Sổ tay bảo trì hệ thống phức tạp

Springer Series in Reliability Engineering


Series Editor
Professor Hoang Pham
Department of Industrial Engineering
Rutgers
The State University of New Jersey
96 Frelinghuysen Road
Piscataway, NJ 08854-8018
USA

Other titles in this series
The Universal Generating Function in Reliability Analysis and Optimization
Gregory Levitin
Warranty Management and Product Manufacture
D.N.P Murthy and Wallace R. Blischke
Maintenance Theory of Reliability
Toshio Nakagawa
System Software Reliability
Hoang Pham
Reliability and Optimal Maintenance
Hongzhou Wang and Hoang Pham
Applied Reliability and Quality
B.S. Dhillon
Shock and Damage Models in Reliability Theory
Toshio Nakagawa
Risk Management
Terje Aven and Jan Erik Vinnem
Satisfying Safety Goals by Probabilistic Risk Assessment
Hiromitsu Kumamoto

Offshore Risk Assessment (2nd Edition)
Jan Erik Vinnem
The Maintenance Management Framework
Adolfo Crespo Márquez
Human Reliability and Error in Transportation Systems
B.S. Dhillon


Khairy A.H. Kobbacy • D.N. Prabhakar Murthy
Editors

Complex System
Maintenance Handbook

123


Khairy A.H. Kobbacy, PhD
Management and Management Sciences
Research Institute
University of Salford
Salford, Greater Manchester
M5 4WT
UK

D.N. Prabhakar Murthy, PhD
Division of Mechanical Engineering
The University of Queensland
Brisbane 4072
Australia

ISBN 978-1-84800-010-0

e-ISBN 978-1-84800-011-7

DOI 10.1007/978-1-84800-011-7
Springer Series in Reliability Engineering series ISSN 1614-7839
British Library Cataloguing in Publication Data
A Complex system maintenance handbook. - (Springer series in
reliability engineering)
1. Maintenance 2. Reliability (Eningeering) 3. Maintenance
- Management
I. Murthy, D. N. P. II. Kobbacy, Khairy A. H.
620'.0046
ISBN-13: 9781848000100
Library of Congress Control Number: 2008923781
© 2008 Springer-Verlag London Limited
Watchdog Agent™ is a trademark of the Intelligent Maintenance Systems (IMS) Center, University of
Cincinnati, PO Box 210072, Cincinnati, OH 45221, USA. www.imscenter.net
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 of licences issued by the Copy-right Licensing
Agency. Enquiries concerning reproduction outside those terms should be sent to the publishers.
The use of registered names, trademarks, etc. in this publication does not imply, even in the absence of a
specific statement, that such names are exempt from the relevant laws and regulations and therefore free for
general use.
The publisher makes no representation, express or implied, with regard to the accuracy of the information
contained in this book and cannot accept any legal responsibility or liability for any errors or omissions that
may be made.

Cover design: deblik, Berlin, Germany
Printed on acid-free paper
9 8 7 6 5 4 3 2 1
springer.com


To our wives
Iman and Jayashree
for their patience, understanding and support


Preface

Modern societies depend on the smooth operation of many complex systems
(designed and built by humans) that provide a variety of outputs (products and
services). These include transport systems (trains, buses, ferries, ships and aeroplanes), communication systems (television, telephone and computer networks),
utilities (water, gas and electricity networks), manufacturing plants (to produce industrial products and consumer durables), processing plants (to extract and process
minerals and oil), hospitals (to provide services) and banks (for financial transactions) to name a few.
Every system built by humans is unreliable in the sense that it degrades with
age and/or usage. A system is said to fail when it is no longer capable of delivering
the designed outputs. Some failures can be catastrophic in the sense that they can
result in serious economic losses, affect humans and do serious damage to the
environment. Typical examples include the crash of an aircraft in flight, failure of a
sewerage processing plant and collapse of a bridge. The degradation can be controlled, and the likelihood of catastrophic failures reduced, through maintenance
actions, including preventive maintenance, inspection, condition monitoring and
design-out maintenance. Corrective maintenance actions are needed to restore a
failed system to operational state through repair or replacement of the components
that caused the failure.
Maintenance has moved from being an engineering activity after a system has
been put into operation into an important issue that needs to be addressed during

the design and manufacturing or building of the system. Maintenance impacts on
reliability (a technical issue) with serious economic and commercial implications.
This implies that operators of complex systems need to look at maintenance from
an overall business perspective that integrates the technical and commercial issues
in an effective manner.
The literature on maintenance is vast. Over the last 50 years, there have been
dramatic changes due to advances in the understanding of the physics of failure, in
technologies to monitor and assess the state of the system, in computers to store


