From Prognostics and Health Systems Management to Predictive Maintenance 2 E-Book

Brigitte Chebel-Morello

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

Seitenzahl: 221

Veröffentlichungsjahr: 2017

Beschreibung

This book is the second volume in a set of books dealing with the evolution of technology, IT and organizational approaches and what this means for industrial equipment. The authors address this increasing complexity in two parts, focusing specifically on the field of Prognostics and Health Management (PHM). Having tackled the PHM cycle in the first volume, the purpose of this book is to tackle the other phases of PHM, including the traceability of data, information and knowledge, and the ability to make decisions accordingly. The book concludes with a summary analysis and perspectives regarding this emerging domain, since without traceability, knowledge and decision, any prediction of the health state of a system cannot be exploited.

Leseprobe

Table of Contents

Cover

Title

Copyright

Introduction

PART 1: Traceability of Information and Knowledge Management

1 Intelligent Traceability of Equipment

1.1. Introduction

1.2. State-of-the-art intelligent products

1.3. Knowledge management approach

1.4. Intelligent product: data flow and distributed memory

1.5. Support service for component’s recycling

1.6. Conclusion

2 A Knowledge-oriented Maintenance Platform

2.1. Introduction

2.2. Software architectures of maintenance support systems

2.3. Projects and works on e-maintenance

2.4. Maintenance, e-maintenance, s-maintenance

2.5. Conclusion

3 Intelligent Traceability Application

3.1. Introduction

3.2. Description of the equipment to be maintained

3.3. Infrastructure of intelligent equipment

3.4. The s-maintenance platform

3.5. Web services

3.6. Service for monitoring diagnostic and prognostic

3.7. Knowledge capitalization service

3.8. Decision support service for the recycling of hydraulic jacks

3.9. Conclusion

PART 2: Post-prognostic Decision

4 Position of Decision within the PHM Context

4.1. Introduction

4.2. Definition of post-prognostic decision

4.3. Which objectives?

4.4. Types of decisions

4.5. The subject of a decision

4.6. Typology of decisions in PHM (temporal, granularity and objective types)

4.7. Decision methods

4.8. Summary

5 Towards a Policy of Predictive Maintenance

5.1. Decision problem

5.2. Hypotheses and data

5.3. Implementation: an approach of maintenance planning supported by prognostic information

5.4. Summary

6 Maintenance in Operational Conditions

6.1. Statement of the problem

6.2. Properties and study of complexity

6.3. Optimal approach

6.4. Sub-optimal solution

6.5. Simulation results

6.6. Summary

Conclusion

Bibliography

Index

End User License Agreement

Guide

Cover

Table of Contents

Begin Reading

Pages

C1

iii

iv

v

ix

x

xi

xii

xiii

xiv

xv

1

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

61

62

63

65

67

68

69

70

71

72

73

74

75

76

77

78

79

80

81

82

83

84

85

86

87

88

89

90

91

92

93

94

95

96

97

99

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

129

130

131

132

133

135

136

137

138

139

140

141

142

143

144

145

147

148

G1

G2

G3

G4

G5

G6

Reliability of Multiphysical Systems Setcoordinated byAbdelkhalak El Hami

Volume 7

From Prognostics and Health Systems Management to Predictive Maintenance 2

Knowledge, Traceability and Decision

Brigitte Chebel-Morello

Jean-Marc Nicod

Christophe Varnier

First published 2017 in Great Britain and the United States by ISTE Ltd and John Wiley & Sons, Inc.

Apart from any fair dealing for the purposes of research or private study, or criticism or review, as permitted under the Copyright, Designs and Patents Act 1988, this publication may only be reproduced, stored or transmitted, in any form or by any means, with the prior permission in writing of the publishers, or in the case of reprographic reproduction in accordance with the terms and licenses issued by the CLA. Enquiries concerning reproduction outside these terms should be sent to the publishers at the undermentioned address:

ISTE Ltd27-37 St George’s RoadLondon SW19 4EUUKwww.iste.co.uk

John Wiley & Sons, Inc.111 River StreetHoboken, NJ 07030USAwww.wiley.com

© ISTE Ltd 2017

The rights of Brigitte Chebel-Morello, Jean-Marc Nicod and Christophe Varnier to be identified as the authors of this work have been asserted by them in accordance with the Copyright, Designs and Patents Act 1988.

