Resource Service Management in Manufacturing Grid System E-Book

Fei Tao

• Herausgeber: John Wiley & Sons
• Kategorie: Fachliteratur
• Sprache: Englisch

Seitenzahl: 632

Veröffentlichungsjahr: 2012

Beschreibung

This book includes discussion on advance computer technologies such as cloud computing, grid computing, and service computing. In addition, it furthers the theory and technology of grid technologies that is used in manufacturing, and accelerates the development of service-oriented manufacturing.

Leseprobe

Contents

Cover

Half Title page

Title page

Copyright page

Dedication

Acknowledgements

Preface

Abbreviations

Chapter 1: Introduction to Manufacturing Grid

1.1 Introduction

1.2 Proposal of Manufacturing Grid

1.3 Concept of MGrid

1.4 Basic Features of MGrid

1.5 The Connotation of MGrid

1.6 Comparison between MGrid and Networked Manufacturing

1.7 Comparison between MGrid and Computing Grid

1.8 Key Research Contents and Technologies of MGrid

1.9 Summary

Chapter 2: Resource Service Optimal-Allocation System in MGrid

2.1 Introduction

2.2 The Architecture of MGrid

2.3 MGrid Collaborative Manufacturing Platform

2.4 MGrid Resource Service Optimal-Allocation System (MGRSOAS)

2.5 The Key Issues and Technologies for Realizing RSOAS

2.6 Summary

Chapter 3: Digital Description of MGrid Resource Service

3.1 Introduction

3.2 Classification of MGrid Resource Service and Its Application

3.3 Requirements of DDoRS in MGrid

3.4 MGrid and Ontology

3.5 Establishing the Method of MGrid-Ontology

3.6 Selection of Describing Language for MGrid-Ontology

3.7 MGrid Ontology

3.8 DDoRS Based on MGrid-Ontology

3.9 Application Case: MGrid-Ontology Based MGrid Resource Service Discovery

3.10 Summary

Chapter 4: MGrid Resource Service Match and Search

4.1 Introduction

4.2 Related Works

4.3 Framework of Resource Service Match and Search in MGrid

4.4 SMAs: Similarity Matching Algorithms (SMAs)

4.5 RS-Matcher: Resource Service Matcher

4.6 Case Study

4.7 Performance Results and Discussion

4.8 Summary

Chapter 5: Resource Service QoS Modeling and Evaluation

5.1 Introduction

5.2 Related Works

5.3 Evaluation Indices System of MGrid Resource Service

5.4 Evaluation of SEIs and IEIs

5.5 Classification and Modeling of MGrid QoS

5.6 Evaluation of MGrid QoS Attribute Parameter

5.7 Application Case: QoS-based MGrid Resource Service Management

5.8 Summary

Chapter 6: Resource Service Trust-QoS Evaluation

6.1 Introduction

6.2 Related Works

6.3 Resource Management and Trust Relationship Management in MGrid

6.4 MGrid Resource Service Trust-QoS Relationship Model

6.5 MGrid Resource Service Trust-QoS Evaluation Model

6.6 Data Structure Design

6.7 Trust-QoS Evaluating and Updating Algorithms

6.8 Real-time and Dynamical Updating Algorithm of Trust-QoS Degree

6.9 Trust-QoS Evaluation Case Study

6.10 Application Case: Trust-QoS Based MGrid Resource Service Scheduling

6.11 Summary

Chapter 7: Resource Service Optimal-selection and Composition Framework

7.1 Introduction

7.2 MGrid-RSOSCF: MGrid Resource Service Optimal-selection and Composition

7.3 T-Layer: Task Layer

7.4 S-Layer: Resource Service Match and Search Layer

7.5 Q-Layer: Resource Service QoS Synthetically Processing Layer

7.6 O-Layer: Resource Service Optimal-selection Layer

7.7 C-Layer: Resource Service Composition Layer

7.8 Summary

Chapter 8: Resource Service Optimal-selection Based on Intuitionistic Fuzzy Set and Non-functionality QoS

8.1 Introduction

8.2 Framework of Resource Service Selection

8.3 Resource Service Optimal-selection Based on IFS in MGrid

8.4 Case Study

8.5 Performance Analysis and Discussion

8.6 Summary

Chapter 9: Correlation Relationship Management in Resource Services Composition

9.1 Introduction

9.2 Related Works

9.3 Motivation

9.4 Correlation Relationship among Resource Services

9.5 QoS Computation Model of Correlation-aware Resource Services Composition

9.6 Case Study: Correlation-aware Resource Services Composition

9.7 Summary

Chapter 10: Resource Service Composition Optimal-selection

10.1 Introduction

10.2 Problem Formulation and Review

10.3 Review of Standard PSO

10.4 Improved PSO for MO-MRSCOS Problem

10.5 Performance Analysis and Discussion

10.6 Summary

Chapter 11: Resource Services Composition Flexibility

11.1 Introduction

11.2 Related Works

11.3 The Analysis, Definition and Classification of RSC Flexibility

11.4 The Measurement of RSC Flexibility

11.5 Case Study and Experiment Results

11.6 Summary

Chapter 12: Resource Services Composition Network

12.1 Introduction

12.2 Scale-free Network (SFN)

12.3 The Theoretical Hypothesis: Composition Service Network is a Scale-free Network

12.4 Concepts and Definition in CoRCS-Net

12.5 The Evolving Behavior of CoRCS-Net

12.6 Theoretical Proof of the Scale-free Characteristics of CoRCS-Net

12.7 Summary

Chapter 13: Failure Detection and Recovery in Resource Service Optimal-Allocation

13.1 Introduction

13.2 Related Works

13.3 Define and Classification of MGrid Failure

13.4 Architecture of MGrid Failure Management System

13.5 Detection of MGrid Failure

13.6 MGrid Failure Recovery Based on ECA Rules

13.7 Implementation and Simulation

13.8 Conclusion

Chapter 14: Summary of the Application of Grid Technology in Manufacturing

14.1 Introduction

14.2 Review of MGrid Theories

14.3 Investigation of Application Research on MGrid

14.4 Key Future Research Issues

14.5 Summary

Chapter 15: Cloud Manufacturing: Development and Commerce Realization of MGrid

15.1 Introduction

15.2 Concept and Architecture of Cloud Manufacturing

15.3 Core Enabling Technologies for Cloud Manufacturing

15.4 Typical Characteristics of Cloud Manufacturing

15.5 Difference and Relationship between Cloud Computing and Cloud Manufacturing

15.6 Classification of Cloud Manufacturing Service Platform

15.7 Key Technologies and Main Research Contents of Cloud Manufacturing

15.8 Key Advantages and Challenges of Cloud Manufacturing

15.9 Summary

Bibliography

Index

Also of Interest

Resource Service Management in Manufacturing Grid System

Scrivener Publishing3 Winter Street, Suite 3Salem, MA 01970

Scrivener Publishing Collections Editors

James E. R. CouperKen DragoonRichard ErdlacRafiq IslamNorman LiebermanPeter MartinW. Kent MuhlbauerAndrew Y. C. NeeS. A. SherifJames G. Speight

Publishers at Scrivener

Martin Scrivener ([e-mail])

Phillip Carmical ([e-mail])

Copyright © 2012 by Scrivener Publishing LLC. All rights reserved.

