Multi-agent System Architecture Research Articles

An advanced agent-based order planning system for dynamic networked enterprises

There is an increasing interest in exploring the opportunities for competitive advantage that arise from reinforcing core competencies and innovative capabilities of the individual companies, and by forming integrated supply networks. In complex and dynamic environments such as the automotive and semiconductor industries, managing and co-ordinating the procurement of materials, their transformation into intermediate and finished products, and the distribution to the final customers, are very demanding tasks in terms of information systems. In general, current available software packages do not provide the full support needed for networked and distributed organizations, and are clearly insufficient in what concerns the planning and coordination activities needed in these heterogeneous environments. In order to adequately tackle these problems, this work proposes a multi-agent system architecture for real-time customer-order planning in distributed manufacturing enterprises, addressing the requirements of a make-to-order environment. The present research work resulted from the Co-OPERATE European Project. Enhanced visibility of information, early warning of disturbances, synchronized production and collaborative planning in the supply chain are the general goals of the project. In particular, for aggregate planning and order promising, a distributed and decentralized information system, based on an architecture of agents and extensively using the internet, was designed and implemented. This system aims at responding to the basic requirements of cooperativeness, integration and configurability, and at providing new and more powerful decision support.

Knowledge source network configuration approach to knowledge logistics

Today multidisciplinary real-world problem solving requires an intensive knowledge integration and exchange between participants of the global information environment. Along with a large number of available distributed knowledge sources (KSs) this has led to the appearance of a new direction in knowledge management called knowledge logistics (KL). The paper describes a KSNet-approach, being developed to build an intelligent support system addressing problems of KL through knowledge fusion (KF). The approach is mainly based on such technologies as ontology management and intelligent agents. In the KSNet-approach, a multi-agent system architecture based on FIPA Reference Model was chosen as a technological basis for definition of agent properties and functions and FIPA ontology definition was used for knowledge representation. As a general model of ontology representation the object-oriented constraint network paradigm was proposed. The most important ideas of the proposed KSNet-approach were implemented and verified via a developed software prototype of the system “KSNet”.

A re-configurable multi-agent system architecture for error recovery in production systems

Multi-agent systems for manufacturing systems appear to provide adequate response to abrupt disturbances on the shop floor. To date, most of the work has been focused on planning and scheduling but very little work has been done on issues pertaining to monitoring, diagnostics and error recovery. Our approach addresses the issue of combining the discipline of hierarchical systems with the agility of multi-agent systems. Within the context of a hierarchy, the focus is on the workstation level and, in particular, the construction of a re-configurable system having production agents, error recovery agents, and a mediator agent structure connecting production and recovery agent hierarchies. In addition, the relationship to a multi-level, multi-layer hierarchy control is established. This latter hierarchy, based on Petri Net constructs, serves, in one sense, as a retrieval based resource for process planning and generation of recovery plans for production and recovery agents within the proposed multi-agent system. An objective of this effort is to provide a test-bed for comparison of hierarchical systems, heterarchical, and a hybrid combination which is the focus of the investigation presented here.

Automated monitoring of medical protocols: a secure and distributed architecture

The control of the right application of medical protocols is a key issue in hospital environments. For the automated monitoring of medical protocols, we need a domain-independent language for their representation and a fully, or semi, autonomous system that understands the protocols and supervises their application. In this paper we describe a specification language and a multi-agent system architecture for monitoring medical protocols. We model medical services in hospital environments as specialized domain agents and interpret a medical protocol as a negotiation process between agents. A medical service can be involved in multiple medical protocols, and so specialized domain agents are independent of negotiation processes and autonomous system agents perform monitoring tasks. We present the detailed architecture of the system agents and of an important domain agent, the database broker agent, that is responsible of obtaining relevant information about the clinical history of patients. We also describe how we tackle the problems of privacy, integrity and authentication during the process of exchanging information between agents.