Library of Congress Control Number: 2017942387

British Library Cataloguing-in-Publication Data

A CIP record for this book is available from the British Library

ISBN 978-1-84821-938-0

Introduction

Intelligent Maintenance

Due to the evolution of technology, IT, and organizational approaches, industrial equipment is becoming more and more complex and automated. This complexity is a source of various incidents and faults that cause considerable damage to items, the environment and people. Obviously, the reliability of the equipment has an impact on the safety of items and people and, when maintenance is neglected, it can lead to incidents involving prohibitive costs, stemming from interruption of production, replacing items, etc. A lack of maintenance and its impact on the reliability of the equipment can lead to catastrophic consequences for the environment in cases of contamination. This can entail evacuation operations and environmental cleaning without, nevertheless, being able to completely remove the pollution in the area.

In order to prevent risks, companies must use reliable equipment, which should be well maintained by an efficient and well-organized maintenance system. Correct maintenance extends the lifetime of the equipment while contributing to better global performance. For this reason, maintenance has a strategic role in industry, and today it represents an essential task within a production system.

Sustainable process

An effective maintenance policy provides technical, economic and social advantages. It is coherent with the idea of sustainable development and makes it possible, on the one hand, to increase the availability of industrial systems and, on the other hand, to lengthen their lifecycle. From the point of view of economics, it reduces the cost of failures and, as a result, increases the profit of the product. The emergence of predictive maintenance, based on fault prognostics and, more generally, on PHM (Prognostics and Health Management) enables:

1) the anticipation of faults in systems’ critical elements;

2) the prevention of industrial risks (in nuclear plants, oil platforms, etc.);

3) and the safety of people and items to be maintained.

Predictive maintenance

Classical maintenance strategies such as corrective, preventive and predictive maintenance are composed of business processes such as the upkeep, repair, or monitoring of an equipment’s health state, its monitoring, fault detection, failure diagnostics or fault prognostics. Although these processes can be studied separately, it is wiser to integrate them into a PHM cycle, which can be considered as an adaptation of the OSA-CBM architecture. This cycle is described in part 1 of the book From Prognostics and Health Systems Management to Predictive Maintenance 1: Monitoring and Prognostics [GOU 16], starting from data acquisition, its processing by means of different modules of the cycle (data processing, detection, diagnostics, prognostics, decision and human machine interface (HMI)) and finishing with the decision and its presentation via suitable HMI interfaces for maintenance operators. This part is dedicated to the first modules of the cycle, from data acquisition to prognostics, proposing different monitoring and prognostic methods.

The purpose of the present book, which follows [GOU 16], is to tackle the other phases of PHM. These include the traceability of data, information and knowledge (first part of the book), and the ability to make decisions accordingly (second part of the book).

Maintenance implementation requires qualifications and contributes to the development of maintenance technicians. Our work, within a context of quality policy advocating for the continuous improvement of practices, is in line with a present challenge for a company, which is to provide the employees with the right information at the right moment, in order to allow them to work in the best conditions, and therefore to improve their skills. To maintain the items in a well-functioning state and to anticipate any failure, the maintenance operators need to be able to access all kinds of support services related to different maintenance strategies.

Companies should progress by transforming their activities through the development of a learning culture, which is the only alternative for maintaining a permanent state of innovation. Learning culture means sharing knowledge and cooperative work among members of a company. As a result, knowledge is considered to be the driving force of productivity and economic growth. An emergence of the knowledge management problem is taking place. Creating, capitalizing and sharing knowledge thus become a challenge that any company faces.

In this book, we address the expert maintenance knowledge of a maintenance company, the formalization and the manipulation of this knowledge. The maintenance operations, combined with technical advancements and new information and communication technology, have entailed an evolution of maintenance systems towards systems that integrate smart modules, which communicate and collaborate among each other. It is in this industrial and scientific context that the works described in the first three chapters are entirely situated. A knowledge management approach has been implemented in order to analyze the maintenance processes used by a maintenance company, with the goal of making an overview of support systems to be developed, and of being able to make them available as maintenance support services for the company’s employees.

Actually, timely access to information concerning a product or a piece of equipment provides a better monitoring of the latter and allows us, for example, in the case of dangerous products, a better handling of this product, thus improving safety.

Information traceability

Information traceability is a vital element for ensuring the monitoring of a product or equipment in time, during its whole lifecycle. Over the last few years, Product Lifecycle Management (PLM) systems have been increasingly used to manage business operations and data generated by events and actions that involve the product1.

Lifecycle refers to a set of phases that can be identified as the different stages of life of a product, from its creation to dismantlement. It is composed of three main phases:

–

Beginning Of Life (BOL),

which includes the design and the manufacturing of the product.

–

Middle Of Life (MOL),

which is related to the distribution, the usage and the maintenance operations.

–

End Of Life (EOL),

which concerns the moment when the product ends its usage phase and is retrieved within the company in order to be recycled or eliminated.