viii

Preface

and process large amounts of relevant data and in the tools and techniques needed
to build model to determine the optimal maintenance strategies.
The aim of this book is to integrate this vast literature with different chapters
focusing on different aspects of maintenance and written by active researchers
and/or experienced practitioners with international reputations. Each chapter reviews the literature dealing with a particular aspect of maintenance (for example,
methodology, approaches, technology, management, modelling analysis and optimisation), reports on the developments and trends in a particular industry sector or,
deals with a case study. It is hoped that the book will lead to narrowing the gap
between theory and practice and to trigger new research in maintenance.
The book is written for a wide audience. This includes practitioners from industry (maintenance engineers and managers) and researchers investigating various
aspects of maintenance. Also, it is suitable for use as a textbook for postgraduate
programs in maintenance, industrial engineering and applied mathematics.
We would like to thank the authors of the chapters for their collaboration and
prompt responses to our enquiries which enabled completion of this handbook on
time. We also wish to acknowledge the support of the University of Salford and the
award of CAMPUS Fellowship in 2006 to one of us (PM). We gratefully acknowledge the help and encouragement of the editors of Springer, Anthony Doyle and
Simon Rees. Also, our thanks to Sorina Moosdorf and the staff involved with the

production of the book.


Contents

Part A An Overview
Chapter 1: An Overview
K. Kobbacy and D. Murthy ...................................................................................... 3
Part B Evolution of Concepts and Approaches
Chapter 2: Maintenance: An Evolutionary Perspective
L. Pintelon and A. Parodi-Herz.............................................................................. 21
Chapter 3: New Technologies for Maintenance
Jay Lee and Haixia Wang....................................................................................... 49
Chapter 4: Reliability Centred Maintenance
Marvin Rausand and Jørn Vatn .............................................................................. 79
Part C Methods and Techniques
Chapter 5: Condition-based Maintenance Modelling
Wenbin Wang........................................................................................................ 111
Chapter 6: Maintenance Based on Limited Data
David F. Percy ..................................................................................................... 133
Chapter 7: Reliability Prediction and Accelerated Testing
E. A. Elsayed ........................................................................................................ 155


x

Contents

Chapter 8: Preventive Maintenance Models for Complex Systems
David F. Percy ..................................................................................................... 179

Chapter 9: Artificial Intelligence in Maintenance
Khairy A. H. Kobbacy ......................................................................................... 209
Part D Problem Specific Models
Chapter 10: Maintenance of Repairable Systems
Bo Henry Lindqvist............................................................................................... 235
Chapter 11: Optimal Maintenance of Multi-component Systems: A Review
Robin P. Nicolai and Rommert Dekker ................................................................ 263
Chapter 12: Replacement of Capital Equipment
P.A. Scarf and J.C. Hartman................................................................................ 287
Chapter 13: Maintenance and Production: A Review of Planning Models
Gabriella Budai, Rommert Dekker and Robin P. Nicolai ................................... 321
Chapter 14: Delay Time Modelling
Wenbin Wang........................................................................................................ 345
Part E Management
Chapter 15: Maintenance Outsourcing
D.N.P. Murthy and N. Jack ................................................................................. 373
Chapter 16: Maintenance of Leased Equipment
D.N.P. Murthy and J. Pongpech .......................................................................... 395
Chapter 17: Computerised Maintenance Management Systems
Ashraf Labib ......................................................................................................... 417
Chapter 18: Risk Analysis in Maintenance
Terje Aven ............................................................................................................ 437
Chapter 19: Maintenance Performance Measurement (MPM) System
Uday Kumar and Aditya Parida .......................................................................... 459
Chapter 20: Forecasting for Inventory Management of Service Parts
John E. Boylan and Aris A. Syntetos .................................................................... 479


Contents

xi

Part F Applications (Case Studies)
Chapter 21: Maintenance in the Rail Industry
Jørn Vatn ............................................................................................................. 509
Chapter 22: Condition Monitoring of Diesel Engines
Renyan Jiang, Xinping Yan ................................................................................. 533
Chapter 23: Benchmarking of the Maintenance Process at Banverket
(The Swedish National Rail Administration)
Ulla Espling and Uday Kumar ............................................................................. 559
Chapter 24: Integrated e-Operations–e-Maintenance: Applications in North Sea
Offshore Assets
Jayantha P. Liyanage ........................................................................................... 585
Chapter 25: Fault Detection and Identification for Longwall Machinery
Using SCADA Data
Daniel R. Bongers and Hal Gurgenci .................................................................. 611
Contributor Biographies ....................................................................................... 643
Index ..................................................................................................................... 653


