FEATURESu Smallest Size, Lightest weight Camera

on the market today

u Meets IP-54 Environmental

Specifications.

u 10-Point Temperature Display and

Emissivity Correction

u Spatial Filter for improved Image

Quality

u Allows Voice and Visible Recording

along with IR Image Capture

u IEEE1394 Interface BUS for Real-Time

Image Transfer to PC

u Variable Image Frame Rates

u Direct Video Output

u 14 Bit A/D Resolution

APPLICATIONSu Electrical Switchboards, Switchgear,

Transmission Lines and Transformers.

u Motors and Generators.

u Ovens, Furnaces, Boilers, Kilns,

Moving Machinery, Extruders, Dies

and Moulding Equipment

u Circuit Boards, Components.

u Refrigeration and Air Conditioning

Systems.

u Research and Development

u Process Control Applications

u Medical Research and Diagnosis.

u Optional Items include a full range of

lens configurations and Report

Generation and Analysis Software.

NEC Model TH7102WXVUncooled IR FPA Predictive Maintenance System

Infratherm Pty Ltd PO Box 117, NORTH RICHMOND, NSW 2756 AUSTRALIATel: +61 2 4579 7334 or +61 2 4322 2100 Facsimile: +61 2 4579 6333 or +61 2 4323 7439

E-mail: jrobinson@infratherm.com.au - mratne@infratherm.com.au.

web: www.infratherm.com.au

[Advertiser]Full Page Advertisement

3

The Entek product family - completeenterprise health is now a reality!Leading automation technology company,Rockwell Automation, has developed theEntek product family into a compre h e n s i v esuite of condition-based monitoring solutionsand services offering optimal scalability andf l e x i b i l i t y. From hand-held vibration sensorsto surveillance systems and plant-wide on-line protection networks, RockwellAutomation has you covered - now and inthe future !

On-line continuous machinemonitoring and protection: Solutionsranging from the configurable, singlepoint Sentinel vibration monitor through tothe fully scalable, modular Entek XMdistributed protection system

On-line surveillance: Bridging the gapbetween portable data collectors andcontinuous monitoring, are the Enwatch on-line surveillance systems, a powerful yetextremely cost effective surveillancesolution.

Diagnosis and analysis tools: A widerange of on- and off-line data analysis anddiagnostic solutions including the EmonitorOdyssey V3.0 complete machineryinformation software system, and theEmonitor Enshare advanced Plant AssetManagement [PAM] system

Data collectors and meters: A wideselection of condition monitoring datacollectors and meters, from the low-costVistec portable data vibration meter throughto the fully featured DataPAC 1500 andEnpac portable data collectors/analysers.

Industrial sensors: A complete suite ofindustrial sensors including accelerometers,laser tachometers, and non-contactproximity probes for a range of industries.

Condition monitoring audits andanalysis services: Our experiencedtechnical team are avaliable to provide a complete range of on-site or remoteservices.

For more information, visit us at:www.entek.com orwww.rockwellautomation.com.auor contact us on:Tel: +61 [0]3 9896 0300 AustraliaTel: +64 [0]9 276 3070 New Zealand

A journal for all those interested in themaintenance, monitoring, servicing andmanagement of plant, equipment,buildings and facilities.

Volume 17, No 1.February 2004

Published by:Engineering Information Transfer Pty Ltd

Publisher and Managing Editor:Len Bradshaw

Publishing Dates:Published in February, May, August andOctober.

Material Submitted:Engineering Information Transfer Pty Ltdaccept no responsibility for statementsmade or opinions expressed in articles,features, submitted advertising,advertising inserts and any other editorialcontributions.

Copyright:This publication is copyright. No part ofit may be reproduced, stored in aretrieval system or transmitted in anyform by any means, including electronic,mechanical, photocopying, recording orotherwise, without the prior writtenpermission of the publisher.

For all Enquiries Contact:Engineering Information Transfer Pty LtdPO Box 703, Mornington, Victoria 3931, AustraliaPhone: (03) 5975 0083, Fax: (03) 5975 5735,E-mail: mail@maintenancejournal.comWeb Site: www.maintenancejournal.com

CMMS - A Black Hole or a Black BoxAsgraf W. Labib

6

What Do You Need To Know About Software MaintenanceAlain April, A. Abran and R Dumke

10

Winning The Battle With Downtime Using RBITony Musgrave

14

Revolutionising Naval Maintenance With RCMDr Alun Roberts

18

Terotechnology In The Competitive EdgeDr. Bimal Samanta, Dr. Bijan Sarkar

42

Enterprise Asset Management Benchmark SurveyHow do you measure your maintenance performance

50

2004 Survey Of Maintenance Data Collection SystemsPrepared by Ian Bradshaw

69

A TPM3 Journey To Excellence - Case StudyKeith Saul

24

The Total EAM VisionDaryl Mather

32

Maintenance Improvement - Where To Start

Allan Hutton36

Maintenance NewsCurrent Maintenance andProduct News

72

Forthcoming EventsSeminars and Conferences

78

Positions VacantMaintenance Job Vacancies

79

Subscription FormSubscribe to either thePrint or eMJ versions ofThe Maintenance Journal

80

Regular Features

February 2004Contents

Something Loose Inside Your Machine?The February 2004 Cover Shot shows aRecovery Tool attachment which maybe used with Olympus videoscopes orfibrescobes. For more information see:www.olympusindustrial.com

RCMeasy... Workshop and Software$1995 +gst

A professional two-day how to workshopon the RCM process, supported byuncomplicated reliability software

(THAT YOU KEEP!)