A multi-agent and distributed ruler based approach to production scheduling of agile manufacturing systems

A multi-agent and distributed ruler based approach to production scheduling in agile manufacturing systems is proposed in this paper. Considering special and complex requirements of the agile manufacturing, a function-based decomposition method is suggested to fulfil the agent-based system modelling process, where a real industrial system can be turned into a multi-agent system (MAS). Three classes of agents are defined, which populate the production-scheduling environment: Management agents, resource agents and part agents. All these agents are arranged on the heterarchical framework. Consequently, with the distribution of agents, the complicated scheduling problem is divided into several sub-problems, such as the self-supporting and decision-making of each agent for the local near-optimal schedule and the cooperating or coordinating between agents for system global performance. A ruler-based decision making mechanism is proposed to ensure orderly agent running, and the ability to respond to customer's needs or environmental changes rapidly. Meanwhile the heterarchical MAS architecture offers regroup capabilities that support those rulers to be reconfigured and reused flexibly. Case studies illustrate the feasibility and efficiency of the approach.

A multi-agent system architecture for coordination of just-in-time production and distribution

A multi-agent system architecture for coordination of just-in-time production and distribution is presented. The problem to solve is twofold: first the right amount of resources at the right time should be produced, then these resources should be distributed to the right consumers. In order to solve the first problem, which is hard when the production and/or distribution time is relatively long, each consumer is equipped with an agent that makes predictions of future needs that it sends to a production agent. The second part of the problem is approached by forming clusters of consumers within which it is possible to redistribute resources fast and at a low cost in order to cope with discrepancies between predicted and actual consumption. Redistribution agents are introduced (one for each cluster) to manage the redistribution of resources. The suggested architecture is evaluated in a case study concerning management of district heating systems. Results from a simulation study show that the suggested approach makes it possible to control the trade-off between quality of service and degree of surplus production. We also compare the suggested approach to a reference control scheme (approximately corresponding to the current approach to district heating management), and conclude that it is possible to reduce the amount of resources produced while maintaining the quality of service. Finally, we describe a simulation experiment where the relation between the size of the clusters and the quality of service was studied.

Research on theory and method in parallel designing environment based on coupling

Understanding the relation between multiple agents is very important in developing a multi-agent system. Two concepts, multi-agent system architecture and multi-agent system architecture prototype are brought forward, and the method of descrbing the relation between agents using the concepts is explored. The structure and components are designed. Based on the fact that cooperation between distributed agents is erratic, a controlled natural language communication function is designed and realized. Applying the offered interfaces, the user’s workload consumed in agent communication and management is reduced, and at the same time, robustness of the multi-agent system is improved.

Supporting evolution in a multi-agent cooperative design environment

New architectures for next-generation of computer-aided design environments need to be distributed, evolutive, and cooperative. This paper presents a multi-agent design environment that supports cooperative and evolutionary design by cooperation of a group of agents. Multi-agent system architecture offers a promising framework with its novel approach for dynamically creating and managing agent communities in widely distributed and ever-changing design environments; while the genetic algorithm based approach provides a foundation for supporting evolutionary and innovative design abilities. The computational and information processing processes involved in evolutionary and distributed environment are discussed in this paper.

A multi-agent system infrastructure for software component market-place

In this paper, we introduce a multi-agent system architecture and an implemented prototype for software component market-place. We emphasize the ontological perspective by discussing the ontology modeling for component market-place, UML extensions for ontology modeling, and the idea of ontology transfer which makes the multi-agent system to adapt itself to the dynamically changing ontologies.

EVALUATING THE DEPLOYMENT OF FIPA STANDARDS WHEN DEVELOPING APPLICATION SERVICES

We draw upon various practical experiences of designing and implementing complex systems through a multiagent approach which supports engineering of dynamic open distributed services. The general scope of multiagent system software engineering is reviewed with focus on the analysis and evaluation of certain aspects of the current specification standards provided by FIPA (Foundation of Intelligent Physical Agents). The benefits and drawbacks of a multiagent approach, using the FIPA standards as a benchmark, are evaluated and further illustrated through the deployment of an audio–video entertainment broadcasting (AVEB) system. The development and testing of the AVEB application was part of an EU project called FACTS (acts AC317). A main result of using agent engineering paradigm for complex distributed development, especially apparent in FIPA standards, has been the identification of the usefulness and power of its protocols. The reason for the importance of the protocols in developing multiagent systems (MAS) is it provides a degree of expressing cooperation within MAS architecture. As the protocols stand currently they are not sufficient to capture a complete explicit model of the cooperative requirements in multiagent systems. However, they provide a basis from which to start. We examine this feature of FIPA further in order to evaluate its role as a bridge between the mental agency and social agency requirements in the development of cooperation in multiagent systems.

