Catálogo de publicaciones - libros

AOSE methodologies and models borrow abstractions and concepts from organizational and social disciplines. Although they all view multi-agent systems as organized society, the organizational abstractions, assumptions, concepts, and models in them are actually used in different ways. It is therefore desirable to have a systematic way of analyzing and comparing the organizational and social concepts in AOSE. The contribution of this paper is threefold. Firstly, we identify some premises behind the social conceptions adopted in multi-agent systems. Secondly, we define levels of modeling constructs and classify organizational and social concepts in the AOSE literature into categories according to their organizational abstractions. Finally, we analyze two representative AOSE methodologies and their models, explaining how they use organizational and social concepts to analyze and specify multi-agent system, reflecting various social premises at different levels.

Several modeling techniques exist to represent agent interaction protocols mostly based on work done in distributed systems. These modeling techniques do not take the agent features such as the autonomy into account. Agent Interaction Protocol designers are now considering specific modeling techniques that contain these features. In this paper, we present the second version of the Agent UML interaction diagrams dedicated to interaction protocols, and based on UML 2.0.

The is a semi-formal visual modeling language, specified as an extension to UML 2.0. It is a consistent set of modeling constructs designed to capture the aspects of multi-agent systems. The ultimate objective for AML is to provide a means for software engineers to incorporate aspects of multi-agent system engineering into their analysis and design processes. This paper presents an introductory overview of AML, discussing the motivations driving the development of the language, the scope and approach taken, the specific language structure and optional extensibility. The core AML modeling constructs are explained and demonstrated by example where possible. Extensions to OCL and CASE tool support are also discussed.

In this paper we introduce an approach for defining semantics for AUML agent interaction protocol diagrams using Petri net code structures. This approach is based on the usage of net components which provide basic tasks and the structure for Petri nets. Agent interaction protocol diagrams are used to model agent conversations on an abstract level. By mapping elements of the diagrams to net components we are able to translate the diagrams into Petri nets, i.e to generate code structures from the drawings. We provide tool support for this approach by combining a tool for net components with a tool for drawing agent interaction protocol diagrams. This combined tool is available as a plug-in for (eference et orkshop).

Several agent-oriented methodologies have been proposed over the last few years. Unlike the object-oriented domain and unfortunately for designers, most of the time, each methodology has its own purposes and few standardization works have been done yet, limiting the impact of agent design on the industrial world. By studying three existing methodologies – ADELFE, Gaia and PASSI – and the concepts related to them, this paper tries to find a means to unify their meta-models. Comparing a certain number of features at the agent or system level (such as the agent structure, its society or organization, its interactions capacities or how agents may be implemented) has enabled us to draw up a first version of a unified meta-model proposed as a first step toward interoperability between agent-oriented methodologies.

Societies need patterned behavior to exist. Large-scale agent societies may contain a diversity of agents, each with differing abilities and functionalities. When such an agent system is given a task, it must dynamically muster together a group of agents that collectively have the capability to accomplish the task. To do this, the agent society needs to be able to understand its agents and their potential interactions.

This paper contains a proposed superstructure specification that defines the user-level constructs required to model agents, their roles and their groups. These modeling constructs provide the basic foundational elements required in multi-agent systems to foster dynamic group formation and operation. As agent systems scale beyond the point where an individual organization can track and control their behavior, the use of these concepts within the society will facilitate dynamic, controlled, task-oriented group formation. This in turn will enhance the predictability, reliability and stability of the agent system as a whole, as well as facilitating the analysis of both group and system behavior.

Current agent-oriented methodologies focus mainly on multi-agent systems analysis and design, but without providing straightforward connections to the implementation of such systems on any of the existing agent platforms (e.g. FIPA-OS, Jade, or Zeus), or just forcing the use of specific agent platforms. In this paper we show how the Model Driven Architecture (MDA) can be effectively used to derive agent implementations from agent-oriented designs, independently from both the methodology used and the concrete agent platform selected. Furthermore, this transformation process can be defined in an scalable way, and partly automated thanks to the use of a platform-neutral agent model, called .

Engineering non-trivial open multi-agent systems is a challenging task. Our research focusses on situated multi-agent systems, i.e. systems in which agents are explicitly placed in an environment which agents can perceive and in which they can act. Situated agents do not use long-term planning to decide what action sequence should be executed, but select actions based on the locally perceived state of the world and limited internal state. To cope with change and dynamism of the system, situated agents must be able to adapt their behavior. A well-known family of agent architectures for adaptive behavior are free-flow architectures. However, building a free-flow architecture based on an analysis of the problem domain is a quasi-impossible job for non-trivial agents. To tackle the complexity of designing adaptive agent behavior based on a free-flow architecture, suitable abstractions are needed to describe and structure the agent behavior. The abstraction of a role is obviously essential in this respect. A modeling language is needed as well to model the behavior of the agents. We propose a statechart modeling language to support the design of roles for situated agents. In this paper we describe a design process for adaptive behavior of situated agents as part of a multi-agent oriented methodology. The design process integrates the abstraction of a role with a free-flow architecture. Starting from the results of analysis of the problem domain, the designer incrementally refines the model of the agent behavior. The resulting class diagram serves as a basis for implementation. We illustrate the subsequent design steps with a case study on controlling a collection of automated guided vehicles.

More and more effort is made to provide methodologies for the development of agent–based systems. Awareness has grown that these are necessary to develop high quality agent systems. In recent years a number of proposals have been given. Based on our experiences we argue that a complete evaluation of methodologies cannot be done without considering target platforms, because the differences between available implementations are too fundamental to be ignored. In order to conduct a suitable comparison we present a flexible evaluation framework that takes platform specific criteria into account. Part of this framework is a procedure to derive relevant criteria from the evaluated platforms and methodologies. In combination with a set of platform dependent and independent criteria our framework allows evaluation of the appropriateness of methodologies with respect to platforms. As a consequence, also the suitability of methodologies for an individual platform, or vice versa of several platforms for an individual methodology can be examined. To show the usefulness of our proposal, we evaluate the suitability of different methodologies for an example platform.

While there are many useful models of agents and multi-agent systems, they are typically defined in an informal way and applied in an ad-hoc fashion. Consequently, multi-agent system designers have been unable to fully exploit these models commonalities and specialise or reuse them for specific problems. In order to fully exploit these models and facilitate their reuse we propose a formal approach based upon organisational concepts. The formal notation is the result of the composition of Object-Z and statecharts. The semantics of this multi-formalisms is defined by transition systems. This operational semantics enables validation and verification of specifications. We present this approach through the specification of the satisfaction-altruism model which has been used to design situated multi-agent systems. We put the emphasis on the specification of a mobile robot architecture based on the refinement of this model. The availability of such generic models is a fundamental basis for reuse. We also show how to analyse the specification by validation and verification.