Lecture Notes in Artificial Intelligence
Edited by J. G. Carbonell and J. Siekmann
Subseries of Lecture Notes in Computer Science
2699
3
Berlin
Heidelberg
New York
Hong Kong
London
Milan
Paris
Tokyo
Michael G. Hinchey James L. Rash
Walter F. Truszkowski Christopher Rouff
Diana Gordon-Spears (Eds.)
Formal Approaches
toAgent-Based Systems
Second International Workshop, FAABS 2002
Greenbelt, MD, USA, October 29-31, 2002
Revised Papers
13
Series Editors
Jaime G. Carbonell, Carnegie Mellon University, Pittsburgh, PA, USA
Jörg Siekmann, University of Saarland, Saarbrücken, Germany
Volume Editors
Michael G. Hinchey
James L. Rash
Walter F. Truszkowski
NASA Goddard Space Flight Center
Mailstop 588.0
Greenbelt, MD 20771, USA
Christopher Rouff
SAIC
1710 SAIC Drive
McLean, VA 22102, USA
Diana Gordon-Spears
University of Wyoming
Computer Science Department
Laramie, WY 82070, USA
Cataloging-in-Publication Data applied for
A catalog record for this book is available from the Library of Congress.
Bibliographic information published by Die Deutsche Bibliothek
Die Deutsche Bibliothek lists this publication in the Deutsche Nationalbibliografie;
detailed bibliographic data is available in the Internet at <http://dnb.ddb.de>.
CR Subject Classification (1998): I.2.11, I.2, D.2, F.3, I.6, C.3, J.2
ISSN 0302-9743
ISBN 3-540-40665-4 Springer-Verlag Berlin Heidelberg New York
This work is subject to copyright. All rights are reserved, whether the whole or part of the material is
concerned, specifically the rights of translation, reprinting, re-use of illustrations, recitation, broadcasting,
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 New York
a member of BertelsmannSpringer Science+Business Media GmbH
http://www.springer.de
© Springer-Verlag Berlin Heidelberg 2003
Printed in Germany
Typesetting: Camera-ready by author, data conversion by PTP-Berlin GmbH
Printed on acid-free paper
SPIN: 10928011
06/3142
543210
Preface
The idea of a FAABS workshop was first conceived in 1998 at the NASA Goddard
Space Flight Center, while the Agent Technology Development Group in the
Advanced Architectures and Automation Branch (Code 588) was developing a
prototype agent community to automate satellite ground operations.
While developing this system, several race conditions arose within and between
agents. Due to the complexity of the agents and the communications between them, it
was decided that a formal approach was needed to specify the agents and the
communications between them, so that the system could be checked for additional
errors.
A formal model of the inter-agent communications was developed, with the
expectation that this would enable us to find more errors. Success in this convinced us
of the importance of using formal methods to model agent-based systems. To share
our own experiences and to learn how others were approaching these issues, we
decided to hold a workshop on formal methods and agent-based systems.
The response was overwhelming. The result was the first FAABS workshop, which
was held at the NASA Goddard Space Flight Center. Posters, paper presentations,
panels, and an invited talk by J Moore stimulated much discussion and subsequent
collaboration.
This proceedings contains papers from FAABS-II, the second workshop held at the
Greenbelt Marriott Hotel (near the NASA Goddard Space Flight Center) in October
2002 and sponsored in conjunction with the IEEE Computer Society. Participants
from around the world joined together to present papers and posters, participate in
panels, and hear an enlightening invited presentation by Prof. Sir Roger Penrose.
We would like to express our sincere thanks to all those who attended the
workshop, presented papers or posters, and participated in panel sessions and both
formal and informal discussions. Our thanks to NASA Goddard Code 588 and Code
581 (Software Engineering Laboratory), the Naval Research Laboratory, and CTA,
Inc. for their financial support and to the IEEE Computer Society for their
sponsorship of this event. Thanks also to Springer-Verlag for once again publishing
the proceedings.
We trust that the reader will find this compilation to be of interest, and we look
forward to welcoming some of you to FAABS-III, tentatively planned for early 2004.
Greenbelt, MD
May 2003
Organizing Committee
Mike Hinchey, NASA Goddard Space Flight Center
Jim Rash, NASA Goddard Space Flight Center
Walt Truszkowski, NASA Goddard Space Flight Center
Chris Rouff, SAIC
Diana Gordon-Spears, University of Wyoming
Table of Contents
“What Is an Agent and Why Should I Care?” . . . . . . . . . . . . . . . . . . . . . . . .