Co-published by John Wiley & Sons, Inc. Hoboken, New Jersey, and Scrivener PublishingLLC, Salem, Massachusetts.Published simultaneously in Canada.

No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, [phone], fax [phone], or on the web at www.copyright.com. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, [phone], fax [phone], or online at http://www.wiley.com/go/permission.

Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Neither the publisher nor author shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.

For general information on our other products and services or for technical support, please contact our Customer Care Department within the United States at [phone], outside the United States at [phone] or fax [phone].

Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic formats. For more information about Wiley products, visit our web site at www.wiley.com.

For more information about Scrivener products please visit www.scrivenerpublishing.com.

Library of Congress Cataloging-in-Publication Data:

ISBN 978-1-118-12231-0

“To our families for their continuous love, encouragement, and support.”

Acknowledgements

This book is a summary of Dr. Tao’s research on resource service management in manufacturing grid (MGrid) and cloud manufacturing (CMfg) system during his study and work from September 2005 to August 2011 in Wuhan University of Technology (WHUT), the University of Michigan-Dearborn (UMD), and Beihang University (BUAA). Therefore, Dr. Tao would like to send special thanks to Professors YF Hu, ZD Zhou, MZ Yang, BY Shen, and YF Ding in WHUT, Professor D Zhao in UMD, and Prof. L Zhang in BUAA.

The authors would like to express their special thanks to China Machine Press and journal publishers. This book is an English-language version of the authors’ Chinese book (Theory and Practice: Optimal Resource Service Allocation in Manufacturing Grid by F Tao, YF Hu and L Zhang, published by China Machine Press in 2010) but with more than over 40% new content. Some of the material has been published in IEEE Transactions on Industrial Informatics(TII), International Journal of Production Research (IJPR), International Journal of Advanced Manufacturing (IJAMT), European Journal of Operational Research (EJOR), International Journal of Computer Integrated Manufacturing (IJCIM), International Journal of Manufacturing Technology and Management (IJMTM), Enterprise Information Systems(EIS), Knowledge and Information Systems (KIS), Proceedings of the Institution of Mechanical Engineers, Part B, Journal of Engineering Manufacture. (JEM), Chinese Mechanical Engineering, and some conferences such as IEEM’2009 and INDIN’2006.

Dr. Tao would also like to express his special thanks to Professor A. Y. C. Nee for his encouragement, invaluable help and advice over the years and to his suggestion of writing this book.

Some of the book’s research and writing were made possible with the financial support of the following research projects: Excellent Doctoral Dissertation Fund from WHUT (Wuhan University of Technology), Nature Science Foundation of China (No.51005012, No.61074144 and NO.50335020), Hubei Digital Manufacturing Key Laboratory Opening Fund (No.SZ0621), Aeronautic Bairen (One Hundred Outstanding People) Plan and Weishi Youth Foundation (No.YWF-10-02-007) from Beihang University.

Thanks for Dr. H. Guo’s contribution to the Chapters 9 and 11. Dr. Tao would like to give special thanks to his master students for proofing some chapters: Y. Cheng, Y.L. Liu, and L. LV.

Thanks for the help from Martin D. Scrivener, the President of the Scrivener Publishing LLC., as well as the hard and efficient work by the other people at the publishers.

Fei Tao, Lin Zhang, Yefa HuBeijing, September 2011

Preface

In order to realize the goals of TQCSEFK (fastest Time-to-market, highest Quality, lowest Cost, best Service, cleanest Environment, greatest Flexibility, and high Knowledge), many advanced manufacturing technologies and models such as computer-integrated manufacturing (CIM), networked manufacturing (NM) virtual manufacturing (VM), intelligent manufacturing (IM), green manufacturing (GM), agile manufacturing (AM), concurrent engineering (CE) have been proposed. These technologies or models have played very important roles in manufacturing related fields, and have made great contributions to the development of digital manufacturing. Manufacturing grid (MGrid) is a new manufacturing model combining grid technology with the supports of some common and unified architecture, standards, and criteria computing (e.g., web service, WSDL, UDDI, SOAP, WSRF, OGSA, OGSI).

The aim of this book is to make advanced computing technologies (service computing, grid computing, and cloud computing) to be fully used in manufacturing so as to enhance the utilization and sharing of manufacturing resources, and to speed up the transformation from production-oriented manufacturing to service-oriented manufacturing. It achieves this by constructing a concrete theory for MGrid and by detailing implementation methods for MGrid resources and services. Specifically, the book:

(1) Breaks through the application research field of grid technology, turning grid technology from traditional large-scale science computing to application in manufacturing.

(2) Establishes the theoretical foundation for MGrid Resource Service Optimal Allocation (RSOA). From the perspectives of manufacturing science, systematic, ontology, fuzzy mathematics, information theory, set theory, and social psychology, this book explains the basic theory and the implementation method for addressing MGrid RSOA in many aspects such as implementation framework of RSOA, digital description, match and search, Quality of Service modeling and evaluation, optimal-selection, composition of resource service, and resource service combination network, failure detection and recovery of RSOA.

(3) Provides relevant theories and methods of MGRSOA, such as resource service match and search, QoS evaluation, optimal selection, composition, failure-tolerance management. The theories can be used not only in addressing the RSOA problem in MGrid, but also in resolving the RSOA problem in other related SOA-based distributed system.

Outline of This Book

This book consists of 15 chapters.

In Chapter 1, the motivations and driving forces of MGrid are introduced. The connotation of MGrid, including the concept, basic features, and differences between MGrid, networked manufacturing, and computing grid are investigated. The key research contents and technologies of MGrid, including its four commonly known categories and thirty one items, are also studied.

In Chapter 2, the service-oriented architecture of MGrid is proposed, as well as an MGrid collaborative manufacturing prototype platform. The resource service optimal-allocation system which supports the running of the MGrid collaborative manufacturing platform is investigated. The key functions and components for the system are described, as well as its key implementation technologies.

In Chapter 3, the issue of digital description of resource service (DDoRS) in MGrid is investigated. A method for establishing an MGrid-Ontology is presented and an MGrid-Ontology is built. A method for DDoRS based on established MGrid-Ontology is proposed. An MGrid-Ontology based MGrid resource service discovery framework is proposed.

In Chapter 4, the resource services match and search (RSMS) is studied. The describing information of resource services are classified into four categories: word concept information, sentence information, number information, and entity class (or data structure) information. The similarly matching algorithms (SMAs) for the four kinds of basic describing information are presented, including word matching algorithms (WMAs), sentence matching algorithms (SeMAs), number matching algorithms (NMAs), and entity class matching algorithms (ECMAs). Under the supports of the proposed SMAs, the process of resource services match and search are divided into four steps: basic-matching, I/O-matching, QoS-matching, and integrated-matching.

Chapter 5 studies the evaluation indices system of resource service including special evaluation indices (SEIs), individual evaluation indices (IEIs), and general quality of service (QoS) evaluation indices. The evaluation framework and method for SEIs and IEIs are studied. The modeling of MGrid QoS from the points of QoS whole-lifecycle management, MGrid architecture views, and QoS attributes parameters are investigated. A QoS-based MGrid resource service management framework is proposed.