Part A

An Overview


1
An Overview
K.A.H. Kobbacy and D.N.P. Murthy
K. Kobbacy and D. Murthy

1.1 Introduction
The efficient functioning of modern society depends on the smooth operation of
many complex systems comprised of several pieces of equipment that provide a
variety of products and services. These include transport systems (trains, buses,
ferries, ships and aeroplanes), communication systems (television, telephone and
computer networks), utilities (water, gas and electricity networks), manufacturing
plants (to produce industrial products and consumer durables), processing plants
(to extract and process minerals and oil), hospitals (to provide services) and banks
(for financial transactions) to name a few. All equipment is unreliable in the sense
that it degrades with age and/or usage and fails when it is no longer capable of
delivering the products and services. When a complex system fails, the consequences can be dramatic. It can result in serious economic losses, affect humans
and do serious damage to the environment as, for example, the crash of an aircraft
in flight, the failure of a sewage processing plant or the collapse of a bridge.
Through proper corrective maintenance, one can restore a failed system to an
operational state by actions such as repair or replacement of the components that
failed and in turn caused the failure of the system. The occurrence of failures can
be controlled through maintenance actions, including preventive maintenance,
inspection, condition monitoring and design-out maintenance. With good design
and effective preventive maintenance actions, the likelihood of failures and their
consequences can be reduced but failures can never be totally eliminated.
The approach to maintenance has changed significantly over the last one
hundred years. Over a hundred years ago, the focus was primarily on corrective
maintenance delegated to the maintenance section of the business to restore failed
systems to an operational state. Maintenance was carried out by trained technicians
and was viewed as an operational issue and did not play a role in the design and
operation of the system. The importance of preventive maintenance was fully
appreciated during the Second World War. Preventive maintenance involves
additional costs and is worthwhile only if the benefits exceed the costs. Deciding

4

K. Kobbacy and D. Murthy

the optimum level of maintenance requires building appropriate models and use of
sophisticated optimisation techniques. Also, around this time, maintenance issues
started getting addressed at the design stage and this led to the concept of maintainability. Reliability and maintainability (R&M) became major issues in the
design and operation of systems.
Degradation and failure depend on the stresses on the various components of the
system. These depend on the operating conditions that are dictated by commercial
considerations. As a result, maintenance moved from a purely technical issue to a
strategic management issue with options such as outsourcing of maintenance, leasing
equipment as opposed to buying, etc. Also, advances in technologies (new materials,
new sensors for monitoring, data collection and analysis) added new dimensions
(science, technology) to maintenance. These advances will continue at an everincreasing pace in the twenty-first century.
This handbook tries to address the various issues associated with the maintenance of complex systems. The aim is to give a snapshot of the current status and
highlight future trends. Each chapter deals with a particular aspect of maintenance
(for example, methodology, approaches, technology, management, modelling
analysis and optimisation) and reports on developments and trends in a particular
industry sector or deals with a case study. In this chapter we give an overview of
the handbook. The outline of the chapter is as follows. Section 1.2 deals with the
framework that is needed to study the maintenance of complex systems and we
discuss some of the salient issues. Section 1.3 presents the structure of the book
and gives a brief outline of the different chapters in the handbook. We conclude
with a discussion of the target audience for the handbook.

1.2 Framework for Study of Maintenance
A proper study of maintenance requires a comprehensive framework that incorporates all the key elements. However, not all the elements would be relevant for a
particular maintenance problem under consideration.
The systems approach is an effective approach to solving maintenance problems. In this approach, the real world relevant to the problem is described through

a characterisation where one identifies the relevant variables and the interaction
between the variables. This characterisation can be done using language or a
schematic network representation where the nodes represent the variables and the
connected arcs denote the relationships. This is good for qualitative analysis. For
quantitative analysis, one needs to build mathematical models to describe the
relationships. Often this requires stochastic and dynamical formulations as system
degradation and failures occur in an uncertain manner. In this section, we discuss
the various key elements and some related issues.
We use the term “asset” to denote a complex system or individual equipment. It
can include infrastructures such as buildings, bridges etc. in addition to those listed
in Section 1.1.


An Overview

5

1.2.1 Stakeholders
For an asset there can be several stakeholders as indicated in Figure 1.1.

Figure 1.1. Stakeholders for maintenance of an asset

The number of parties involved would depend on the asset under consideration.
For example, in case of a rail network (used to provide a service to transport people
and goods) the customers can include the rail operators (operating the rolling
stock) and the public. The owner can be a business entity, a financial institution or
a government agency. The operator is the agency that operates the track and is
responsible for the flow of traffic. The service provider refers to the agency
carrying out the maintenance (preventive and corrective). It can be the operator (in
which case maintenance is done in-house) or some external agent (if maintenance

is outsourced) or both (when only some of the maintenance activities are outsourced). The regulator is the independent agency which deals with safety and risk
issues. They define the minimum standards for safety and can impose fines on the
owner, operator and possibly the service provider should the safety levels be
compromised. Government plays a critical role in providing the subsidy and
assuming certain risks. In this case all the parties involved are affected by the
maintenance carried out on the asset. If the line is shut either frequently and/or for
long duration, it can affect customer satisfaction and patronage, the returns to the
operators and owners and the costs to the government.
1.2.2 Different Perspectives
We focus our attention on the case where the asset is owned by the owner and
maintenance is outsourced. In this case, we have two parties – (i) owner (of the
asset) and (ii) service agent (providing the maintenance). Figure 1.2 is a very
simplified system characterisation of the maintenance process where the main-