The PLM concept is much more than an issue of visualizing and transforming data. It includes processes (the flow of data among the operators and the flow of resources according to competency) and methods (practices and techniques established along the process by using product data generated during each life stage of the product). This translates into three fundamental elements that constitute the basis of the PLM concept: ICT managing remote information systems, the processes and the methodology, which evolve along the lifecycle phases of a product (Figure I.1).

PLM services based on the Web, contrarily to PDM (Product Data Management) systems, do not limit themselves to facilitating the exchange of information regarding the product among heterogeneous product data management systems [GUN 08], but they can be a platform of collaborative development with the integration of data originated at scattered locations. PLM widens the field of application of PDM systems in order to provide a large company with more information concerning their product.

The availability of information during each phase of a product’s lifecycle enables the sharing of information among the players of different cycle stages and the exploitation of this knowledge to improve the decisions to be made with respect to the product.

Figure I.1.Elements of the PLM concept

During the management phase of the product’s middle of life, or MOL, a lot of data is gathered on the field for monitoring and controlling the product’s life state and for keeping a record. Information issued from the product’s beginning of life, BOL, is necessary for analyzing the product’s structure and for understanding its behavior.

Within the context of safety of items and people, standards and laws are imposed in order to be able to trace the history of each product with the aim of ensuring a reliable, safe and traceable supply, and enabling the recovery of information required to understand post-mortem any anomalous event, whichever its seriousness.

Decision and strategy of maintenance

The second part of the book focuses on the concept of decisions based on expert knowledge of the system and on estimations provided by prognostics.

Indeed, sharing information regarding the product and its lifecycle process is vital for ensuring its durability. Knowledge of the product’s history sheds light on this management, and it can provide information regarding the implemented maintenance policy, which has a non-negligible effect on the product’s operating state. Actually, an effective maintenance policy yields technical, economic and social advantages:

– from a technical point of view, it allows an increase of the useful lifespan, availability and durability performance of a product;

– from an economic point of view, it reduces the cost of failures and, consequently, increases the profit of the product;

– finally, from a social point of view, it reduces to a minimum the number of incidents and risky situations.

Today, technological evolution enables the equipment to communicate and to provide information regarding the different ongoing events related to it. Therefore, it is possible to trace its operation and malfunctioning during its lifecycle. One of the key features of PLM systems is the availability of information, which can be easily accessed by the operators related to the product, and the intelligence integrated in their lifecycle. The idea is to propose a set-up of a so-called intelligent equipment, as in McFarlane’s definition, which facilitates access to embedded and remote information.

In this book, we address the product’s middle of life and, in particular, the implemented maintenance policy, as well as its impact on the other lifecycle phases. In fact, information issued from the MOL phase can be used:

– to evolve, within the BOL phase, the product’s design by improving the product with respect to its usage, as in [STA 15];

– to define, within the MOL phase, the different kinds of decision (tactical, operational or strategic) in the best way possible. According to the monitoring problem, the decisions can be automatic or controlling decisions, decisions of online scheduling of diagnostics, or those of re-configuration of tasks or maintenance intervention planning;

– to improve the recycling procedure within the EOL phase of a component according to its health state and to the maintenance policy implemented on the product. Without information, decisions are made with respect to an approximate inspection, which is insufficient if the safety of people is at stake [PAR 04].

Chapter 1 illustrates a smart tracing system and its architecture connected to a remote maintenance platform. An infrastructure of smart products is proposed, which enables the equipment to be connected and capable of knowledge capitalization based on maintenance ontology.

Chapter 2 proposes a maintenance platform with a particular attention to knowledge, in order to guarantee the traceability of information along its lifecycle, and thus to be able to implement decision support systems.

An application of this intelligent traceability has been implemented on a ski lift and its brief description is given in Chapter 3.

Chapter 4 illustrates a bibliographic overview of different decision-making approaches in the context of PHM. The aspects of scalability of this decision phase (temporal granularity and description degree) are illustrated as well. This chapter is an opportunity to show the importance of decisions within the PHM process.

A first implementation of decisions is the object of Chapter 5. It adds a new strategic dimension to maintenance by means of the anticipation which it enables. Therefore, we speak of predictive maintenance. This chapter illustrates an example of optimization of predictive maintenance starting from information that is issued from the prognostic phase of PHM. This optimization consists of reducing the maintenance related costs via appropriate planning.

Finally, Chapter 6 develops an original approach for involving production resources with respect to demand. A further dimension is added to the planning phase by varying the utilization conditions of each piece of equipment with respect to its health state, with the aim of lengthening the production lifespan of the whole system before maintenance.

The book concludes with a summary analysis and perspectives regarding this emerging domain, since without traceability, knowledge and decision, any prediction of the health state of a system cannot be exploited.