In Chapter 6, the concept of resource service trust-QoS is presented and introduced in order to enhance the validity and success rate of MGrid resource scheduling, and provide high credible resource service abilities and results to user. The trust problems existing in the resource service transaction are put forward. The trust-QoS relationship model which is capable of capturing a comprehensive range of trust relationships exist in MGrid system is put forward. A two-layer resource service trust-QoS evaluation model (intra-domain trust-QoS and inter-domain trust-QoS evaluation models) are put forward. The quantitative evaluating algorithms of trust-QoS degree value are proposed, as well as the real-time and dynamic updating algorithms of trust-QoS degree value. A trust-QoS based MGrid resource service scheduling framework and associated realizing algorithms are proposed to illustrate the application.

In Chapter 7, an MGrid resource service optimal-selection and composition framework (MGrid-RSOSCF) is investigated. The process of resource service optimal-selection and composition is divided into five steps in MGrid-RSOSCF and the five key problems to realize MGrid-RSOSC are analyzed. The proposed MGrid-RSOSCF consists of five layers and each layer provides the corresponding necessary services and algorithms to address one problem. The five layers are: (1) T-Layer is responsible for MGrid task decomposition, (2) S-Layer is responsible for resource service match and search, (3) Q -Layer is responsible for QoS processing, (4) O-Layer is responsible for evaluating and ranking the candidate resource service, and (5) C-Layer is responsible for resource service composition and optimal-selection.

In Chapter 8, user’s feeling is taken into account in resource service optimal selection (RSOS) in MGrid system. The non-functionality QoS evaluation of resource services is based on users’ feeling and transaction experiences using intuitionistic fuzzy set (IFS). The dynamics of non-functionality QoS is considered, and a time-decay function is introduced into non-functionality QoS evaluation. A method is proposed for resource service optimal-selection based on IFS and non-functionality QoS. The performance and advantage of the proposed method are discussed.

In Chapter 9, the issue of correlation-aware composite resource service in MGrid is considered. Three kinds of correlations relationship (i.e., combinable correlation, business entity correlation, and statistical cooperate correlation) in services composition are investigated. The impact of each kinds of correlation relationship on the whole quality of resource service composition is investigated, and QoS computation model based on the three correlations is proposed. The case study indicates that the higher quality of services composition can be achieved when considering the correlations in resource services composition.

In Chapter 10, the multi-objective MGrid resource service composition and optimal-selection (MO-MRSCOS) problem is studied. The formulation is presented for an MO-MRSCOS problem with the given multi-objective (e.g., time minimization, cost minimization and reliability maximization) and multi-constraints. The basic resource service composite modes (RSCM) for composite resource service are described, and the principles for translating a complicated RSCM into a simple sequence RSCM are presented for simplifying the resolving process and complexity of MO-MRSCOS problem. A method based on the principles of particle swarm optimization (PSO), is proposed for solving MO-MRSCOS problem. Unlike previous works: (a) the proposed PSO algorithms combine the non-dominated sorting technique to achieve the selection of global best position and private best position; (b) the parameters of particle updating formulation in PSO are dynamical generated in order to make a compromise between the global exploration and local exploitation abilities of PSO; and (c) To maintain diversity of solutions in population, permutation-based and objective-based population trimming operators are applied in PSO.

In Chapter 11, the concept and the classification of resource service composition (RSC) flexibility are presented, and the measurement method of RSC flexibility is investigated to achieve the optimal-selection of RSC based on flexibility.

In Chapter 12, the resource service composition network based on complex network theory is investigated. The principles for establishing and modeling combinable relationship-based composition service network (CoRCS-Net) are studied, and nine combinable relationships among services in CoRCS-Net were investigated and fourteen elementary evolving operators for CoRCS-Net dynamic evolution are designed.

In Chapter 13, the potential failures that would generate during MGrid resource service scheduling are investigated. Thirteen failures are defined in detail, which are classified into four classifications: (a) virtual link related failures, (b) resource service related failures, (c) task related failures, and (d) application related failures. A failure management system is proposed, which provides failure-tolerance service in MGrid resource service scheduling. Corresponding detection mechanisms and methods to each defined failure are presented in detail, associated with the corresponding failure recovery methods.

In Chapter 14, the related works on the application of grid technology in manufacturing are investigated, including research on manufacturing grid (MGrid) theories and applications, and then several key future research issues of MGrid are discussed.

In Chapter 15, combing the new technologies and existing theories and technologies of current enterprise information, a computing and service-oriented manufacturing model, i.e., cloud manufacturing (CMfg), which is the future commercial realization of MGrid, is discussed based on the previous work of this book. The concept, architecture, core enabling technologies, and typical characteristics of CMfg are abstractly studied. Four typical CMfg service platforms, i.e., public, private, community, and hybrid CMfg service platforms are investigated. The key advantages and challenges for implementing CMfg are analyzed, as well as the key technologies and main research contents.

Abbreviations

AM: Agile manufacturing

App_AccessRight_Failure: Accessing right failures

App_DesignCode_Failure: Application design or coding failures

ASR: Application System Resource

Bandwidth_Failure: Bandwidth failure

BuC: Business entity correlation

CEAgent: Chief Evaluation Agent

CG: Computing grid

CIM: Computer-integrated manufacturing

C-Layer: Resource service composition layer

CMfg: Cloud manufacturing

CoR: Combinable relationship

CoRCS-Net: Combinable relationship based composition service network

CR: Computational Resource

CRS: Composite resource service

CRSS: Candidate resource service set

DDoRS: Digital description of resource service

Dep-phase: Deploy phase

Des-phase: Design phase

E&M-phase: Execution and monitor phase

E-Agent: Evaluation agent

ECA: Event–condition–action

ECMAs: Entity class matching algorithms

EEAgent: Evaluation expert agent

eiCoR: Equivalent input combinable relationship

EIS-Agent: Evaluation indices set agent

eoCoR: Equivalent output combinable relationship

eqCoR: Equivalent or competition combinable relationship

ERP: Enterprise resource planning

exCoR: Exact combinable relationship

FD: Failure detector

FR: Failure recovery

GA: Genetic algorithms

GRAM: Grid resource allocation management

IaaS: Infrastructure as a service

IEIs: Individual evaluation indices

IFS: Intuitionistic fuzzy set

IM: Intelligent manufacturing

IOPE: Inputs, outputs, preconditions, and effects

IoT: Internet of thing

irCoR: Input replaceable combinable relationship

MatchEngine: Resource service match engine

MCRS: Mixed composite resource service

MCS: Manufacturing cloud service

MCSs: Manufacturing cloud services

MDS: Miscomputing Discovery Service

MGJMS: MGrid job management system

MGrid: Manufacturing grid

MGrid-Ontology: MGrid ontology

MGrid-RSOSCF: MGrid resource service optimal-selection and composition framework

MGRS: MGrid resource service

MGRSOA: MGrid resource service optimal allocation

MO-MRSCOS: Multi-objectives MGrid resource service composition and optimal-selection