6

K. Kobbacy and D. Murthy

tenance activities are defined through a maintenance service contract. The problem
is to determine the terms of the service contract.

Figure 1.2. System characterisation for maintenance out-sourcing

Each of the elements of Figure 1.2 involves several variables. For example, the
maintenance service contract involves the following: (i) duration of contract, (ii)
price of contract, (iii) maintenance performance requirements, (iv) incentives and
penalties, (v) dispute resolution, etc. The maintenance performance requirements
can include measures such as availability, mean time between failures and so on.
The characterisation of the owner’s decision-making process can involve costs,

asset state at the end of the contract, risks (service agent not providing the level and
quality of service) and so on. The interests and goals of the owner are different
from that of the service agent.
The study of maintenance is complicated by the unknown and uncontrollable
factors. It could be rate of degradation (which depends on several factors such as
material properties, operating environment etc) and other commercial factors (high
demand for power in the case of a power plant due to very hot weather).
1.2.3 Key Issues and the Need for Multi-disciplinary Approach
The key issues in the maintenance of an asset are shown in Figure 1.3. The asset
acquisition is influenced by business considerations and its inherent reliability is
determined by the decisions made during design. The field reliability and degradation is affected by operations (usage intensity, operating environment, operating
load etc.). Through use of technologies, one can assess the state of the asset. The
analysis of the data and models allow for optimizing the maintenance decisions
(either for a given operating condition or jointly optimizing the maintenance and
operations). Once the maintenance actions have been formulated it needs to be
implemented.


An Overview

7

Figure 1.3. Key Issues in maintenance of an asset

To execute effective maintenance one needs to have a good understanding of a
variety of concepts and techniques for each of the issues. Another issue is the
computer packages that allow one to collect and analyze data and build models and
derive the optimal solutions.
The linking of the technical and commercial issues is indicated in Figure 1.4
and this requires an inter-disciplinary approach.

Figure 1.4. Linking technical and commercial issues


8

K. Kobbacy and D. Murthy

The disciplines involved are as follows
1.2.3.1 Engineering
The degradation of an asset depends to some extent on the design and building (or
production) of the asset. Poor design leads to poor reliability that in turn results in
high level of corrective maintenance. On the other hand, a well-designed system is
more reliable and hence less prone to failures. Maintainability deals with maintenance issues at the design and development stage of the asset.
1.2.3.2 Science
This is very important in the understanding of the physical mechanisms that are at
play and have a significant influence on the degradation and failure. Choosing the
wrong material can have a serious consequence and impact on the subsequent
maintenance actions needed.
1.2.3.3 Economic
Maintenance costs can be a significant fraction of the total operating budget for a
business depending on the industry sector. There are two types of costs – annual
cost and cost over the life cycle of the asset. The costs can be divided into direct
(labour, material etc.) and indirect (consequence of failure).
1.2.3.4 Legal
This is important in the context of maintenance out-sourcing and maintenance of
leased equipment. In both cases, the central issue is the contract between the
parties involved. Of particular importance is dispute resolution when there is a
disagreement between the parties in terms of the violation of some terms of the
contract.

1.2.3.5 Statistics
The degradation and failures occur in an uncertain manner. As such, the analysis of
such data requires the use of statistical techniques. Statistics provide the concepts
and tools to extract information from data and for the planning of efficient collection systems.
1.2.3.6 Operational Research
Operation research provides the tools and techniques for model building, analysis
and optimization. Often, analytical approaches fail and one needs to use simulation
approach to evaluate the outcomes of different decisions and to choose the optimal
(or near optimal) strategies.
1.2.3.7 Reliability Theory
Reliability theory deals with the interdisciplinary use of probability, statistics and
stochastic modelling, combined with engineering insights into the design and the
scientific understanding of the failure mechanisms, to study the various aspects of
reliability. As such, it encompasses issues such as (i) reliability modelling, (ii)
reliability analysis and optimization, (iii) reliability engineering, (iv) reliability
science, (v) reliability technology and (vi) reliability management.


