Many researchers have shown that there is no single best organization or coordination mechanism for all environments. This paper discusses the design and implementation of an extendable family of coordination mechanisms, called Generalized Partial Global Planning (GPGP). The set of coordination mechanisms described here assists in scheduling activities for teams of cooperative computational agents. The GPGP approach has several unique features. First, it is not tied to a single domain. Each mechanism is defined as a response to certain features in the current task environment. We show that different combinations of mechanisms are appropriate for different task environments. Secondly, the approach works in conjunction with an agent's existing local planner/scheduler. Finally, the initial set of five mechanisms presented here generalizes and extends the Partial Global Planning (PGP) algorithm. In comparison to PGP, GPGP allows more agent heterogeneity, it exchanges less global ...

The construction of large-scale embedded software systems demands the use of design methodologies and modelling techniques that support abstraction, inheritance, modularity, and other mechanisms for reducing complexity and preventing error. If multi-agent systems are to become widely accepted as a basis for large-scale applications, adequate agentoriented methodologies and modelling techniques will be essential. This is not just to ensure that systems are reliable, maintainable, and conformant, but to allow their design, implementation, and maintenance to be carried out by software analysts and engineers rather than researchers. In this paper we describe an agent-oriented methodology and modelling technique for systems of agents based upon the Belief-Desire-Intention (BDI) paradigm. Our models extend existing Object-Oriented (OO) models. By building upon and adapting existing, well-understood techniques, we take advantage of their maturity to produce an approach that can be easily lear...

There are many formal approaches to specifying how the mental state of an agent entails that it perform particular actions. These approaches put the agent at the center of analysis. For some questions and purposes, it is more realistic and convenient for the center of analysis to be the task environment, domain, or society of which agents will be a part. This paper presents such a task environment-oriented modeling framework that can work hand-in-hand with more agent-centered approaches. Our approach features careful attention to the quantitative computational interrelationships between tasks, to what information is available (and when) to update an agent's mental state, and to the general structure of the task environment rather than single-instance examples. A task environment model can be used for both analysis and simulation, it avoids the methodologicalproblems of relying solely on single-instance examples, and provides concrete, meaningful characterizations with which ...

Coordination, as the act of managing interdependencies between activities, is one of the central research issues in Distributed Artificial Intelligence. Many researchers have shown that there is no single best organization or coordination mechanism for all environments. Problems in coordinating the activities of distributed intelligent agents appear in many domains: the control of distributed sensor networks; multi-agent scheduling of people and/or machines; distributed diagnosis of errors in local-area or telephone networks; concurrent engineering; `software agents' for information gathering. The design of coordination mechanisms for group...

There are many formal approaches to specifying how the mental state of an agent entails the particular actions it will perform. These approaches put the agent at the center of analysis. For some questions and purposes, it is more realistic and convenient for the center of analysis to be the task environment, domain, or society of which agents will be a part. This paper presents such a task environment-oriented modeling framework that can work hand-in-hand with more agent-centered approaches. Our approach features careful attention to the quantitative computational interrelationships between tasks, to what information is available (and when) to update an agent's mental state, and to the general structure of the task environment rather than single-instance examples. A task environment model can be used for both analysis and simulation, it avoids the methodological problems of relying solely on single-instance examples, and provides concrete, meaningful characterizations with which to sta...

The study of computational agents capable of rational behaviour has received a great deal of attention in recent years. A number of theoretical formalizations for such multiagent systems have been proposed. However, most of these formalizations do not have a strong semantic basis nor a sound and complete axiomatization. Hence, it has not been clear as to how these formalizations could be used in building agents in practice. This paper explores a particular type of multi-agent system, in which each agent is viewed as having the three mental attitudes of belief (B), desire (D), and intention (I). It provides a family of multi-modal branching-time BDI logics with a semantics that is grounded in traditional decision theory and a possible-worlds framework, categorizes them, provides sound and complete axiomatizations, and gives constructive tableaubased decision procedures for testing the satisfiability and validity of formulas. The computational complexity of these decision procedures is n...

This paper shows how the distributions of objective parameters such as "the number of VLM methods seen by the maximally loaded agent" (

viewpoints. It is a tool for building and testing computational theories of coordination. TÆMS is compatible with both formal computational agent-centered approaches and experimental approaches. The framework allows us to both mathematically analyze (when possible) and quantitatively simulate the behavior of multi-agent systems with respect to interesting characteristics of the computational task environments of which they are part. We believe that it provides the correct level of abstraction for meaningfully evaluating centralized, parallel, and distributed control algorithms, negotiation strategies, and organizational designs. This chapter will briefly describe the TÆMS modeling framework for representing abstract task environments, concentrating particularly on its support for simulation. I will describe how to model each of several different multi-agent problem-solving environments, such as This work was supported by DARPA contract N0

A generic, component based architecture is proposed as a basis for designing the agents used within Multi-Agent Systems. The architecture, based on Sun's Java Beans, is explored, and both domain independent and applied component examples are described in detail. Designs for theoretical new components for the applied agent are also proposed and examined. 1 Overview Component based architectures are a relatively new introduction to software development. They attempt to effectively encapsulate the functionality of an object while respecting interface conventions, the goal being to easily combine groups of components to create stand alone applications. This type of design promotes software reusability - the ability to painlessly transport source code from one project to another - which is a long sought after but infrequently achieved goal of software engineering. In this paper I will describe such a component architecture designed for use in the domain of agent construction. Much of ...

Information gathering agents are required in many software agent applications to answer queries, posed by other agents, using a variety of available information sources. We formally consider the problem of designing information gathering agents, and make two important contributions. First, we examine the key issue of integrating knowledge from external sites into our knowledge base, and present an expressive language for this purpose. A noteworthy feature of our language is its ability to capture the knowledge that some external sites have complete information of a certain kind, using rich semantic constraints. Given a query on the knowledge base, it is important for the agent to first determine the set of external sites that contain information relevant to answering the query, and then access those sites. Our second contribution is to show that, given a query and the descriptions of the external sites in our language, it is possible to determine minimal subsets of sites that are neede...