MRCM: MGrid resource conceptual model

MROM: MGrid resource objective model

MRSCOS: MGrid resource service composition and optimal-selection

MRSOAS: MGrid resource service optimal-allocation system

MRSRTask: Multi-resource service request task

NIS: Negative idea solution

NM: Networked manufacturing

NMAs: Number matching algorithms

OGSA: Open grid service architecture

OGSI: Open grid service infrastructure

O-Layer: Optimal-selection layer

orCoR: Output replaceable combinable relationship

OWL-S: Ontology web language for services

PaaS: Platform as a service

PIS: Positive idea solution

ppCoR: Partial pre-order combinable relationship

prCoR: Partial replaceable combinable relationship

psCoR: Partial successor combinable relationship

PSO: Ppartial swarm optimization

Q-Layer: QoS synthetically processing layer

QoS: Quality of Service

QoS-RSM: QoS-based resource service management

RMS: Resources Management System

RS_AbilityChange_Failure: Resource service ability changed failure

RS_Composition_Failure: Resource service composition failure

RS_Overload_Failure: Resource service overload or saturation failure

RS_Quit_Failure: Resource service quit failure

RSC-Coordinator: Resource service coordinator

RSCE-Controller: Resource service executing controller

RSC-Engine: Resource service composition engine

RSCEP: Resource service composition executing path

RSCEP-Generator: Resource service composition executing paths generator

RSCEP-Selector: Resource service composition executing paths selector

RSC-Monitor: Resource service monitor

RSD: User enterprise or resource service demander

RSIC: Rresource service information center

RS-Matcher: Resource service matcher

RSMS: Resource service match and search

RSOS: Resource service optimal-selection

RSOSC: Resource service optimal-selection and composition

RSP: Resource enterprise or resource service provider

SaaS: Software as a service

SCM: Supply chain management

SCRS: Sequence composite resource service

SEIs: Special evaluation indices

SeMAs: Sentence matching algorithms

SFN: Scale-free network

SLA: Service level agreement

S-Layer: Resource service match and search layer

SMAs: Similarity matching algorithms

SMEs: Small and medium-sized enterprises

SOAP: Simple object access protocol

SRSRTask :Single resource service request task

StC: Statistical cooperate correlation

Task_Cancel_Failure: Task cancellation failure

Task_Require_Change_Failure: Changed task requirements failure

Task_RS_Mimatch_Failure: Mismatch failure between a task and resource service

Task_Suspension_Failure: Task suspension failure

T-Layer: Task layer

TQCSEFK: Fastest Time-to-market, highest Quality, lowest Cost, best Service, cleanest Environment, greatest Flexibility, and high Knowledge

THFNMAs: Triangular fuzzy numbers matching algorithms

UDDI: Universal description, discovery and integration

VL: Virtual link

VL_Disconnect_Failure: Virtual link disconnected failure

VM: Virtual manufacturing

VO: Virtual organization

WMAs: Word matching algorithms

WSDL: Web service description language

WSRF: Web service resource framework

Chapter 1

Introduction to Manufacturing Grid

1.1 Introduction

Grid technology has been recognized as a promising paradigm for the next generation of manufacturing systems. Researchers have attempted to apply grid technology to product design, manufacturing enterprise resource integration and sharing, enterprise management, enterprise collaboration, optimal manufacturing resource allocation and scheduling, and to enable the digitalization of enterprise information as an implementation methodology. It has been already close to a decade since the first appearance of the concept of manufacturing grid (MGrid). Many projects on the application of grid technology in manufacturing have been carried out, and a large number of research papers on MGrid have been published. In order for MGrid to be better known and received by researchers in the manufacturing community, this chapter will cover (a) the motivations and driving forces of MGrid, (b) a detailed description of the connotation of MGrid, such as its concept and basic features and (c) the differences between MGrid, networked manufacturing, and computing grid. The key research contents and technologies of MGrid, including its four commonly known categories and thirty one items, will also be investigated (Tao et al., 2011d).

1.2 Proposal of Manufacturing Grid

1.2.1 Several Issues of Manufacturing

Modern manufacturing has changed significantly due to intense global competition, economic globalization, resource globalization, and the rapid development of advanced manufacturing, information, computer, and management technologies. The key mission of manufacturing has changed from enlarging production scale in the 1960s, reducing production cost in the 1970s, promoting product quality in the 1980s, rapidly responding to markets in the 1990s, to emphasizing knowledge and service in the present decade. The introduction of computer and information technologies in manufacturing, and the rapid development and application of Internet technologies have sped up the development of manufacturing. Manufacturing is now moving towards the direction of achieving globalization, digitalization, integration, and intelligence. Some of the current issues that have emerged are as follows (Tao et al., 2010a; Tao et al., 2011d; Tao et al., 2008a):

1. There is a contradiction between the scarcity and redundancy of manufacturing resources.

Currently, on the one hand, some manufacturing enterprises, especially small and medium-sized enterprises (SMEs), are unable to accomplish their manufacturing orders because of their lack of advanced high precision equipment. On the other hand, some enterprises own this equipment, but it may have relatively little manufacturing mission and utility.

2. The design process of a new product has become more complex.

The entire life-cycle of a product must be considered during its development phase. High performance computing, complex simulation, multi-disciplinary optimization and cooperation are often involved in a new product development. It is impossible for an enterprise to own all of the manufacturing capabilities and resources. Therefore, an enterprise will not be able to complete the entire development process of a new product relying on its in-house resources alone.

3. There is an inadequacy in the traditional resource and capability sharing mode.

Sharing of resources, such as document and picture transmission, is inadequate to meet the requirements of global competition, and resource and cooperation globalization. Deep sharing such as full connectivity, remote access, and interoperability of computing resources (e.g., high performance computing devices), data and information resources (e.g., parts library), application and service (e.g., online simulation, design, and analysis), equipment (e.g., CNC, manufacturing cell), application systems (e.g., large-scale and application-specific software) are required. Novel technologies have to be introduced in order to meet the new requirements of resource sharing.

4. There is a dependency on information and software systems.

Modern manufacturing enterprises use many information and software systems such as ERP (enterprise resource planning), PDM (product data management), SCM (supply chain management), CRM (customer relationship management), CAD (computer-aided design), MIS (manufacturing information system), and CAPP (computer-aided process planning). Undoubtedly, these systems are required and are very useful for an enterprise, but they are very expensive if an enterprise needs to own all of them. In addition, it is very difficult for an enterprise to integrate and share the data among different information systems. Therefore, new methodologies are needed to enable each enterprise to benefit from these information systems at low cost, and in a convenient way.

5. High subcontracting cost hinder the development of SMEs.

The cost of subcontracting can represent a significant portion in the entire cost of a new product. Reducing subcontracting cost (e.g., fees on renting manufacturing resources from other enterprises) and improving the competitive power of an enterprise are some of the important issues to be considered.

Taking into account the above issues, it is apparent that it will be very difficult for an enterprise to remain competitive using its in-house resources and capabilities alone. In order to survive and grow in a competitive market, it has to cooperate with other enterprises, and make use of the global resources in its production methods (Ding et al., 2005a). In other words, it has to realize the general sharing of global manufacturing resources in order to achieve the goal of TQCSEFK (fastest Time-to-market, highest Quality, lowest Cost, best Service, cleanest Environment, greatest Flexibility, and high Knowledge) (Tao et al. 2010a; Tao et al. 2006b).

1.2.2 Proposal of MGrid

In order to realize the goals of TQCSEFK, many advanced manufacturing technologies and models such as computer-integrated manufacturing (CIM) (Barash 1980, Allen and Bryan 1986, Sabbaghi and Montazemi 2004), networked manufacturing (NM) (D’Amours et al., 1999, Akkermans and Van der Horst 2002, Poler et al., 2008) virtual manufacturing (VM) (Onosato and Iwata 1993, Wadhwa et al., 2009), intelligent manufacturing (IM) (Márkus 1987, Gholamian et al., 2007, Oztemela and Tekezb 2009), green manufacturing (GM) (Dickinson et al., 1995, Jiang et al., 2008a), agile manufacturing (AM) (Fravret et al., 2001, Elkins et al., 2004), and concurrent engineering (CE) (Thomas 1996, Koufteros et al., 2001) have been proposed. These technologies or models have played very important roles in manufacturing-related fields, and have made great contributions to the development of digital manufacturing.