An Overview

9

1.2.3.8 Information Technology and Computer Science
The operation and maintenance of complex assets generates a lot of data. One
needs efficient ways to store and manipulate the data and to extract relevant
information from data. Computer science provides a range of artificial intelligence
techniques such as data mining, expert systems, neural networks etc., which are
very important in the context of maintenance.
1.2.4 Maintenance Management
Maintenance management deals with the overall management of the maintenance

of an asset. The management needs to be done at three different levels (strategic,
tactical and operational) as indicated in Figure 1.5.
- BUSINESS PERSPECTIVE
- TECHNICAL & COMMERCIAL
- IN-HOUSE vs. OUT-SOURCING
- REPLACEMENT / DESIGN CHANGES

MAINTENANCE
STRATEGY

STRATEGIC
LEVEL

- DEGRADATION (RELIABILITY SCIENCE)
- MAINTENANCE POLICIES
- LOGISTICS (SPARES, FACILITIES ETC)

MAINTENANCE
PLANNING AND
SCHEDULING

TACTICAL
LEVEL

- DATA COLLECTION
- DATA ANALYSIS (ROOT CAUSE, OTHER
FACTORS)

MAINTENANCE
WORK

EXECUTION

OPERATIONAL
LEVEL

Figure 1.5. Maintenance management

The strategic level deals with maintenance strategy. This needs to be formulated so that it is consistent and coherent with other (production, marketing,
finance, etc.) business strategies. The tactical level deals with the planning and
scheduling of maintenance. The operational level deals with the execution of the
maintenance tasks and collection of relevant data.

1.3 Structure of the Handbook
The handbook integrates the vast literature on maintenance with each chapters
focussing on a different aspect of maintenance and written by active researchers
with international reputation and/or experienced practitioners from industry. Each
chapter either reviews the literature dealing with a particular aspect of maintenance
(for example, methodology, approaches, technology, management, modelling ana-


10

K. Kobbacy and D. Murthy

lysis and optimisation), reports on developments and trends in a particular industry
sector, or deals with a case study.
The book is structured into five parts and each of the last four parts contains
several chapters. The topic of the different chapters is as indicated below.
Part A:

An Overview

Chapter 1:

An Overview (Khairy Kobbacy and Pra Murthy)

Part B:

Evolution of Concepts and Approaches

Chapter 2:
Chapter 3:
Chapter 4:

Maintenance: An Evolutionary Perspective (Liliane Pintelon and
Alejandro Parodi Herz)
New Technologies for Maintenance (Jay Lee and Haixia Wang)
Reliability Centred Maintenance (Marvin Rausand and Jorn Vatn)

Part C:

Methods and Techniques

Chapter 5:
Chapter 6:
Chapter 7:
Chapter 8:
Chapter 9:

Condition-based Maintenance Modelling (Wenbin Wang)

Maintenance Based on Limited Data (David F. Percy)
Reliability Prediction and Accelerated Testing (Elsayed A. Elsayed)
Preventive Maintenance Models for Complex Systems
(David F. Percy)
Artificial Intelligence in Maintenance (Khairy A.H. Kobbacy)

Part D:

Problem Specific Models

Chapter10:
Chapter 11:

Chapter 14:

Maintenance of Repairable Systems (Bo Henry Lindqvist)
Optimal Maintenance of Multi-component Systems: A Review
(Robin P. Nicolai and Rommert Dekker)
Replacement of Capital Equipment (Philip A. Scarf and Joseph
C. Hartman)
Maintenance and Production: A Review of Planning Models
(Gabriella Budai, Rommert Dekker and Robin P. Nicolai)
Delay Time Modelling (Wenbin Wang)

Part E:

Management

Chapter 15:
Chapter 16:

Maintenance Outsourcing (Pra Murthy and Nat Jack)
Maintenance of Leased Equipment (Pra Murthy and Jarumon
Pongpech)
Computerised Maintenance Management Systems (Ashraf Labib)
Risk Analysis in Maintenance (Terje Aven)
Maintenance Performance Measurement (MPM) System
(Uday Kumar and Aditya Parida)
Forecasting for Inventory Management of Service Parts
(John E. Boylan and Aris A. Syntetos)

Chapter 12:
Chapter 13:

Chapter 17:
Chapter 18:
Chapter 19:
Chapter 20:


An Overview

11

Part F:

Applications (Case Studies)

Chapter 21:
Chapter 22:

Maintenance in the Rail Industry (Jorn Vatn)
Condition Monitoring of Diesel Engines
(Renyan Jiang and Xinping Yan)
Benchmarking of the Maintenance Process at Banverket
(The Swedish National Rail Administration)
(Ulla Espling and Uday Kumar)
Integrated e-Operations–e-Maintenance: Applications in North Sea
Offshore Assets (Jayanta P. Liyanage)
Fault Detection and Identification for Longwall Machinery Using
SCADA Data (Daniel Bongers and Hal Gurgenci)