Formal specification of a multi-agent system architecture for manufacture: the contribution of the π-calculus

We first recall the interest of multi-agent systems (MAS) in computer-aided production engineering. We argue that the notion of agent needs to be clarified and that MAS architectures must consequently be subject to formal descriptions. Concurrent specification languages, such as process calculi, can be regarded as suitable models for this purpose. The π-calculus is therefore introduced through its syntax and an informal description of its semantic. We then support the idea that this calculus fulfils the requirements for MAS architecture description, especially the aspects related to the interactions between agents. Then, we focus on the Supply Chain Management problem: two agents and their sub-components are specified using the π-calculus. After a short evaluation of the proposed method, we end by pointing out the potential applications and some areas of research related to our approach.

A formulation of the collaboration mechanism for integrated abnormal situation management

To better support computer integrated abnormal situation management (ASM), a multiagent system architecture is proposed within which user and functional agents cooperate based on a collaboration mechanism. Each agent participating in the collaboration is composed of a local module and a collaboration module. Different types of collaboration are systematically analyzed. In particular, interactions among functional agents are represented by an ASM activity diagram and an extended signed digraph (SDG). A prototype operation support system for the simulated Tennessee Eastman process is used to illustrate the application of the framework and the collaboration mechanism.

A MULTI-AGENT SYSTEM ARCHITECTURE FOR DISTRIBUTED COMPUTER VISION

We present a multi-agent system architecture dedicated to model computer vision systems, which provides the vision system with a great degree of flexibility. The basic idea of this architecture is to model a vision system as a society of autonomous agents, where each agent is responsible for specific vision tasks, the control strategy of a vision system is decentralized, and agents communicate using a flexible but easy understandable communication language. This directly leads to self-organizing vision systems, which accomplish vision tasks by goal-driven communication processes. We describe in detail the basic concepts of the proposed multi-agent system approach including the agent architectures, the communication language and network, as well as the interaction strategies. As a testbed for the proposed architecture we have modeled an object recognition system as an assembly of agents which organize themselves according to a given recognition task by employing communication.

Uma proposta de arquitetura de software baseada em agentes

Agent-based software development has been the object of several studies which analyze its application to distributed systems and look for techniques capable of minimizing the difficulties inherent to this new approach. A suggestion of software architecture for multi-agent systems has been provided. It has been elaborated from studies on software agents, multi-agent systems and agent-based software architectures. Some of these concepts have been given, including the main pitfalls found in the construction of agent-based architectures. Further, two models that served as subsidy for their elaboration have also been described

MetaMorph: An adaptive agent-based architecture for intelligent manufacturing

Global competition and rapidly changing customer requirements are forcing major changes in the production styles and configuration of manufacturing organizations. Traditional centralized manufacturing systems are not able to meet such requirements. This paper proposes an agent-based approach for dynamically creating and managing agent communities in such widely distributed and everchanging manufacturing environments. After reviewing the research literature, an adaptive multi-agent manufacturing system architecture called MetaMorph is presented and its main features are described. Such architecture facilitates multi-agent coordination by minimizing communication and processing overheads. Adaptation is facilitated through organizational structural change and two learning mechanisms: learning from past experiences and learning future agent interactions by simulating future dynamic, emergent behaviours. The MetaMorph architecture also addresses other specific requirements for next generation manufacturing systems, including scalability, reliability, stability, maintainability, flexibility, real-time planning and scheduling, standardized communication, fault tolerance, and security. The proposed architecture is implemented as a multi-agent virtual manufacturing system, in simulation form, which incorporates heterogeneous manufacturing agents within different agent-based shop floors or factories. The experimental results have shown the potential of the agent-based approach for advanced manufacturing systems.

Distributed scheduling to support a call center: A cooperative multiagent approach

This article describes a multiagent system architecture to increase the value of 24-hour-a day call center service. This system supports call centers in making appointments with clients on the basis ofknowledge ofemployees and their schedules. Relevant activities are scheduled for employees in preparation ofsuch appointments. The multiagent system architecture is based on principled design, using the compositional development method for multiagent systems DESIRE. To schedule procedures in which more than one employee is involved, each employee is represented by its own personal assistant agent, and a work manager agent coordinates the schedules of the personal assistant agents and clients through the call center. The multiagent system architecture has been applied to the banking domain, in cooperation with and partially funded by the Rabobank.