Although each of the above-mentioned advanced manufacturing technologies or models has its own emphasis, they all center on network, cooperative work, and resource sharing. Without the support of common and unified architecture, standards, and criteria, the technologies would lack expandability, openness, and flexibility, thereby limiting their implementation. In addition, resource sharing is the bottleneck that further hinders their implementation. Therefore, it is imperative to find a unified architecture based on international standards, and to establish a virtual manufacturing platform based on network and information technologies.

At the same time, after the proliferation of Internet and web technology, information technology is now experiencing a third wave, i.e., grid technology. The grid concept was first proposed in the 1990s, and was introduced based on the concept of a power grid. A grid is a group of emerging technologies based on the Internet, and is known as the next generation Internet.

Ian Foster, widely known as the “Father of the Grid,” and his co-authors stated in their article “The Anatomy of the Grid” (Foster et al. 2001a), that grid computing is concerned with “Coordinated resource sharing and problem solving in dynamic, multi-institutional virtual organizations“. The key concept is the ability to negotiate resource-sharing arrangements among a set of participating parties (providers and consumers) in order to use the resulting resource pool for some specific purposes. In his subsequent article, “What is the Grid? A Three Point Checklist” (Foster 2002a), he suggested that the essence of the definition of grid computing can be captured in the following three point checklist (Foster 2002a):

Coordinate resources that are not subject to centralized control,Use standard, open, general-purpose protocols and interfaces, andDeliver nontrivial qualities of service.

Apparently, the key concepts of grid are resource sharing and coordinated problem solving in a virtual organization. However, resource sharing and coordinated work are also the two bottlenecks that hinder the development of the current modes of manufacturing. This is where grid technology can be effectively used to solve these bottlenecks.

Grid technology can enable full connectivity of distributed and heterogeneous resources located in disparate places, enterprises, organizations, and yet realize resource sharing and enjoy interoperability. In a grid system, users can obtain most of the services they need for establishing a professional application system according to specific manufacturing requirements. Furthermore, the grid has the characteristics of being distributed, dynamic, autonomous, and transparent, which are useful features in the realization of the sharing of manufacturing resources. Therefore, VM organizations based on grid technologies have gradually evolved into a new manufacturing paradigm under a network-centric environment. MGrid has been put forward under this scenario, and it has been named the next generation manufacturing model (Qiu 2002).

Furthermore, the rapid development and prevalent application of international standards and protocols such as web service, grid service, WSDL (Web Service Description Language) (WSDL 2009), UDDI (Universal Description, Discovery and Integration) (UDDI 2009), SOAP (Simple Object Access Protocol) (SOAP 2009), WSRF (Web Service Resource Framework) (WSRF 2009), OGSA (Open Grid Service Architecture) (OGSA 2009), OGSI (Open Grid Service Infrastructure) (OGSI 2009), Globus Toolkit (Globus 2009), etc., have provided the basis for implementing MGrid.

Since the concept of MGrid was first introduced in 2003, many projects applying grid technology in manufacturing have been carried out, and a large number of research papers on MGrid have been published. In order for MGrid to be better understood and accepted by researchers in the field, this chapter will aim to describe the connotations of MGrid by investigating the existing research pertaining to it.

1.2.3 Technological Driving Forces of MGrid

There are primarily five Technological Driving Forces (TDFs) of MGrid (Tao et al., 2008a):

1. The first TDF of MGrid is the rapid development of grid technology. The appearance and development of grid technology has established a reliable and effective foundation for a general resource sharing system (Foster et al., 2004a). This makes the common sharing and mutual operation of all kinds of resources possible, including computational, memory, communication, data, and knowledge resources, etc., and eliminates the resource and information island (Czajkowski et al., 2001; Foster et al., 2001a).

2. The second TDF of MGrid is the requirement for continuous change and innovation of the product design and manufacture model. In addition to changes from the initial independent design and manufacture, the current integrated design and manufacture, the collaborative design and manufacture, etc., the requirements for high performance computation, complicated simulation, collaborative working in heterogeneous environments and multi-fields is extremely urgent. MGrid presents an opportunity to overcome any insufficiencies in innovation and low technological value added in the process of a new product’s design and development.

3. The third TDF of MGrid are changes to the resource sharing model and its ability in modern enterprise. Competitive globalization, resource globalization and cooperation globalization have already brought a huge change to enterprises regarding their resource sharing model and its ability. Large amounts of information have already brought about a great challenge for networked manufacturers. In the MGrid system, resources’ multiplicity, complexity, and dynamic property go far beyond the current Internet (Ye and Gu et al., 2004). The modern manufacturing enterprise’s resource sharing is not only simple document and picture transmission, but is also the comprehensive sharing of all kinds of heterogeneous resources geographically distributed and owned by different organizers, including computational resource sharing (e.g., high performance calculating device), data and information sharing (e.g., network components storehouse), application and service sharing (e.g., on-line analysis processing), equipment sharing (e.g., senior processing center), application system sharing (e.g., large-scale expensive software), etc., in order to solve the problems faced due to collaboration.

4. The fourth TDF is the expansion of information technology (IT) application and the difficulties of enterprise informatization. The application of IT in enterprises has become more extensive, and has been applied from the original departments to the enterprise’s interior integration, leading to today’s integration among enterprises. From a depth perception, it has gone from original information integration and process integration, to today’s advanced manufacturing technology application integration and knowledge integration. All these changes have brought about higher costs, longer cycles, poor controllability, difficult integration, poor adaptability, and many other problems to an enterprise’s informatization. Grid technology can fully make use of the idle resources of software and hardware in an enterprise’s IT system, and can actualize the integration of all existing information systems in the form of grid service, thereby reducing the investment of an enterprise’s informatization.

5. The fifth TDF is the rapid development and application of related technologies and criteria such as Web Service, Grid Service, Web Service Description Language, Universal Description, Discovery and Integration, Simple Object Access Protocol, Web Service Resource Framework, Open Grid Service Architecture, Open Grid Service Infrastructure, Globus Toolkit and so on. These are fundamental for the realization of MGrid.

1.3 Concept of MGrid

1.3.1 A Brief Outline of Grid and its Applications

The traditional Internet has realized the connection of computer hardware, and the Web has realized the connection of different homepages. Grid has made it possible for all resources to connect comprehensively on the Internet. Grid is a group of emerging technologies based on the Internet. It is a computation and data processing virtual system, which integrates many kinds of services provided by distributed resources, and provides synthesized problems a processing environment. It merges the high speed Internet, high performance computer, large-scale database, sensor, long-distance equipment, and so on, into an integrated whole. It also enables the comprehensive sharing of global computation resources, memory resources, data resources, information resources, knowledge resources, expert resources and equipment resources, etc. Hence, it can make it easy for people to use all kinds of resources such as those involving computation and memory. So far, from its emergence to its commercial application, the development of Grid Computation has roughly experienced three stages as shown in Table 1.1.