Chapter 23:

Chapter 24:
Chapter 25:

A brief outline of each chapter is as follows
Chapter 2: Maintenance: An Evolutionary Perspective
In the past few decades industrial maintenance has evolved from a non-issue into a
strategic concern. During this period the role of maintenance has drastically been
transformed. This chapter, while considering the fundamental elements of maintenance and its environment, describes the evolution path of maintenance management and the driving forces of such changes. It basically explains how and why
maintenance practice has evolved in time. It includes basic notions of maintenance
and clearly classifies and distinguishes between different types of maintenance
actions, policies and concepts currently available. The chapter concludes by enlightening the reader with some new challenges in maintenance
Chapter 3: New Technologies for Maintenance
Predictive maintenance is critical to any engineering system, especially complex
systems, in order to avoid system breakdown. With the recent advances in pervasive
computing, prognostics can be easily embedded in any devices and systems. When
smart machines are networked and remotely monitored, and when their data is

modelled and continually analyzed with sophisticated embedded systems, it is
possible to go beyond mere “predictive maintenance” to intelligent “prognostics”, the
process of pinpointing exactly which components of a machine are likely to fail and
then autonomously trigger service and order spare parts. This chapter addresses the
paradigm shift in modern maintenance systems from the traditional “fail and fix”
practices to a “predict and prevent” methodology. Recent advances in prognostic
technologies and tools are presented, and future work directions are discussed.
Chapter 4: Reliability Centred Maintenance
This chapter gives an introduction to reliability centred maintenance (RCM). The
RCM analysis process is divided into 12 distinct steps. Each step is thoroughly
described and discussed. The main RCM process is similar to the processes
outlined in RCM standards and guidelines, but has more focus on the optimization
of maintenance intervals. A new approach is proposed based on generic RCM
analyses related to specified classes of consequences. The new approach will
significantly reduce the workload of the RCM analysis. A computer tool OptiRCM


12

K. Kobbacy and D. Murthy

that has been developed by the authors, is used to illustrate the new approach.
Several examples from railway applications are provided.
Chapter 5: Condition-based Maintenance Modelling
This chapter presents a model for supporting condition based maintenance decision
making. The chapter discusses various issues related to the subject, such as the
definition of the state of an asset, direct or indirect monitoring, relationship between observed measurements and the state of the asset, and current modelling
developments. In particular, the chapter focuses on a modelling technique used
recently in predicting the residual life via stochastic filtering. This is a key element
in modelling the decision making aspect of condition based maintenance. A few

key condition monitoring techniques are also introduced and discussed. Methods of
estimating model parameters are outlined and a numerical example based on real
data is presented.
Chapter 6: Maintenance-based on Limited Data
Reliability applications often suffer from paucity of data for making informed
maintenance decisions. This is particularly noticeable for high reliability systems
and when new production lines or new warranty schemes are planned. Such issues
are of great importance when selecting and fitting mathematical models to improve
the accuracy and utility of these decisions. This chapter investigates why reliability
data are so limited and proposes statistical methods for dealing with these
difficulties. It considers graphical and numerical summaries, appropriate methods
for model development and validation, and the powerful approach of subjective
Bayesian analysis for including expert knowledge about the application area.
Chapter 7: Reliability Prediction and Accelerated Testing
This chapter presents an overview of accelerated life testing (ALT) methods and
their use in reliability prediction at normal operating conditions. It describes the
most commonly used models and introduces new ones which are “distribution
free”. Design of optimum test plans in order to improve the accuracy of reliability
prediction is also presented and discussed. The chapter provides, for the first time,
the link between accelerated life testing and maintenance actions. It develops
procedures for using the ALT results for estimating the optimum preventive
maintenance schedule and the optimum degradation threshold level for degrading
systems. The procedures are demonstrated using two numerical examples.
Chapter 8: Preventive Maintenance Models for Complex Systems
Preventive maintenance (PM) of repairable systems can be very beneficial in
reducing repair and replacement costs, and in improving system availability.
Strategies for scheduling PM are often based on intuition and experience, though
considerable improvements in performance can be achieved by fitting mathematical models to observed data. For simple repairable systems comprising few components or many identical components, compound renewal processes are appropriate.
This chapter reviews basic and advanced models for complex repairable systems
and demonstrates their use for determining optimal PM intervals. Computational