Tim Menzies, Adrian Pearce, Clinton Heinze, Simon Goss
1
Organising Logic-Based Agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Michael Fisher, Chiara Ghidini, Benjamin Hirsch
15
A Statechart Framework for Agent Roles that Captures Expertise
and Learns Improved Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Bahram Kimiaghalam, Abdollah Homaifar, Albert C. Esterline
28
Formal Specification of Interaction in Agent Societies . . . . . . . . . . . . . . . . . .
Virginia Dignum, John-Jules C. Meyer, Frank Dignum, Hans Weigand
37
Formal Verification for a Next-Generation Space Shuttle . . . . . . . . . . . . . . .
Stacy D. Nelson, Charles Pecheur
53
Automated Protocol Analysis in Maude . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Jeffrey Van Baalen, Thomas Böhne
68
Towards Interaction Protocol Operations for Large Multi-agent
Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Joaquı́n Peña, Rafael Corchuelo, José Luis Arjona
79
Formal Modeling and Supervisory Control of Reconfigurable Robot
Teams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Kiriakos Kiriakidis, Diana F. Gordon-Spears
92
Computational Models for Multiagent Coordination Analysis:
Extending Distributed POMDP Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
Hyuckchul Jung, Ranjit Nair, Milind Tambe, Stacy Marsella
Bounded Model Checking for Interpreted Systems: Preliminary
Experimental Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
A. Lomuscio, T. L
3 asica, W. Penczek
Verifiable Middleware for Secure Agent Interoperability . . . . . . . . . . . . . . . . 126
Ramesh Bharadwaj
Distributed Implementation of a Connection Graph Based on
Cylindric Set Algebra Operators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
Silvana Zappacosta Amboldi
Using Statecharts and Modal Logics to Model Multiagent Plans and
Transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
Albert C. Esterline
VIII
Table of Contents
Qu-Prolog: An Implementation Language for Agents with Advanced
Reasoning Capabilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
Peter J. Robinson, Mike Hinchey, Keith Clark
A Model for Conformance Testing of Mobile Agents in a MASIF
Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
Mikaël Marche, Yves-Marie Quemener
Analysis of a Phase Transition in a Physics-Based Multiagent System . . . . 193
Diana F. Gordon-Spears, William M. Spears
You Seem Friendly, But Can I Trust You? . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208
Tim Menzies, David Owen, Bojan Cukic
Taking Intelligent Agents to the Battlefield . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
Jeffrey Hicks, Richard Flanagan, Plamen Petrov, Alexander Stoyen
Panel Session on “Applications”
Naval Applications of Secure Multi-agent Technology . . . . . . . . . . . . . . . . . . 235
Ramesh Bharadwaj
Challenges Arising from Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
Charles Pecheur
Agents Applied to Autonomous Vehicles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
Craig Schlenoff
Using XML for Interprocess Communications in a Space Situational
Awareness and Control Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
Stuart Aldridge, Alexander Stoyen, Jeffrey Hicks,
Plamen Petrov
Panel Session on “Asimov’s Laws”
Asimov’s Laws: Current Progress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
Diana F. Gordon-Spears
Asimov’s Laws . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
James P. Hogan
On Laws of Robotics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264
Yoji Kondo
Panel Session on “Tools and Education”
Challenges Arising from Applications of Agent-Based System . . . . . . . . . . . 269
Walt Truszkowski
Table of Contents
IX
Tools and Education towards Formal Methods Practice . . . . . . . . . . . . . . . . 274
John-Jules C. Meyer
Poster Presentations
Evaluating Agent-Based Modeling as a Tool for Economists . . . . . . . . . . . . 283
Margo Bergman
Modeling Traffic Control through Deterrent Agents . . . . . . . . . . . . . . . . . . . . 286
Michel Rudnianski, Hélène Bestougeff
Towards a Formal Representation of Driving Behaviors . . . . . . . . . . . . . . . . 290
Craig Schlenoff, Michael Gruninger
Formal Analysis of an Agent-Based Medical Diagnosis Confirmation
System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 292
Alexander Hoole, Issa Traore, Michael Liu Yanguo
Agent Programming in Dribble: From Beliefs to Goals with Plans . . . . . . . 294
Birna van Riemsdijk, Wiebe van der Hoek, John-Jules C. Meyer
Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297
“What Is an Agent and Why Should I Care?”
