Abstract: To ease large-scale realization of agent applications there is an urgent need for frameworks, methodologies and toolkits that support the effective development of agent systems. Moreover, since one of the main tasks for which agent systems were invented is the integration between heterogeneous software, independently developed agents should be able to interact successfully. In this paper, we present JADE (Java Agent Development Environment), a software framework to build agent systems for the management of networked information resources in compliance with the FIPA specifications for inter-operable intelligent multi-agent systems. The goal of JADE is to simplify development while ensuring standard compliance through a comprehensive set of system services and agents. JADE can then be considered to be an agent middle-ware that implements an efficient agent platform and supports the development of multi-agent systems. It deals with all the aspects that are not peculiar to agent internals and that are independent of the applications, such as message transport, encoding and parsing, or agent life-cycle management. Copyright © 2001 John Wiley & Sons, Ltd.

Abstract: Agent Theories I.- Optimistic and Disjunctive Agent Design Problems.- Updating Mental States from Communication.- Sensing Actions, Time, and Concurrency in the Situation Calculus.- Agent Development Tools and Platforms.- Developing Multiagent Systems with agentTool.- Layered Disclosure: Revealing Agents' Internals.- Architectures and Idioms: Making Progress in Agent Design.- Developing Multi-agent Systems with JADE.- Agent Theories II.- High-Level Robot Control through Logic.- Determining the Envelope of Emergent Agent Behaviour via Architectural Transformation.- Models of Agent Communication and Coordination.- Delegation and Responsibility.- Agent Theory for Team Formation by Dialogue.- Task Coordination Paradigms for Information Agents.- Autonomy and Models of Agent Coordination.- Plan Analysis for Autonomous Sociological Agents.- Multiagent Bidding Mechanisms for Robot Qualitative Navigation.- Performance of Coordinating Concurrent Hierarchical Planning Agents Using Summary Information.- Agent Languages.- Agent Programming with Declarative Goals.- Modeling Multiagent Systems with CASL - A Feature Interaction Resolution Application.- Generalised Object-Oriented Concepts for Inter-agent Communication.- Specification of Heterogeneous Agent Architectures.- Planning, Decision Making, and Learning.- Improving Choice Mechanisms within the BVG Architecture.- Planning-Task Transformations for Soft Deadlines.- An Architectural Framework for Integrated Multiagent Planning, Reacting, and Learning.- Panel Summary: Agent Development Tools.- Panel Summary: Agent Development Tools.- Panel Summary: Autonomy -Theory, Dimensions, and Regulation.- Again on Agents' Autonomy: A Homage to AlanTuring - Panel Chair's Statement.- Autonomy as Decision-Making Control.- Autonomy: Theory, Dimensions, and Regulation.- Situated Autonomy.- Autonomy: A Nice Idea in Theory.- Adjustable Autonomy: A Response.

Abstract: To solve complex problems, agents work cooperatively with other agents in heterogeneous environments. We are interested in coordinating the local behavior of individual agents to provide an appropriate system-level behavior. The use of intelligent agents provides an even greater amount of flexibility to the ability and configuration of the system itself. With these new intricacies, software development is becoming increasingly difficult. Therefore, it is critical that our processes for building the inherently complex distributed software that must run in this environment be adequate for the task. This paper introduces a methodology for designing these systems of interacting agents.

Abstract: The architecture of a multi-agent system can naturally be viewed as a computational organisation. For this reason, we believe organisational abstractions should play a central role in the analysis and design of such systems. To this end, the concepts of agent roles and role models are increasingly being used to specify and design multi-agent systems. However, this is not the full picture. In this paper we introduce three additional organisational concepts--organisational rules, organisational structures, and organisational patterns--that we believe are necessary for the complete specification of computational organisations.We view the introduction of these concepts as a step towards a comprehensive methodology for agent-oriented systems.

Abstract: This paper presents an approach for cooperative search of a team of distributed agents. We consider two or more agents, or vehicles, moving in a geographic environment, searching for targets of interest and avoiding obstacles or threats. The moving agents are equipped with sensors to view a limited region of the environment they are visiting, and are able to communicate with one another to enable cooperation. The agents are assumed to have some "physical" limitations including possibly maneuverability limitations, fuel/time constraints and sensor range and accuracy. The developed cooperative search framework is based on two inter-dependent tasks: (1) online learning of the environment and storing of the information in the form of a "search map"; and (2) utilization of the search map and other information to compute online a guidance trajectory for the agent to follow. The distributed learning and planning approach for cooperative search is illustrated by computer simulations.