1

CIMData: Product Lifecycle Management (PLM) Definition. Available online at:

http://www.cimdata.com/PLM/plm.htmls

PART 1Traceability of Information and Knowledge Management

1Intelligent Traceability of Equipment

1.1. Introduction

Over the last few decades, the business awareness of companies has been based on their intangible capital, the knowledge and expertise of their employees. Many works have been dedicated to creating, sharing and capitalizing on this expert knowledge.

However, just like expertise and practical skills, data also contributes to the intangible capital of companies and it is carefully recorded. Banks jealously guard their databases. Companies distrust the cloud due to security issues related to data, which may fall into the wrong hands and be exploited by competitors or harm data owners. In the age of connected objects, of easy data harvesting and instantaneous remote access, it is evident that well-exploited data represents a gold mine. Indeed, data is becoming a resource which can be exploited by the economy. Data must be secured and exploited and this is the basis of informed decision-making within a domain.

The process of knowledge capitalization corresponds to the notion of knowledge management, which was defined by Davenport as the process of collecting, distributing and effectively using knowledge [DAV 94], an approach developed by [DUH 98]. Traceability is an essential element of the capitalization of knowledge related to different stages of a product’s evolution [BIS 08]. Several works have highlighted the notion of the growth of products’ lifecycle management as one of knowledge management [AME 05] and [STA 15]. In fact, the research consortium of the project FP6 IP 507 100 PROMISE (PROduct lifecycle Management and Information tracking using Smart Embedded systems) has remarked that traditional PLM systems lack product knowledge and visibility in the two MOL and EOL phases, and has recommended developing traceability and knowledge capitalization during the lifecycle.

The aim of our work is to design and develop a solution for processing and capitalizing knowledge related to industrial equipment throughout its lifecycle and for making it available for operators to easily access this equipment in an understandable way and at the required moment.

New PLM possibilities [RAN 11] are being introduced, thanks to continuous developments in the domain of information systems regarding radio-frequency identification (RFID), sensor network technology and, more generally, in product embedded information devices (PEID). A new generation of products called smart or intelligent products is being developed [KIR 11]. According to [YAN 09], it makes the information easily accessible for designers, users or disassemblers of the product or equipment. However, although these intelligent products are capable of gathering data during their lifecycle, they lack the means of extracting information and acquiring knowledge from this data. To reach this goal, we have tackled three challenges:

– Creation of a so-called intelligent product which allows users access to reliable information capable of being read or manipulated, as well as an available deduction related to the current health state of the product;

– Transformation of data in knowledge in a memory that stores all the information concerning the product during its lifecycle and which can be accessed from the product;

– Proposition for decision support services, online prognostics and monitoring of the health state of equipment and support services for maintenance and recycling of products. These services should be available via an information system which can be easily accessed through the product.

We will address these challenges in three stages: (i) after having defined what intelligent equipment is and having considered the work in this domain, we will orient ourselves toward the data exchange infrastructure CL2M, featuring RFID tags connected to an e-maintenance platform and equipped with deduction tools. (ii) We have developed a knowledge capitalization process that stores the knowledge in an operating memory which is distributed on the equipment and the e-maintenance platform. The knowledge is formalized according to the maintenance ontology IMAMO_RFID and made is available by means of an intelligent product infrastructure that ensures knowledge sharing. (iii) We have developed web services that require the availability of information regarding the state of (mal)functioning of the equipment along its lifecycle. In this way, we propose different decision support services:

– a support service for recycling components (products), in which data is indispensable in order to provide such a service [SIM 00];

– a support service for the monitoring and prognostic processes of the health state of the component;

– a support service for maintenance action planning.

As a result, this chapter will begin by outlining some state-of-the-art intelligent products followed in section 1.3 by a presentation of a knowledge capitalization process that monitors a component’s health state along its lifecycle, and the proposition of an ontology called IMAMO_RFID, defined for intelligent products. Section 1.4 will be dedicated to the infrastructure of an intelligent product with the exchange of data and information and the implementation of decision support services.

1.2. State-of-the-art intelligent products

1.2.1. Definition of intelligent products

In order to monitor a product during its MOL phase, this product has to be intelligent in McFarlane’s sense. [MCF 03] defines the product via a physical and informational representation stored in a database, which is associated with an intelligence provided by a decision support agent. An intelligent product is characterized by five main properties: possession of a unique identify, a capability to communicate effectively with its environment, ability to retain or store data about itself, and a potential to participating in or making decisions relevant to its own destiny. Other definitions of intelligent products exist [MEY 09, MCF 03] as listed [KIR 11] by Kiritsis, who synthesized them by defining an intelligent product as a system containing sensing, memory, data processing, reasoning and communication capabilities. He