Table 1.1 The stages of grid technology development (Tao et al. 2008a)

StageDateAchievement and Key Research QuestionThe first stageThe early 1990s–1797The stage of Grid emergence. The key point was the Meta-computing problem, namely, to provide the ability of high performance computation.The second stage1998–2001Ian Foster described Grid in detail, and proposed five-hour glasses structures grid system structure. The key point was to support the large-scale data and the development of computation middleware, so as to solve the grid platform isomerism, the extension and the compatible question. It also was applied in the scientific research field.The third stageFeb. 2002In the Global Grid Forum, Globus Alliance and IBM issued OGSA, (Open Grid Services Architecture) Jun. 2003Globus Alliance etc issued the bottom standard of OGSA—OGSI (Open Grid Services Infrastructure), connected the Grid and the commercial Web service, and issued GT3. Jan. 2004In the Globus World conference, Globus Alliance, IBM and HP etc issued the new standard draft of Grid Service–WSRF (Web Services Resource Framework), and converted OGSI into WSRF; WSRF defined five series of standards of using WS to visit the condition resources: WS–ResourceProperties, WS-ResourceLifetime, WS-BaseFaults, WS-ServiceGroup, WS-RenewableReferences and WS-Notification.The third stageApril, 2005The platform of Grid–GT4 based WSRF was issued. It was a symbol that Grid entered the commercial application field. Until nowSome new research correlative with Grid, such as Knowledge Grid, Intelligent Grid, Semantic Grid, Manufacturing Grid etc are being researched. The application of Grid in each domain is the key research.

Grid technology is featured in heterogeneity distribution and sharing, scalability, adaptability and dynamic, structural unpredictability, and multi-level management domain. At present, the mature grid application and research mainly concentrate on distributed supercomputing, distributed instrumentation system, tele-immersion and information integration.

Due to the endless promotion and development of grid applications, researchers all over the world are paying more and more attention to grid technology. Many new research areas and new concepts along with grid are emerging, such as computational grid, data grid, knowledge grid, information grid, semantic grid, intelligent grid, service grid, manufacturing grid, etc. Among all of these, manufacturing grid is becoming one of the most important fields of grid technology.

1.3.2 Concept of MGrid

The concept of MGrid was first put forward in 2003. Research on MGrid is still at its infancy and there has been no unified and well accepted definition till now. Following are several related influential concepts of MGrid:

The definition of MGrid by Fan et al., (2003a) from Tsinghua University is: “MGrid is an integrated supporting environment both for the sharing and integration of resources in enterprises and socially for the cooperating operation and management of the enterprises“. A nine-layered architecture for the MGrid system was proposed. The nine layers in the architecture include: network, unit and protocol, resource encapsulation, grid middleware, enabling, application, portal and enterprise cooperation, and MGrid operation and management system.

In 2004 “IEEE International Conference on System, Man and Cybernetics (SMC2004),” Qiu (2004) from Penn State University defined MGrid as the next generation of manufacturing. He pointed out that MGrid was created for low-cost, high productivity production through controlled resource sharing across enterprises just like a computational grid. All participants in MGrid provide designated production service for making different products. A layered MGrid architecture based on five-layered grid architecture was proposed and a simple prototype application of MGrid was presented.

McFarlan from the Institute for Manufacturing (IfM) at the University of Cambridge gave a definition of grid manufacturing in 2004 as: “The (dynamic) harnessing of significant, disparate manufacturing capabilities and resources in order to satisfy one or more business requirements” (McFarlan 2004). In this definition, capabilities, connections/linkages, and interoperability are the key elements as shown in Figure 1.1. The corresponding characteristics, driving forces, key issues, and background for grid manufacturing were analyzed.

Figure 1.1 Connotation of grid manufacturing defined by the Institute for Manufacturing at University of Cambridge (McFarlan 2004).

Based on the above definitions of MGrid and the investigations of related studies on MGrid, the authors came up with the following definition of MGrid (Tao et al., 2011d):

“MGrid is to utilize grid, information, computer, and advanced management and manufacturing technologies, etc. to overcome the barrier resulting from spatial distances in collaboration among different corporations to allow dispersed manufacturing resources (including design, manufacturing, human, and application system resources) to be fully connected, and it is a manufacturing service pool supporting manufacturing resource sharing, integration, and inter-operability among different enterprises.“

In MGrid, all the manufacturing resources distributed in heterogeneous systems and locations can offer various manufacturing services to users in a transparent way by encapsulating and integrating them into different corresponding resource services. Users can therefore use all the resources in MGrid as conveniently as they use local resources. MGrid is indeed a realization of sharing various resources.

MGrid is developed from traditional networked manufacturing (NM), and it is a new stage of NM. The differences between MGrid and NM from the aspects of distribution and centralization, maintenance of resources, dynamics of solutions, and standardization have been investigated by Tao et al., (2008b), Zhang (2006a), Liu (2004), and Wen (2005). In addition, some key technologies for implementing MGrid were adopted from computing grid (CG), but compared with CG, MGrid has the characteristics of more flexible interaction, high real-time requirement, multiparty cooperation, resource multiplicity, high reliability requirements, etc. (Tao et al., 2008b).

1.4 Basic Features of MGrid

Combined with research achievements and the features of networked manufacturing and grid technology, the basic features of MGrid are as follows (Tao et al., 2008b):

1.MGrid is a new advanced manufacturing model (Qiu 2004). It is a method based on grid technology, information technology, and computing technology, that is employed to organize and manage an enterprise’s whole process of design, manufacturing, sales, service, etc., under a networking (a next generation network) environment.

2.The goal of MGrid is to respond to the market’s requirements promptly. MGrid is a concrete substantial form of Contemporary Integrated Manufacturing (CIM) theory and Agile Manufacturing (AI) system under networking. The goal of MGrid is to achieve rapid design, rapid manufacture, rapid detection, and rapid recombination to promote an enterprise’s core competitive power and resolve the TQCSEFK problems.

3.MGrid has broken through the resources’ geography limit and traditional sharing model. MGrid utilizes grid technologies, information technologies, and computer and advanced management technologies to overcome the barricade resulting from the spatial distance in collaboration among different corporations by enabling all kinds of dispersive manufacturing resources (e.g., design resources, manufacturing resources, human resources, application system resources, etc.) to be fully connected, thereby breaking through the resources’ geography limit and traditional sharing model.

4.MGrid provides a transparent manufacturing service. In an MGrid system one can only see some units which can provide special services (e.g., computing, storage, foundry, assembly etc.). Perhaps in the future, when someone wants an individual product, what he will need to do is just login MGrid and submit his requirements, and then there will be a group of given units working collaboratively and the MGrid will return the result to the user.

5.MGrid emphasizes the collaborative work and service among enterprises. Through the resources sharing and cooperation among enterprises, MGrid will strengthen the interoperation both in resources and techniques. That is to say, through MGrid, any enterprise can cooperate with other enterprises or organizations to develop a new product. At the same time, it can perfect the product’s collaborative service among enterprises and the model of manufacturing will develop into an integrated, multi-level distribution system.

