Adriana Giret

Holons and agents

The future of the manufacturing sector in the world will be determined by how it meets the challenges of the 21st century. Two paradigms promise to meet these challenges, multi-agent systems (MAS) and holonic system (HS). Currently, there is some misunderstanding about the relationships between these two approaches. The aim of this work is to make a study of all the characteristics of the HS and the MAS approach, in order to illustrate a comprehensive comparison of holons and agents, and, at the same time, to present a survey about the main research works in both areas.

An abstract architecture for virtual organizations: The THOMAS approach

Today, the need for architectures and computational models for large-scale open multi-agent systems is considered to be a key issue for the success of agent technology in real-world scenarios. This paper analyzes the significant unsolved problems that must be taken into account in order to develop real, open multi-agent systems. It identifies requirements and related open issues, discusses how some of these requirements have been tackled by current technologies, and explains how the THOMAS architecture is able to give support to these open issues. This paper also describes the THOMAS abstract architecture and computational model for large-scale open multi-agent systems based on a service-oriented approach that specifically addresses the design of virtual organizations. An application example for the management of a travel agency system, which demonstrates the new features of the proposal, is also presented.

Engineering Holonic Manufacturing Systems

Holonic Manufacturing Systems (HMS) was first proposed as a new manufacturing paradigm at the beginning of 1990 and has subsequently received a lot of attention in academic and industrial research. The application of holonic concepts to manufacturing was initially motivated by the inadequacy of existing manufacturing systems in the following two aspects: (i) dealing with the evolution of products within an existing production facility, and (ii) maintaining a satisfactory performance outside of normal operating conditions. In spite of the large number of developments reported in Holonic Control Architectures and Holonic Control Algorithms, there is very little work reported on Methodologies for HMS. Manufacturing requirements impose significant properties on HMS making the modelling process of HMS a complex and difficult task. In this work we describe how to engineer Holonic Manufacturing Systems using a Multi-agent System method. Our approach is based on holonic systems specific requirements, and it incorporates software engineering principles in order to assist the system designers at each development stage. It implements a collaborative-engineering process for HMS, and provides clear and unambiguous analysis and design guidelines.

An open architecture for service-oriented virtual organizations

Recent technological advances in open systems have imposed new needs on multi-agent systems. Nowadays, open systems require open autonomous scenarios in which heterogeneous entities (agents or services) interact to fulfill the system goals. The main contribution of this paper is the definition of an open architecture and computational model for largescale open multi-agent systems based on a service-oriented approach. This work requires a high-level design of all features and needs for systems of this kind. The new proposed architecture, called THOMAS, is specifically addressed for the design of virtual organizations. A simplified example for the management of a travel agency system, which shows the features of the proposal, is also included.

Energy-efficient dynamic scheduling for a flexible flow shop using an improved particle swarm optimization

Manufacturing companies are facing the emergent challenges to meet the demand of sustainable manufacturing.Energy-efficient dynamic scheduling is a NP-hard problem presented in manufacturing systems.A novel particle swarm optimization algorithm based on Hill function is presented to minimize makespan and energy consumption.The relationship between makespan and energy consumption is conflicting.The results show that the proposed algorithm outperforms the behavior of state of the art algorithms. Due to increasing energy requirements and associated environmental impacts, nowadays manufacturing companies are facing the emergent challenges to meet the demand of sustainable manufacturing. Most existing research on reducing energy consumption in production scheduling problems has focused on static scheduling models. However, there exist many unexpected disruptions like new job arrivals and machine breakdown in a real-world production scheduling. In this paper, it is proposed an approach to address the dynamic scheduling problem reducing energy consumption and makespan for a flexible flow shop scheduling. Since the problem is strongly NP-hard, a novel algorithm based on an improved particle swarm optimization is adopted to search for the Pareto optimal solution in dynamic flexible flow shop scheduling problems. Finally, numerical experiments are carried out to evaluate the performance and efficiency of the proposed approach.

An engineering framework for Service-Oriented Intelligent Manufacturing Systems

HighlightsA complete software development method is presented for Service Oriented Intelligent Manufacturing Systems.The software development method provides a CASE tool that automatically generates executable code for the system execution.A specific and complete set of service guidelines for the development of Service Oriented Intelligent Manufacturing Systems is provided.The completeness of the approach is evaluated under two views: an empirical by means of case studies; and a qualitative analysis. Nowadays fully integrated enterprises are being replaced by business networks in which each participant provides others with specialized services. As a result, the Service Oriented Manufacturing Systems emerges. These systems are complex and hard to engineer. The main source of complexity is the number of different technologies, standards, functions, protocols, and execution environments that must be integrated in order to realize them. This paper proposes a framework and associated engineering approach for assisting the system developers of Service Oriented Manufacturing Systems. The approach combines multi-agent system with Service Oriented Architectures for the development of intelligent automation control and execution of manufacturing systems.