Abstract: Foundations of Multi-agent Systems.- Perspectives on Organizations in Multi-agent Systems.- Multi-agent Infrastructure, Agent Discovery, Middle Agents for Web Services and Interoperation.- Logical Foundations of Agent-Based Computing.- Standardizing Agent Communication.- Standardizing Agent Interoperability: The FIPA Approach.- Distributed Problem Solving and Planning.- Automated Negotiation and Decision Making in Multiagent Environments.- Agents? Advanced Features for Negotiation and Coordination.- Social Behaviour, Meta-reasoning, and Learning.- Towards Heterogeneous Agent Teams.- Social Knowledge in Multi-agent Systems.- Machine Learning and Inductive Logic Programming for Multi-agent Systems.- Relational Reinforcement Learning.- From Statistics to Emergence: Exercises in Systems Modularity.- Emotions and Agents.- Applications.- Multi-agent Coordination and Control Using Stigmergy Applied to Manufacturing Control.- Virtual Enterprise Modeling and Support Infrastructures: Applying Multi-agent System Approaches.- Specialised Agent Applications.- Agent-Based Modelling of Ecosystems for Sustainable Resource Management.- Cooperating Physical Robots: A Lesson in Playing Robotic Soccer.- A Multi-agent Study of Interethnic Cooperation.

Abstract: Multi-agent systems can very naturally be viewed as computational organisations. For this reason, we believe organisational abstractions offer a promising set of metaphors and models that can be exploited in the analysis and design of such systems. To this end, the concept of role models is increasingly being used to specify and design multi-agent systems. However, this is not the full picture. In this paper we introduce three additional organisational concepts - organisational rules , organisational structures, and organisational patterns - and discuss why we believe they are necessary for the complete specification of computational organisations. In particular, we focus on the concept of organisational rules and introduce a formalism, based on temporal logic, to specify them. This formalism is then used to drive the definition of the organisational structure and the identification of the organisational patterns. Finally, the paper sketches some guidelines for a methodology for agent-oriented systems based on our expanded set of organisational abstractions.

Abstract: New techniques for grid monitoring, protection, and control have been perfected, and their judicious application can help reduce the frequency and severity of catastrophic failures. This article describes the development of an innovative multiagent approach to prevent and control catastrophic failures in large power systems. The research has created understanding of the origin and nature of catastrophic failures. This is achieved by an analysis of hidden failures in protection systems. Vulnerabilities associated with the power system, the information network, and the communication network are also evaluated. The approach is characterized by the extensive use of real-time information from diverse sources, coupled with the development of an evolving dynamic decision event tree. A novel multiagent based platform is used to evaluate system vulnerability to catastrophic events taking in account the market environment and competing entities. Several new concepts associated with wide-area measurements and controls, networked sensors, and adaptive self healing are used to reconfigure the network to minimize the system vulnerability. Approaches developed in this research will provide important solutions for power systems and other interconnected networks of the future.

Abstract: A new class of genetic algorithms (GA) is presented. It is based on the idea of outsourcing, a popular trend in business. In a human based genetic algorithm (HBGA), all primary genetic operators are outsourced, i.e. delegated to external human agents. A totally outsourced genetic algorithm uses both human evaluation and the human ability of innovation. It is a multi-agent environment and the mediator of communication between multiple heterogenous agents. The advantage of this approach is its ability to address complex problems for which it is hard, not only to evaluate individuals, but to find a good representation for them. These qualities allow HBGA to process flows of information without knowledge of its particular structure and representation. The suggested conceptual approach can also be used as a general model and a way of thinking about different kinds of genetic algorithms.

Abstract: Virtual enterprises paradigm represents an important application field for multi-agent approaches, both in terms of modeling and infrastructure development This article summarizes the main challenges in this field and describes several current Multi-Agent System application approaches. A particular emphasis is given to the creation and operation phases of the virtual enterprise life cycle. Several open challenges in this area are also introduced.

Abstract: In recent years, multi-agent systems (MASs) have received increasing attention in the artificial intelligence community. Research in multi-agent systems involves the investigation of autonomous, rational and flexible behaviour of entities such as software programs or robots, and their interaction and coordination in such diverse areas as robotics (Kitano et al., 1997), information retrieval and management (Klusch, 1999), and simulation (Gilbert & Conte, 1995). When designing agent systems, it is impossible to foresee all the potential situations an agent may encounter and specify an agent behaviour optimally in advance. Agents therefore have to learn from, and adapt to, their environment, especially in a multi-agent setting.