1.5 The Connotation of MGrid

The connotations of MG are plentiful, and the theory of MGrid is developed from many relative theories, such as grid theory, collaborative theory, system theory, information theory, etc. MGrid embodies many traits, such as multi-attributed and transparent traits, which are found in sharing, high abstract, self-similitude, autonomy, and management (Tao et al., 2008b)

1.Resource distribution and share. On the one hand, the manufacturing resources existing in MGrid are geographically distributed and belong to different systems or organizations, and each enterprise participating in MGrid is different in many aspects and has its own special market orientation and goal. MGrid should resolve distribution problems such as distributed task allocation, distributed scheduling, distributed monitoring, etc. On the other hand, although manufacturing resources in MGrid are distributed, they are also shared completely. That is to say, in MGrid, any resource can be used by any other user or organization. Through a network, MGrid connects all kinds of geographically distributed enterprises and resources, and forms a virtual organization (VO) which is centralized in logic but distributed in physics. In this VO, all resources can be shared and all enterprises can be employed to work collaboratively towards the same target of dealing with a distributed manufacturing task or problem.

2.Self-similitude. MGrid has the same dynamic rules and diversity both in large and small scale. In MGrid there is some similitude between the whole and the part. The part usually has some characteristics of the whole, while some characteristics of the whole are also found in the part.

3.Dynanism and diversity. In MGrid, the resources or services existing in the current moment may be broken down so they can’t be used in the next moment, or may exit from MGrid, and the new resources or services may continuously join into MGrid as time passes. The dynamic nature of the manufacturing grid includes dynamic increases and dynamic decreases. In addition, manufacturing resources in MGrid are diversified. Therefore, MGrid must solve the communication and cooperation problems among these heterogeneous resources.

4.Autonomy and multiplicity of management. The resources in MGrid primarily belong to an organization or individual. Therefore, the owners have the highest level of authority over the resources management and should be allowed to have their own management capacity, which is the autonomous character of MGrid. At the same time, grid resources must also be under the unified management of the whole MGrid system, otherwise connecting and sharing among different resources cannot be achieved, so MGrid is not only autonomous but also multiplicitous in regard to management.

5.Highly abstract and transparent. MGrid reflects all resources in a highly abstract way in “power wiring boards” which can be seen by users, and other resources are transparent to the users. Manufacturing grid data integration makes the interaction between the various subsystems take place in a transparent manner, which means that all the mutual details are concealed from users so that they accept the various systems as a fully seamless integrated system. Manufacturing grid has the following aspects of transparency:

Transparency of location: Users in MGrid do not need to know the physical storage location of the resources they use. In MGrid, what the users care about is what service the resources can provide. Just as one do not need to know where the transmission is from, nor what form it is in, when we use electricity.Transparency of namespace: In MGrid, all resources or service are described with the same naming rules (or namespace).Registration transparency: Once a user is enrolled in MGrid, all the services and functions can be accessed with one logging in.Distributed memory transparency: A user can handle any resources distributed in MGrid in the same way they operate their local machines or systems.

1.6 Comparison between MGrid and Networked Manufacturing

MGrid is developed from networked manufacturing (NM). In this section, the differences between MGrid and NM have been analyzed according to the following four aspects: distributed and centralized, maintenance of resources, dynamics of solutions, and standardization (Tao et al., 2008b).

Distributed and centralized. The present mode of NM, or the platform of ASP networked manufacturing, actually is a centralized control mode which embodies the monopolized sharing of services or resources. In MGrid, each node directly corresponds to an enterprise or resource, so the organization of the node chains is the organization of virtual enterprises. But in networked manufacturing, the resources of each enterprise would be assembled together and managed in its own platform, and then the enterprise organizations would be realized according to the above principle, so the maintenance of resources is centralized. The resource management in MGrid is distributed and is centralized in networked manufacturing.The maintenance of resources. Each enterprise realizes the management of its resources through maintaining its own services in MGrid, but the platform of networked manufacturing embodies the centralized control of resources and needs the centralized management of resources. The former embodies the distributed mode that provides the advantage of dynamic expansion, not only in the depth but also in the extent of resources, and has more extensibility and dynamic property.Dynamics of solutions. The application process in MGrid is the process of forming the business chain based on services, thus being mapped into virtual enterprises. Its solution is dynamically generated, namely, firstly generating the business flow and then forming the virtual enterprise. However, networked manufacturing embodies a whole independent system, which provides a fixed amount of resources or an established solution to the customers. The former is more dynamic.Standardization and oneness. Networked manufacturing lacks a uniform platform, uniform technologies, uniform standard and criterion at the present time. But there have existed uniform grid platforms (e.g., Globus Toolkit), uniform technologies (e.g., web service, UDDI, SOAP, etc.), uniform standards and criterions (e.g., OGSA, OGSI, WRSF, etc.), and so on, which can be used to realize a real and dynamic MGrid platform.

A further comparison between MGrid and NM is illustrated in Table 1.2.

Table 1.2 Comparison between MGrid and networked manufacturing

 Networked ManufacturingMGridAimResource sharing and cooperative manufacturingResource or memberEnterprise (fixed)Resource, enterprise, manufacturing ability (enter or leave at any time)Organization and management of resource or serviceCentralized control and managementSelf-organization, distributed control and managementEstablishment of virtual organizationBased on partner selectionBased on dynamically service discovery, selection, and compositionCooperation wayBetween enterprisesBetween resources or servicesImplementing technologies.NET, J2EE, CORBA, etc. (no uniform technology)Uniform technology (e.g., Web service, grid Service) and platform (e.g., Globus Toolkit)Solutions to a problem or a jobFixed service and solutionsSolutions are dynamically generated based on the services in the form of service chain or virtual enterpriseAccess control strategyCentralized access controlOGSI, Public key infrastructure, etc.Standardization criteriaNo uniform standardizationOGSA, WSRF, OGSIExpandabilityLow (e.g., need huge customized work)HighDynamicLow (e.g., fixed member and solutions)High (e.g., members can enter or leave dynamically, solutions are generated dynamically)Application scopeDistributed enterprise, group, regional manufacturingGlobal manufacturing

1.7 Comparison between MGrid and Computing Grid

The main differences between MGrid and computing grid (CG) are as follows (Tao et al., 2008b):

1.More flexible interaction. MGrid services offer interaction between resources and the user. Although users need to communicate with services in CG, there are not many ways to interact, such as online and offline interaction in MGrid.

2.High real-time requirement. In addition to reflecting the connecting state and load state of MGrid resources in real time, the service results and the effects (the finished result of task) of the resources must be reflected in a timely manner. The users’ requirements must also be responded to in time.

3.Long life cycle. In CG, a service or trading may only exist for several seconds. In general, it is shorter than one day. However, trading in MGrid may last a few hours or even longer than several days.

4.Multiparty cooperation. Different from CG, a computing service is assigned to finish a simple computing task. It does not need to complete one task together with other different services, but a task should be completed by the coordinated work of a number of different resources or services in MGrid.

5.Resource multiplicity. In CG, resources are limited to computing resources, communication resources, and storage resources. However, in MGrid, in addition to these resources, equipment, software, technical, materials, and public service resources are also involved.

6.Special knowledge requirement. In MGrid, manufacturing task decomposition, the combination and discovery of resource services are often based on specific manufacturing knowledge. It is more complex than those in CG.

7.Functional complexity. MGrid provides not only the functions of simple computation, transmission, and storage, but also the functions of collaborative design, simulation, fault diagnosis, etc.

8.Data complexity. In MGrid, data involve pictures, NC codes, 2D and 3D graphs, etc., which can take many forms and are more complex than those in CG.

9.Professionalism. Manufacturing is a professional activity, but not in the case of CG which is rather impersonal. To meet the requirements of business, process, function, service, etc., MGrid offers a good degree of professionalism.

