Overijse, Belgium.
Series F: Computer and Systems Sciences VoI. 114
Intelligent Systems:
Safety, Reliability
and Maintainability Issues
Edited by
Okyay Kaynak
Department of Electrical and Electronic Engineering
Bogazi<ti University, 80815 Bebek, Istanbul, Turkey
Ger Honderd
Department of Electrical Engineering, TU Delft, Mekel Weg 4
POB 5031, 2600 GA Delft, The Netherlands
Edward Grant
Department of Computer Science, University of Strathclyde
Glasgow G1 1Xa, Scotland
Springer-Verlag Berlin Heidelberg GmbH
Proceedings of the NATD Advanced Research Workshop on Intelligent Systems:
Safety, Reliability and Maintainability Issues, held in Kusadasi, Izmir, Turkey,
August 25-28, 1992
CR Subject Classification (1991): 8.1.3, F.1.1, C.3
ISBN 978-3-642-63438-3 ISBN 978-3-642-58021-5 (eBook)
DOI 10.1007/978-3-642-58021-5
This work is subject to copyright. AII rights are reserved, whether the whole or part of the material is
concerned, specifically the rights oftranslation, reprinting, reuse of iIIustrations, recitation, broadcast
ing, reproduction on microfilms or in any other way, and storage in data banks. Duplication of this
publication or parts thereof is permitted only under the provisions of the German Copyright Law of
September 9, 1965, in its current version, and permission for use must always be obtained from
Springer-Verlag. Violations are liable for prosecution under the German Copyright Law.
© Springer-Verlag Berlin Heidelberg 1993
Originally published by Springer-Verlag Berlin Heidelberg New York in 1993
Softcover reprint of the hardcover 15t edition 1993
Typesetting: Camera ready by authors
45/3140 -54321 0-Printed on acid-free paper
PREFACE
This book is a collection of some of the papers that were presented during a NATO
Advanced Research Workshop (ARW) on "Intelligent Systems: Safety, Reliability
and Maintainability Issues" that was held in Kusadasi, Turkey during August 24-
28, 1992. Attendance at this workshop was mainly by invitation only, drawing
people internationally representing industry, government and the academic
community. Many of the participants were internationally recognized leaders in the
topic of the workshop.
The purpose of the ARW was to bring together a highly distinguished group of
people with the express purpose of debating where the issues of safety, reliability
and maintainability place direct and tangible constraints on the development of
intelligent systems. As a consequence, one of the major debating points in the
ARW was the definition of intelligence, intelligent behaviour and their relation to
complex dynamic systems. Two major conclusions evolved from the ARW are:
1. A continued need exists to develop formal, theoretical frameworks for
the architecture of such systems, together with a reflection on the
concept of intelligence.
2. There is a need to focus greater attention to the role that the human play
in controlling intelligent systems.
The workshop began by considering the typical features of an intelligent system.
The complexity associated with multi-resolutional architectures was then discussed,
leading to the identification of a necessity for the use of a combinatorial
synthesis/approach. This was followed by a session on human interface issues.
Both these sessions indicated that intelligence in human terms is difficult to
quantify. When designing architectures, they can only be represented formally if
enough human expertise is elicited to allow a formal representation at each level. In
terms of the interface, much of the high level abstraction is left to the human
operation and decisions are based on low-level sensing associated with intelligence.
VI
Numerous other technical areas associated with intelligence were addressed in eight
sessions. These included expert system development and use, virtual reality,
interactive graphics and adaptive non-linear control. The application domains for all
these technologies ranged from robotics and computer aided manufacturing to
cement production, nuclear plant control and air-traffic control. The sessions
therefore reflected the spectrum of work being conducted worldwide.
The conclusion reached unanimously after four days was that there should be close
dialogues between the formal and application communities. The frameworks
developed by the application community should be presented formally with the
goal of incrementally moving towards a general and genetic formal representation
of control architectures. Finally, it became obvious that a better understanding was
required of the role of the human-in-the-loop control systems.
The papers presented during the workshop are organized in six parts in the
following pages. The reader will find that there has been an effort to balance the
theory and the applications; some papers provide introductory and overview
material as well as reviews of the state of the art and yet some others provide
enough detail in specialized areas which can be useful to practitioners in the field.
Part 1 is on the analysis and assessment issues related to intelligent control
systems. The first two papers by the same author discuss the complexity and
reliability of multi-resolutional architectures and propose new control solutions.
The third paper describes a software tool for reliability and safety assessment of
control systems.
In Part 2, the dependability of intelligent systems is discussed where the word
dependability encompasses all aspects of system safety, reliability, availability and
security, as discussed in the first paper of the part. The other papers address the
questions on how to handle intelligent systems and draw the attention to possible
shortcomings of such systems.