Abstract: Intelligent Agents are personal assistants which can provide proactive support to users by executing routine activities like searching on the Internet, the scheduling of meetings, etc. The concept of an Intelligent Agent has its roots in Artificial Intelligence and provides a basis for the construction of a new programming paradigm. A programm is being viewed within this paradigma as a pro-active agent that has specific knowledge to achieve its tasks, is goal-directed, is able to construct appropriate plans and possesses the capabilities to execute those plans. Unfortunately, the construction and the design of Intelligent Agents is ad-hoc and proceeds without a proper theoretical basis. In this thesis, a more structured approach is proposed that provides a programming framework to develop agents. More specific, we design two programming languages that are specifically targeted at the construction of intelligent agents. The basic agent concepts like knowledge, goals, plans and capabilities provide the basis for the language 3APL. One of the unique features of 3APL agents is that they are able to modify and revise their plan. 3APL agents thus are able to modify themselves and can be viewed as self-modifying programs. The noition of a goal in 3APL has an operationel meaning and is very similar to a plan or procedure that the agent knows how to execute. In the agent language called GOAL (goal-oriented agent language), a more descriptive notion of a goal is incorporated. A goal in this second agent language denotes the state of affairs that the agent wants to achieve. The power of both of the agent langauges developed in the thesis consists in the fact that the basic programming concepts are natural and intuitive, and therefore provide a basis to bridge the gap between our common-sense way of thinking and the way in which software engineers develop programs.

Abstract: In this paper we investigate the use of hierarchical reinforcement learning to speed up the acquisition of cooperative multi-agent tasks. We extend the MAXQ framework to the multi-agent case. Each agent uses the same MAXQ hierarchy to decompose a task into sub-tasks. Learning is decentralized, with each agent learning three interrelated skills: how to perform subtasks, which order to do them in, and how to coordinate with other agents. Coordination skills among agents are learned by using joint actions at the highest level(s) of the hierarchy. The Q nodes at the highest level(s) of the hierarchy are configured to represent the joint task-action space among multiple agents. In this approach, each agent only knows what other agents are doing at the level of sub-tasks, and is unaware of lower level (primitive) actions. This hierarchical approach allows agents to learn coordination faster by sharing information at the level of sub-tasks, rather than attempting to learn coordination taking into account primitive joint state-action values. We apply this hierarchical multi-agent reinforcement learning algorithm to a complex AGV scheduling task and compare its performance and speed with other learning approaches, including flat multi-agent, single agent using MAXQ, selfish multiple agents using MAXQ (where each agent acts independently without communicating with the other agents), as well as several well-known AGV heuristics like "first come first serve", "highest queue first" and "nearest station first". We also compare the tradeoffs in learning speed vs. performance of modeling joint action values at multiple levels in the MAXQ hierarchy.

Abstract: In order to generate plans for agents with multiple actuators, agent teams, or distributed controllers, we must be able to represent and plan using concurrent actions with interacting effects. This has historically been considered a challenging task requiring a temporal planner with the ability to reason explicitly about time. We show that with simple modifications, the STRIPS action representation language can be used to represent interacting actions. Moreover, algorithms for partial-order planning require only small modifications in order to be applied in such multiagent domains. We demonstrate this fact by developing a sound and complete partial-order planner for planning with concurrent interacting actions, POMP, that extends existing partial-order planners in a straightforward way. These results open the way to the use of partial-order planners for the centralized control of cooperative multiagent systems.

Abstract: Formations of multi-agent systems, such as satellites and aircraft, require that individual agents satisfy their kinematic equations while constantly maintaining inter-agent constraints. In this paper, we develop a systematic framework for studying formations of multiagent systems. In particular, we consider undirected formations for centralized formations and directed formations for decentralized formations. In each case, we determine differential geometric conditions that guarantee formation feasibility given the individual agent kinematics. Our framework also enables us to extract a smaller control system that describes the formation kinematics while maintaining,all formation constraints.

Abstract: Microscopic traffic simulators can model traffic flow in a realistic manner and are ideal for agent-based vehicle control. In this paper we describe a model of a reactive agent that is used to control a simulated vehicle. The agent is capable of tactical-level driving and has different driving styles. To ensure fast reaction times, the agent's driving task is divided into several competing and reactive behavior rules. The agent is implemented in and tested with a prototype traffic simulator program. The simulator consists of an urban environment with multi-lane roads, intersections, traffic lights, and vehicles. Every vehicle is controlled by a driving agent and all agents have individual behavior settings. Preliminary experiments show that the agents exhibit human-like behavior ranging from slow and careful to fast and aggressive driving behavior.