An ontology for commitments in multiagent systems

Social commitments have long been recognized as an important concept for multiagent systems. We propose a rich formulation of social commitments that motivates an architecture for multiagent systems, which we dub spheres of commitment. We identify the key operations on commitments and multiagent systems. We distinguish between explicit and implicit commitments. Multiagent systems, viewed as spheres of commitment (SoComs), provide the context for the different operations on commitments. Armed with the above ideas, we can capture normative concepts such as obligations, taboos, conventions, and pledges as different kinds of commitments. In this manner, we synthesize ideas from multiagent systems, particularly the idea of social context, with ideas from ethics and legal reasoning, specifically that of directed obligations in the Hohfeldian tradition.

Agent-based production management

Due to the requirements of one-of-a-kind production, a separate, all-embracing concept for shipbuilding was evolved. In this concept, the harmonisation of production and its activities with respect to time and content is given priority. Customer orders are handled as projects and the simultaneous production processes, their activities and the resource requirements are harmonised with multi-projects in mind. Considering the involved software-systems concurrent manufacturing, demands a movement from centralised towards increasingly decentralised models of control and action. Our goal is to support decentralised planning entities, each of them equipped with a GIGROS-system. A central challenge is the co-ordination and synchronization of activities within those complex systems. To improve the system according to today's most modern production paradigms, it is being redesigned towards a multi-agent system architecture. Introducing concepts of distributed artificial intelligence, a co-operation of the modules or of several systems can be achieved to yield solutions for the simultaneous, co-operative and decentralised scheduling and resource planning. After a short description of the special requirements of the one-of-a-kind production, the paper will introduce the features of the graphical interactive rough planning system (GIGROS) focusing on the multi-agent approach.

Maciv a dai based resource management system

Managing resources in the framework of the civil construction sector is usually an extremely complex task. There are many factors contributing to this complexity: the variety and great number of existing resources, both human and material; the diversity of tasks that each working unit is able to execute; the performance of each working unit; the involved costs; and the spatial distribution of all resources over the different places, leading to the need for displacement from one site to another. All these important factors imply a high number of variables, resulting in a somewhat difficult optimization process. On the other hand, these factors are highly dynamic as a result of unpredictable situations responsible for the modification of the initial conditions, e.g., weather conditions, uncertainties attached to task duration, acquisition of new resources, technical problems related with those resources, and accidents. Such dynamics make it mandatory for the systems to have the capability to continuously adapt themselves to the evolving real conditions, overcoming usual limitations of classical inflexible solutions. To handle this problem, we set up MACIV, a project whose goal is to design and implement a computer system, mainly based on distributed artificial intelligence techniques, enabling a decentralized management of the different available resources in civil construction companies. This problem was inspired by an existing large company whose experts are collaborating with us, both in the problem definition and in the system design phases. In this article, the overall multiagent system architecture is presented, and the employed techniques are explained, with special emphasis on our own contributions to the specific negotiation and agents' coalition formation protocols. In order to support the particulars of the application, original intercoalition and intracoalition negotiation algorithms and strategies were developed and are explained here. Finally, an application example is described in order to illustrate how our proposal enables the system to reach a good solution for a concrete scenario.

Designing distributed control architecture for cooperative multi-agent system and its real-time application to soccer robot

The soccer robot system consists of multi-agents, with highly coordinated operations and movements so as to fulfill specific objectives, even under adverse situations. The coordination of the multi-agent is associated with a lot of supplementary work in advance. The associated issues are the position correction, prevention of communication congestion, local information sensing in addition to the need for imitating the human-like decision making. A control structure for soccer robot is designed and several behaviors and actions for a soccer robot are proposed. The action selection mechanism (ASM) is proposed as a high-level controller and is applied to a one-on-one soccer game. The intervention module of ASM uses multilayer perceptrons and its effectiveness is shown by experimental results. Modified zone defense as a basic strategy and several special strategies for fouls as applied to SOTY and MIRO teams for MIROSOT are discussed. The authors have been placed in the second and fourth positions in MIROSOT'96 (Micro-Robot World Cup Soccer Tournament) held at KAIST in November 1996, with the proposed scheme and micro-robots.