Papers on intelligent front ends are grouped in Part 3. A graphical front-end for
grasp planning and programming robots through visual dialogues are some
examples to the topics treated in this part. The section concludes with a paper that
explores a new approach for inspection applications.
The main concern of Part 4 is learning systems. The use of fractal set theory and
neural networks in learning systems are among the topics discussed. Attention is
also focused on qualitative reasoning.
Part 5 is robotics oriented. The papers presented in this part discuss the safety and
reliability of advanced/intelligent robotic systems. Also included in the part is a
paper describing the activities of UK Advanced Robotics Research Center.
VII
Finally the book concludes with a part that presents some examples of industrial
applications of intelligent systems with particular emphasis on reliability. The ftrst
two papers of this part are by the same author and discuss deep knowledge and
rigorous engineering practice and describe a real time intelligent system for critical
applications. A hierarchically structured multi-expert system controller for cement
production is described in the last paper of the part and the book.
It should here be pointed out that this book does not and cannot reflect all that took
place during the workshop. The lively discussions that were carried out, both
formally in the meeting room and informally by the beach or the pool were just as
important as the papers themselves. It is rather unfortunate that discussions of this
kind cannot be compiled and printed for the beneftt of the scientiftc community. On
behalf of those who enjoyed and benefited from these discussions so much, I would
like to thank, once again through this publication, all the participants for being so
dynamic and participative.
Okyay Kaynak
Workshop Director
Bogazici University, Istanbul
nl
Marmara Research Center, Gebze
CONTENTS
Part 1 Intelligent Systems : Analysis and Assessment Issues
Reliability of Intelligence: Issues and Conjectures . 1
A. Meystel
Multiresolutional System : Complexity and Reliability . .11
A. Meystel
MARSA -the Software Tool for Reliablity and Safety Assessment of Control
Systems ........................ 23
T. Nikolov, C. Christov
Part 2 Dependability of Intelligent Systems
Safety Aspects of Security Models . . . . 31
T. Beth
The Intelligent Use of Intelligent Systems .42
E. Hollnagel
Intelligent Systems: The Weakest Link ? . .60
D. Davenport
Part 3 Intelligent Front Ends
Developing Hypermedia Front-Ends for Robot Teleoperation Tasks. . 74
E. Grant, G. K. Adam, B. A. Shepherd
Automated Inspection and Reverse Engineering. . . . . . . . . 95
T. M. Sobh, C. Jaynes, M. Dekhil, T. Henderson
Development of the Intelligent Graphical Front-End for Robot Grasp
Planning . . . . . . . . . . . . . . . . . . . . . . . . 123
A. Kasinski
x
Part 4 Learning Systems
Another Fonnal Approach to Modelling Inference Engine Behaviour . 137
S. Nilsen
Fractal Set Theoretic Analysis of Perfonnance Losses for Tuning Training
Data in Learning Systems . . . . . . . . . . . . . . . . . . 157
A. M. Erkmen
Automatic Construction of Qualitative Models. . . . . . . . . . . 177
A. C. C. Say. S. Kuru
Single and Multiple Memory-based Neural Networks for Efficient Function
Approximation . . . . . . . . . . . . . . . . . . . .. 194
E.Alpaydm
Robustness and Fault Tolerance Issues in Artificial Neural Network Based
Robotic Control Applications. . . . . . . . . . . . . . . . . 205
K. Czl,Z
Part 5 Safety and Reliability of Advanced Robotic Systems
Human Factors in Robotic Systems . 221
B. S. Dhillon
Advanced Robotics -Towards Intelligent Machines . . . . . . . . . 232
J. o. Gray
Intelligent Motion Planning for Safe and Reliable Control of Robot Anns . 243
T-J. Tarn. N. Xi. A. K. Bejczy
Control Issues of the Delft Intelligent Assembly Cell with Emphasis on
Geometrical Trajectory Planning . . . . . . . . . . . . . . . . 257
G. Honderd. T. A. Rieswijk. W. Jongkind
Part 6 Dependability Considerations in Industrial Applications
Deep Knowledge and Rigorous Engineering Practice: Two Ingredients to Build
Dependable Knowledge Based Systems for Industrial Process Plant Control . 285
S. Bologna. H. Viilisuo
Architectural Foundations, Concepts, and Methods Behind ISACS -Real-Time
Intelligent System for Critical Applications . . . . . . . . . . . . 307
S. Bologna. (2j. Berg. K. Haugset. J. Kvalem
XI
Intelligent Industrial Control . . . . ........... 327
R. E. King
Modelling of the Process of Air Traffic Control Under the Expert System
Management .......... ........... 333
K. Bonchev. K. Zografov