Tim Menzies1 , Adrian Pearce2 , Clinton Heinze3 , and Simon Goss3
1
2
Lane Department of Computer Science, West Virginia University,
PO Box 6109, Morgantown, WV, 26506-6109, USA, [email protected]
Department of Computer Science and Software Engineering The University of Melbourne,
Victoria, 3010, Australia, [email protected]
3
Air Operations Division, Aeronautical & Maritime Research Laboratory, Melbourne,
Australia, clinton.heinze|[email protected]
Abstract. A range of agent implementation technologies are reviewed according
to five user-based criteria and via a comparison with object-oriented programming. The comparison with OO shows that some parts of object technology are a
candidate implementation technique for some parts of agent systems. However,
many other non-object-based implementation techniques may be just as useful.
Also, for agents with mentalistic attitudes, the high-level specification of agent
behavior requires numerous concepts outside the object paradigm; e.g. plans,
communication, intentions, roles, and teams.
Keywords: Evaluation, agent-oriented, object-oriented.
1
Introduction
Is there anything really new in agent-oriented software? Are agents a bold step forward
into the future of software? Or is agency just “new wine in old bottles”?
Our users demand answers to these questions, and others. One gruff user always
asked “what are agents and why should I care?”. To such users, the issue in italics is
the key question. Agent technologies are interesting to users only if those technologies
address issues of interest to the users.
After explaining agents to this gruff user, this users next comment was “this sounds
just like OO to me; what’s new here?”. Such comments motivate this article. Our response
to these comments is in three parts:
1. We carefully define the core concepts of agent-oriented software and object-oriented
software.
2. Next, we review the diverse range of software labelled “agents”.
3. This software is then assessed these concepts with respect to certain user-oriented
issues.
The user issues used in this article come from the Australian Workshops on AgentBased systems. Those workshops have debated the relative merits of the agent implementation technologies shown in Figure 1. In those debates, the technologies were assessed
with respect to the problem of building agents for the Air Operations Division (AOD)
of the Australian Defense Science Technology Division.
M.G. Hinchey et al. (Eds.): ’FAABS 2002, LNAI 2699, pp. 1–14, 2003.
c Springer-Verlag Berlin Heidelberg 2003
2
T. Menzies et al.
Name : Notes
Introduced in...
OO : Object-oriented
§2.1
Standard BDI : BDI= beliefs, desires, intentions
§2.2
FORTRAN : How we used to build agents
§3.1
dMARS : A commercial agent-oriented BDI tool
§3.1
Command agents : Heinze and Pearce’s extension to dMARS
§3.1
Behavioural cloning : Machine learning to build agents
§3.1
Petri nets :
§3.3
TACAIR/ SOAR/ PSCM : The problem space computational model (PSCM) is
§3.4
how the rule-based system called SOAR implements
TACAIR, an agent system.
G2 : Gensym’s rule-based expert system shell: includes
§3.5
powerful interface tools.
MBD-based : The model-based diagnosis system used in NASA’s re§3.6
mote agent experiment (RAX).
Fig. 1. Agent implementation technologies discussed in this article
For several years, the Australian Defense Forces have been using agent-oriented
software to assess potential new hardware purchases. Buying planes and helicopters
for the Air Force implies a major commitment to a particular platform. AOD uses operational simulation for answering specific questions about very expensive equipment
requisitions, component capabilities and rehearsing dangerous tactical operations. In
pilot-in-the-loop flight simulation, intelligent pilots (agents) interact with each other in
the computer simulation, as well as the human pilot in the virtual environment. These
dynamic, interactive multi-agent simulations pose a challenge for the integration of valid
pilot competencies into computer controlled agents. This involves modeling pilot perception through recognition of actions and events that occur during simulation. Such
simulators are often used after purchase as training tools. Hence, a core task within
DSTO is the construction and maintenance of agent-oriented systems. These AOD agent
simulations push the state-of-the-art:
– AOD agents interact at high frequency in a dynamic environment with numerous
friendly and hostile agents. For example, AOD agents engage in complex aerial
maneuvers against hostile high-speed aircraft.
– AOD agents co-ordinate extensively to achieve shared goals. For example, a
squadron of fighters may collaborate to shepherd a cargo ship through enemy lines.
– AOD agents may change their roles at runtime. For example, if the lead of a fighter
formation is shot down, then the wing-man may assume the role of fighter lead. As
roles change, agents must dramatically alter their plans.
After discussions with AOD users, the following concerns were identified. These
concerns are the basis for our user-oriented discussion of the merits of different agent
technologies:
– Easy of construction/ modification.
– Provable reliability.