An Overview

13

difficulties are addressed and practical illustrations are presented, based on subsystems of oil platforms and
Chapter 9: Artificial Intelligence in Maintenance
AI techniques have been used successfully in the past two decades to model and
optimise maintenance problems. This chapter reviews the application of Artificial
Intelligence (AI) in maintenance management and introduces the concept of
developing intelligent maintenance optimisation system. The chapter starts with an
introduction to maintence management, planning and scheduling and a brief
definition of AI and some of its techniques that have applications in maintenance
management. A review of literatures is then presented covering the applications of
AI in maintenance. We have focused on five AI techniques namely Knowledge
Based Systems, Case Based Reasoning, Genetic Algorithms, Neural Networks and
Fuzzy Logic. This review also covers “hybrid” systems where two or more AI
techniques are used in an application. A discussion of the development of the
prototype hybrid intelligent maintenance optimisation system (HIMOS) which was
developed to evaluate and enhance PM maintenance routines of complex engineering systems then follows. The chapter ends with a discussion of future
research and concluding remarks.
Chapter 10: Maintenance of Repairable Systems
A repairable system is traditionally defined as a system which, after failing to perform one or more of its functions satisfactorily, can be restored to fully satisfactory
performance by any method other than replacement of the entire system. An
extended definition used in this chapter includes the possibility of additional maintenance actions which aim at servicing the system for better performance, referred to
as preventive maintenance (PM). The common models for the failure process of a
repairable system are renewal processes (RP) and non-homogeneous Poisson processes (NHPP). The chapter considers several generalizations and extensions of the
basic models, for example the trend renewal process (TRP) which includes NHPP
and RP as special cases, and having the property of allowing a trend in processes of

non-Poisson type. When several systems of the same kind are considered, there may
be an unobserved heterogeneity between the systems which, if overlooked, may lead
to wrong decisions. This phenomenon is considered in the framework of the TRP
process. We then consider the extension of the basic models obtained by introducing
the possibility of PM using a competing risks approach. Finally, models for periodically inspected systems are studied, using a combination of time-continuous and
time-discrete Markov chains.
Chapter 11: Optimal Maintenance of Multi-component Systems: A Review
This chapter gives an overview of the literature on multi-component maintenance
optimization focusing on work appearing since the 1991 survey by Cho and Parlar.
A classification scheme primarily based on the dependence between components
(stochastic, structural or economic) is introduced. Next, the papers are also classified on the basis of the planning aspect (short-term vs. long-term), the grouping of
maintenance activities (either grouping preventive or corrective maintenance, or
opportunistic grouping) and the optimization approach used (heuristic, policy


14

K. Kobbacy and D. Murthy

classes or exact algorithms). Finally, attention is paid to the applications of the
models.
Chapter 12: Replacement of Capital Equipment
This chapter deals with models of replacement of capital equipment. Capital replacement models may be classified as economic life models or dynamic programming
models. The former are concerned with determining the optimal lifetime of an item
of equipment taking account of costs over some planning horizon. The latter considers replacement decisions dynamically, determining whether plant should be
retained or replaced after each period. We begin by looking at simple economic life
models. These are applied in a case study on escalator replacement. Economic life
models are then extended to consider first an inhomogeneous fleet and then second a
network system viewed as an inhomogeneous fleet with interacting items. A number
of different dynamic programming models are introduced for singular systems and

then expanded to homogeneous and inhomogeneous fleets and networks of assets.
Chapter 13: Maintenance and Production: A Review of Planning Models
This chapter gives an overview of the relation between planning of maintenance
and production. Production planning and scheduling models where failures and
maintenance aspects are taken into account are considered first. The planning of
maintenance activities are considered next, where both preventive as well as
corrective maintenance are discussed. Third, the planning of maintenance activities
at such moments in time where the items to be maintained are not or less needed
for production, also called opportunity maintenance is considered. Apart from
describing the main ideas, approaches, and results a number of applications are
provided.
Chapter 14: Delay Time Modelling
This chapter presented a modelling tool that was created to model the problems of
inspection maintenance and planned maintenance interventions, namely Delay
Time Modelling (DTM). This concept provides a modelling framework readily
applicable to a wide class of actual industrial maintenance problems of assets in
general and inspection problems in particular. The delay time defines the failure
process of an asset as a two-stage process. The first stage is the normal operating
stage from new to the point that a hidden defect has been identified. The second
stage is defined as the failure delay time from the point of defect identification to
failure. It is the existence of such a failure delay time which provides the opportunity for preventive maintenance to be carried out to remove or rectify the
identified defects before failures. With appropriate modelling of the durations of
these two stages, optimal inspection intervals can be identified to optimise a
criterion function of interest. This chapter first gives an outline of the delay time
concept then introduces two delay time inspection models of a single component
and a complex system respectively. The parameters estimation techniques used in
DTM are discussed. Extensions to the basic delay time model are highlighted and
future research in DTM concludes the chapter.