Abstract: Deploying intelligent agents to do peoples' bidding in environments ranging from Internet marketplaces to Mars has received much attention. Exactly what an agent is and in what sense a computational agent can behave intelligently remain the subject of considerable debate. However, most would agree that coordination, an agent's fundamental capability to decide on its own actions in the context of the activities of other agents around it, is a central concern of intelligent agency. The value of an intelligent agent coordination strategy lies in how well it scales along various dimensions of stress. Understanding the agent population, its task environment, and expectations about its collective behavior are central to mapping the space of potential approaches. The paper discusses agent coordination and dimensions of coordination stress.

Abstract: Agents, Multiagent Systems and Software Engineering.- Basic Concepts in Software Engineering.- The Conceptual Framework of Massive.- Massive Views.- Further Case Studies.- Conclusion.- Toolkits for Agent-Based Applications.- Basic Problem Solving Capabilities of TCS Agents.- Protoz Specification of the Contract-Net Protocol.

Abstract: Active learning is an e ective learning approach. In this paper, we present an intelligent agent assisted environment for active learning. The system is to better support studentcentered, self-paced, and highly interactive learning approach. Students' learning-related pro les, such as learning styles and background knowledge, are used in selecting, organizing, and presenting learning materials. A new approach to course content organization and delivery is being developed based on smart instructional components, which can be integrated into a wide range of courses. The system is being implemented using the prevalent Internet, Web, digital library, and multi-agent technologies.

Abstract: The emergence of ad-hoc pervasive connectivity for devices based on Bluetooth-like systems provides a new way to create applications for mobile systems. We seek to realize ubiquitous computing systems based on the cooperation of autonomous, dynamic and adaptive components (hardware as well as software) which are located in vicinity of one another. In this paper we present this vision. We also describe a prototype system we have developed that implements parts of this vision – in particular a system that combines agent oriented and service oriented approaches and provides dynamic service discovery. We point out why existing systems such as Jini are not suited for this task, and how our system improves on them.

Abstract: Positioning.- Multi-Agent Systems.- Coordination Models and Languages.- ECM and Its Instances.- The ECM Coordination Model.- The STL Coordination Language.- The STL++ Coordination Language.- The Agent & Co Coordination Language.- Case Studies in STL++.- Collective Robotics Simulation.- Trading System Simulation.- Conclusion.

Abstract: Preface. Contributing Authors. 1. Agents' conflicts: new issues C. Tessier, H.-J. Muller, H. Fiorino, L. Chaudron. Part I: Conflicts and Agents: Essentials. 2. Conflicts within and for collaboration C. Castelfranchi, R. Falcone. 3. Their problems are my problems M. Hannebauer. 4. Conflicts in social theory and MAS T. Malsch, G. Weiss. Part II: Conflicts of Operational Agents. 5. Conflicts in agent teams Hyuckchul Jung, M. Tambe. 6. Conflict-based behaviour emergence in robot teams J. Penders. 7. Conflicts in collective robotics F. Chantemargue. Part III: Application Centered Agents' Conflicts. 8. Strategic use of conflicts in tutoring systems E. Aimeur. 9. Conflict handling in collaborative search J. Denzinger. 10. Conflict as a heuristic in the development of an interaction mechanics W.F. Lawless, T. Castelao, C.P. Abubucker. References. Index.

Abstract: Knowledge fusion refers to the process of locating and extracting knowledge from multiple, heterogeneous on-line sources, and transforming it so that the union of the knowledge can be applied in problem-solving. The KRAFT project has defined a generic agent-based architecture to support fusion of knowledge in the form of constraints expressed against an object data model. KRAFT employs three kinds of agent: facilitators locate appropriate on-line sources of knowledge; wrappers transform heterogeneous knowledge to a homogeneous constraint interchange format; mediators fuse the constraints together with associated data to form a dynamically-composed constraint satisfaction problem, which is then passed to an existing constraint solver engine to compute solutions. The paper presents the KRAFT architecture and the three kinds of agent, and includes a description of a demonstration KRAFT application in the domain of telecommunications service provision.

Abstract: This paper studies how to bring flexibilityto fault-tolerant systems. Firstly, multi-agent systems are identified as a very valuable basis for reaching this goal,and reliability is also shown to be a rare and attractive feature for such systems. We then propose a frameworkfor building applications that provide adaptive fault tolerance, and put forward the promising results obtainedwhen testing the implementation of this framework. We conclude with drawing some perspectives of evolution ofour work.