From system requirements to holonic manufacturing system analysis

Research on Holonic Manufacturing Systems is undergoing a huge amount of interest and developments. In spite of the great amount of effort devoted to architectures and algorithms for holonic control of manufacturing systems, there is little work about design methodologies. The present work presents a multi-agent approach for analysis of Holonic Manufacturing Systems, it is based on an abstract agent definition and the PROSA reference architecture for holonic systems. The analysis process to identify and specify holons is organized around five views of the system and integrates the advantages of proven multi-agent methodologies to the manufacturing field. An industrial case example from the ceramic domain is presented to show the characteristics of the approach.

MAS methodology for HMS

Developments in Holonic Manufacturing Systems (HMS) have been reported in three main areas: architectures, algorithms, and methodologies for HMS. Despite the advancements obtained in the first two areas the methodologies for HMS have not received great attention. To date, many of the developments in HMS have been conducted in an almost “empirical way”, without design methodology. There is a definite need to have methodologies for HMS that can assist the system designer at every development steps. This methodology should also provide clear and unambiguous analysis and design guidelines. To this end, in this work we present a Multi Agent Methodology for HMS analysis and design.

A multi agent methodology for holonic manufacturing systems

There is a definitive need to have methodologies for holonic systems (HMS) [4], based on software engineering principles, which assist the system designer in every development steps and provide clear, unambiguous analysis and design guidelines. We believe that methodologies from the Multi Agent Technology (MAS) are good candidates for modeling HMS. Some reasons are: the similarities between the holonic and the agent approaches [2], the wide use of agents as the implementation tool for HMS, and the availability of complete MAS Methodologies. Nevertheless, there are some extensions we have to include to a MAS methodology to be able to model the HMS requirements in a proper way: holons recursive structure [3], systems abstraction levels, HMS specific guidelines and a mixed top-down and bottom-up approach for analysis and design steps. These facts have leaded our research towards the definition of a MAS methodology for HMS analysis and design.In our approach, the HMS is specified by dividing it in more specific characteristics that form different views of the system. These models are agent, organization, interaction, environment and task/goal. The way in which the views are defined is inspired by INGENIAS methodology [5]. The development process of our methodology provides the HMS designer with clear and HMS-specific modeling guidelines. It also provides complete development phases for the HMS life cycle. The phases of our approach are System Requirements Analysis, Holons Identification and Specification. Holons Design, Holons Implementation, SetUp and Configuration, and Operation and Maintenance. The first stage, System Requirements Analysis and the second stage Holons Identification and Specification define the analysis phase of our approach. The aim of the analysis phase is to produce a high-level HMS specification. We have defined domain specific guidelines for the identification and specification of holons, such as: HMS UC Guidelines, to help the designer to identify domains cooperations [1] and the system goals as use cases, and PROSA Guidelines based on PROSA types of holons [6] to identify and specify holons. The analysis adopts a top-down recursive approach. The next step in the development process is the Holons Design stage which is a bottom-up process to produce the system architecture. The aim of the Holons Implementation stage is to produce an Executable Code for the SetUp and Configuration stage. Finally maintenances functions are executed in the Operation and Maintenance stage. Currently, we are working on the evaluation of our methodology with industrial Case Studies, and we are developing a CASE tool for our methodology.

Agent-supported simulation environment for intelligent manufacturing and warehouse management systems

The manufacturing field is an area where the application of simulation is an essential tool for validating methods and architectures before applying them on the factory floor and the warehouse control. Multiagent system technology has demonstrated its utility in manufacturing system modelling and implementation. Agenthood features such as proactivity, reactivity, sociability and trust may also be useful for associating them with the specific simulation needs of the new manufacturing and warehouse management requirements. In this article, we present an agent-supported simulation tool for an intelligent manufacturing system and its influence with the management of the warehouse applied to a metal-mechanic manufacturing enterprise. The main goal is to provide a flexible simulation tool that can be adapted to solve the new manufacturing requirements and optimise and validate methods of management of material in warehouses that appear in real environments. Thus, the benefits of the multiagent theory application for the simulation of intelligent manufacturing and warehouse management system allows the experts of manufacturing domains to optimise the resource usage, control of inventory, providers and to have enough data to make decisions that influence the production level are presented.