10.High reliability requirements. Compared with CG, the services required in MGrid meet not only the time requirements, but also the complex and reliability requirements such as high precision and manufacturability.

11.Online/offline resources sharing. In CG, the sharing of resources is primarily done online. In MGrid, the manufacturing equipment, manufacturing cell, and production line to the production workshop, are not part of the network and online resources, but they can still be shared in MGrid offline.

12.Online/offline result submission. The way to send jobs and receive results in CG is primarily online. Besides online, there is an offline mode in MGrid, e.g., through entity communication such as transport and purchase.

1.8 Key Research Contents and Technologies of MGrid

In this section, the key research contents and technologies for implementing MGrid are presented, and they can be classified into four categories: general technologies, supporting technologies, key enabling technologies, and application technologies (Tao et al., 2011d). The key research contents and technologies of MGrid represented in the Venn diagram are shown in Figure 1.2.

Figure 1.2 Venn diagram of key research content and technologies of MGrid (Tao et al., 2011d).

1.8.1 General Technologies

Item 1Concept and Connotation: The first task in carrying out a new research problem is to investigate its concept and connotation. As mentioned in Section 1.3.2, there has been no complete and unified definition of MGrid until recently. The connotation of MGrid is evolving with the emergence of new technologies, such as cloud computing and cloud service. Therefore, it is necessary to research the concept and connotation of MGrid under a new environment.Item 2Architecture: The basis of the architecture that guides the development and implementation of MGrid, including (a) resource and service organization methods, such as the resource management system based on ecology, and (b) service-oriented MGrid architecture with the characteristics of being dynamic, intelligent, self-organizational, and semantics supporting.Item 3Advanced manufacturing model: The theories and methods which are based on advanced manufacturing models and technologies such as NM, CIM, VM, DA, IM, GM, AM, and CE.Item 4Product life-cycle management: One of the aims of MGrid is to encapsulate all the resources, manufacturing ability, data, and information involved in the entire life-cycle of a product into services, so as to make them easily sharable by other users. Hence, product life-cycle management technologies are very important elements.Item 5Executing management technologies of the grid system: This involves constructing, implementing, and maintaining theories and methods for grid and distributed systems.

1.8.2 Supporting Technologies

Item 6Standards and protocols: The standards and protocols technologies include those which support the modeling of services, service encapsulation, communication, access, and failure tolerance, etc. Existing standards and protocols, such as OGSA, OGSI, UDDI, WSDL, SOAP, WRSF, Globus Toolkit, Web service, and grid service are utilized as much as possible.Item 7System integration technologies: The technologies for system integration, such as enterprise information and application integration technologies.Item 8Network security: The network security technologies in MGrid include (a) general and standard security architecture and systems, (b) authentication methods and mechanisms for registration of MGrid members, and (c) safety accessing control strategies, the security mechanisms for data storage and communication, such as encryption mechanisms.Item 9Web Service/Grid service: The development and application technologies on web and grid services, which are used for developing and managing MGrid service.Item 10Grid technologies (Globus Toolkit): The application of grid programming technologies, core grid services (such as high performance scheduling, high throughput resource management technologies, performance data collection, analyses and visualization technologies, and security technologies), and Globus Toolkit and its core grid middleware.Item 11Distributed database, storage, communication: The distributed database, distributed storage, and distributed communication technologies that support MGrid operations.

1.8.3 Key Enabling Technologies

Item 12Resource virtualization technologies: The technologies that encapsulate the distributed and heterogeneous manufacturing resources or capabilities into services, including modeling of resource information, resource virtualization model, semantic description of services, and so on.Item 13Resource and service publication and discovery: The methods and technologies for resource service publication and discovery, including resource service publication and discovery mechanisms, resource service match algorithms, and resource service search algorithms.Item 14Resource and service scheduling: The methods and technologies responsible for allocating jobs to corresponding resources or services at specific times under multi-objectives and multi-constraints.Item 15Quality of service management: The methods and technologies for QoS managements, such as the establishment of QoS index system, description of QoS, modeling of QoS, evaluation of QoS, QoS ontology, life-cycle management of QoS (Tao et al., 2009c), QoS information storage and extraction.Item 16Service composition: The methods and technologies for assembling several services in some sequence according to the requirements of a complex task, including (a) resource service composition architecture, (b) basic construction model for composite resource service, (c) correlation relationship management of composite resource service, such as the definition of correlation relationship among resource services (e.g., combinatorial correlation, entity or organization correlation, and historical correlation), representation of correlation relationship, and correlation relationship mining, (d) uncertainty management of resource service composition, (e) flexibility of resource service composition, (f) resource service composition optimal selection method, and (g) establishment of composition net and its dynamic characteristics (Tao et al., 2011d; 2008c; 2010e).Item 17Failure tolerance management: The methods and technologies for detecting failure and failure recovery, including (a) the definition, classification, and description of failure in MGrid, such as virtual link related failure, resource service related failure, job related failure, application related failure, (b) dynamic failure detection model and method, and (c) dynamic failure recovery strategies.Item 18Workflow technology: The methods and technologies for workflow management, including (a) concept and architecture of workflow in MGrid, (b) modeling of workflow in MGrid, such as process model, organization model, resource model, and service model, (c) resource service scheduling and composition method based on workflow, and (d) monitoring of workflow.Item 19Job management: The methods and technologies for managing jobs in MGrid, such as job description, decomposition, scheduling, deployment, and allocation.Item 20Cooperative work: Technologies for cooperative design, business, and supply chain management in MGrid.Item 21Reliability and security management: The methods and technologies with regard to reliability and security management, such as (a) definition and classification of reliability (including systems, resources, services, and results), (b) evaluating criteria system of reliability, and evaluation model and algorithms of reliability, (c) dynamic and online collection, statistics and analysis of reliability data, and (d) strategies and methods to improve reliability of MGrid systems.Item 22Data and knowledge mining: The methods and technologies for mining data and knowledge used in the operations of resource discovery, matching, composition, etc. They are primarily the methods and technologies for (a) manufacturing knowledge mining, such as information extraction of manufacturing businesses, (b) semantic information mining of manufacturing processes, (c) service composition relationship mining, correlation relationship mining of service, and (d) correlation-aware QoS mining and evaluation.Item 23Intelligent decision method: The intelligent decision supporting system and methods based on intelligent optimization algorithms (e.g., genetic algorithms, evolutionary algorithms), particle intelligence, fuzzy theory, and analytic hierarchy process (AHP) which are used to support the operations for resource service discovery, match, optimal selection, and composition in MGrid.Item 24Cost and price management: Involves the research on (a) cost factors and its measurement from the perspectives of resource service providers, resource service demanders, and managers of MGrid systems, such as costs for hardware (e.g., cost of new resources, replacement of hardware components), business premises, electricity for computing hardware, software, personal, and data communication (e.g., fee for local area networks and connection to the Internet) (Opitz et al., 2008), and (b) pricing and cost management strategies.Item 25Electronic payment: The method and technologies for electronic accounting, electronic payment, and rewards and punishment strategies.Item 26Measure and evaluation of system performance: The establishment of performance index systems, evaluation models and measuring methods, methods for performance monitoring and dynamic measurement.

1.8.4 Application Technologies

Item 27Human-machine interaction: The methods for portal design of human-machine interaction, data collection, and data input and output methods.Item 28Visualization technologies