An Overview

15

Chapter 15: Maintenance Outsourcing
It is often uneconomical for businesses to carry out their own maintenance on
complex equipment. The alternative is to ‘out-source’ the maintenance function
and use an external agent, under a service contract, to carry out some or all of the
maintenance actions (preventive and corrective). This chapter develops the framework needed to study decision-making for maintenance outsourcing from both the
customer (equipment owner) and service agent perspectives. The relevant literature
is reviewed and a game theoretic approach to maintenance outsourcing and the use
of agency theory is discussed. The link between maintenance outsourcing and
extended warranties is highlighted and the scope for future research in both areas is
examined.
Chapter 16: Maintenance of Leased Equipment
For leased equipment, the lessor has to carry out the maintenance of the equipment
over the lease period. To ensure satisfactory performance and maintenance, the
lease contract has penalty terms which result in the lessor having to compensate the
lessee if the number of failures exceeds some specified number and/or the time to
rectify each failure exceeds some specified value. This implies that the lessor needs
to take into account these penalties in determining the optimal maintenance
strategy. The chapter starts with a conceptual framework to discuss the different
issues involved and then looks at models to help the lessor in developing the
optimal maintenance strategy.
Chapter 17: Computerised Maintenance Management Systems
Computerised maintenance management systems (CMMSs) are vital for the coordination of all activities related to the availability, productivity and maintainability
of complex systems. Modern computational facilities have offered a dramatic scope
for improved effectiveness and efficiency in, for example, maintenance. CMMSs
have existed, in one form or another, for several decades. In this chapter, the
characteristics of CMMSs have been investigated and have highlighted the need for

them in industry and identified their current deficiencies.
A proposed model is then presented to provide a decision analysis capability
that is often missing in existing CMMSs. The effect of such model is to contribute
towards the optimisation of the functionality and scope of CMMSs for enhanced
decision analysis support. The use of AI techniques in CMMSs is illustrated. The
features of next generation maintenance systems are finally highlighted.
Chapter 18: Risk Analysis in Maintenance
Risk analysis can be used for selection and prioritisation of maintenance activities,
and this application of risk analysis has been given increased attention in recent
years. This chapter presents and discusses the use of risk analysis for this purpose.
The chapter reviews some critical aspects of risk analysis important for the
successful implementation of such analyses in maintenance. This relates to risk
descriptions and categorisations, uncertainty assessments, risk acceptance and risk
informed decision making, as well as selection of appropriate methods and tools.
Both qualitative and quantitative approaches are covered. A detailed risk analysis
is outlined showing the effect of maintenance on risk.


16

K. Kobbacy and D. Murthy

Chapter 19: Maintenance Performance Measurement (MPM) System
It is important that factors influencing the performance of maintenance process
should be identified, and measured, so that they can be monitored and controlled for
improvement. In this chapter, besides an overview of performance measurement,
maintenance performance indicators, associated issues and challenges for developing
a maintenance performance measurement framework, and indicators as in use by
different industries are discussed. The framework considers stakeholders, business
environment, multi-criteria and hierarchical needs amongst other.

Chapter 20: Forecasting for Inventory Management of Service Parts
This chapter addresses issues pertinent to forecasting for the inventory management
of service parts. In some sectors, such as the aerospace and automotive industries, a
very wide range of service parts are held in stock, with significant implications for
availability and inventory holding. Their management is therefore an important task.
First, a number of possible approaches to classifying service parts for forecasting
and inventory management related purposes are reviewed. Second, parametric and
non-parametric approaches to forecasting service parts requirements are discussed
followed by the presentation of appropriate metrics for measuring the performance
of the inventory management system. The existing empirical evidence on various
forecasting methods is then summarised. Finally, the conclusions of this work are
presented along with the identification of some natural avenues for further research.
Chapter 21: Maintenance in the Rail Industry
The chapter presents two case studies in railway maintenance. The first case study
presents an optimisation model preventive maintenance of a train bogie. In the model
a dynamic approach to grouping of maintenance activities is used enabling, e.g.,
opportunity maintenance. Data from the Norwegian State Railways have been used
in the calculation example. The second case study present a life cycle cost approach
to prioritization of larger maintenance and renewal projects under budget constraints.
Chapter 22: Condition Monitoring of Diesel Engines
Various techniques have been widely used to monitor the condition of diesel
engines. Analysis of engine lubricant is a most widely used condition monitoring
technique. In this chapter, a case study applying oil analysis technique to monitor
the condition of marine diesel engines is presented. The case study focuses on
analysis and modelling of oil monitoring data. The study first introduces the concept of state discriminant capability of condition variables and uses it to identify
the significant condition variables, and then develops a state discriminant model to
determine the state of the monitored system based on the current observation. The
model parameters are obtained by directly minimizing the misjudgment probability. We believe that the proposed model has a great potential to be used due to its
plausible mathematical basis and simplicity though it needs further testing with
new data.