The R C M e a s y workshop concentrates on a pro v e nsimpler approach, while acknowledging the

fundamentals of the traditionally exhaustive RCMp rocess and is supported by MCPs R C M e a s y s o f t w a re .

Places are limited, early registration is recommended.Ask us about our Early Bird Discounts.

Sydney, March 16th & 17th 2004Visit www.mcp-cg.com/RCMeasy for details

Maintenance Management Resourc e sMCP knows that most maintenance management teamsare well qualified and experienced, we also know that

they need flexible and experienced resources from timeto time - this is where MCP can help.

MCP offers contract maintenance management andengineering personnel of the highest standard. All MCPs

s t a ff are equipped with the latest intellectual pro p e rt y,s o f t w a re tools, methodologies and pro c e d u res. Our staff

roll up their sleeves, become part of your team and developreal world practical solutions that are not tied to

self serving and single solution techniques.

M C P s team has years of experience and knowledge in arange of specialised industries. Recent and current clientsinclude; Santos, QNI, Queensland Rail, Sunwater, Sydney

Catchment Authority and many more .

Visit www. m c p - c g . c o m / re s o u rce for detailsComplimentary,one user license,RCMeasy Software,for each delegateattending theworkshop.

REGISTER TODAY...Ph: 1300 368 820Email: enquiries@mcp-cg.com

NEW FOR 2004

www.mcp-cg.com

Editorial

Welcome to the first edition of the Maintenance Journal for 2004. This editionof the Maintenance Journal has a broad range of topics that includes articles onRCM, TPM, RBI, EAM, CMMS, ERP and to conclude our line up of acronyms whatabout SMCMM? To learn more read the articles.

The major item in this issue is an Enterprise Asset Management Benchmarksurvey that looks at maintenance of plant, equipment, facilities and assets. Thesurvey confirms that a fix it when it is broken attitude is still prevalent amongmany organisations, costing millions of dollars annually. The org a n i s a t i o n ss u rveyed acknowledge that preventative maintenance adds true economic valueto the bottom-line, yet they still are not investing in an integrated, pre v e n t a t i v emaintenance strategy to maximize this pro f i t a b i l i t y. Just how dramaticallyMaintenance can impact profitability is illustrated by re p o rts in the Australianp ress re g a rding a major mineral processing plant. Recent failures at the planthave caused a 4% slump in the company share price. A failure lasting weeks inNovember 2003 wiped $1.5 million a day off the companys pre-tax profit and alsorequired $3 million to repair.

For the first time in the MJ we have included an article on the Maintenance ofS o f t w a re. It is interesting that much of the maintenance terminology used is veryfamiliar. However one new term used in the article, which has a nice sound to it,is Perfective maintenance!

Those of you using CMMS on a regular basis will appreciate what Ashraf Labibhas to say on the Black Hole problem associated with many org a n i s a t i o n s usage of computerised maintenance management systems

Special Feature

in the May2004 issue

MJ SurveyComputerisedMaintenanceManagement

SystemsIf your organisation wishes

to be included in thesurvey, then you may obtainthe appropriate survey form

by contacting Ian Bradshaw at:

m a i l @ m a i n t e n a n c e j o u r n a l . c o m

Completed survey forms mustbe returned by: 20 Feb 2004

February

If yo u re a typical facilitiesm a n a g e r, you spend 95% of yo u r

time doing anything but!Request this information pack and discover how BEIMS can automate your day-to-day operations-

leaving you free to get on with the real business of facility management

Going Beyond standard asset managements o f t w a re, it is the only software that fullyintegrates the facility management process.

For this reason, it has been adopted by morethan 130 organisations, including theUniversity of Melbourne, the Victorian Art sC e n t re, Darling Harbour Authority and theCrown Casino.

Free information packFor a comprehensive information packageabout Beims, telephone Merc u ry ComputerSystems on 03 9602 2255 or e-mail yourdetails to sales@beims.com

This package profiles the many benefitso ff e red by BEIMS and explains how it canhelp you to minimise your costs and impro v ethe efficiency of your facility.

5:55pm: light bulb blows in reception.

5:58pm: maintenance persons mobilephone receives SMS notification.

6:15pm: light bulb changed. job logged,costed and closed

The scenario above demonstrates howBEIMS can automate the day-to-dayoperations that, if left unchecked, can easilyswamp you - leaving you free to get on withthe real business of facility management.

A fully automated facilityControl centre

BEIMS provides you with a control centrethat automates the management ofp reventive maintenance, as well as adhocb reakdowns. Furt h e rm o re, it allows you toautomatically track and manage yourCompanys assets.

Preventive maintenance schedules can beplanned 12 or more months in advance and

adhoc breakdowns or requests logged dire c t l yinto the system via intranet or the intern e t .Work orders are then automatically createdand dispatched at the appropriate time - withcosts and labour charges tracked andrecorded.

Over 200 management reports atthe press of a button

With more then 200 management reports atyour fingertips, BEIMS will allow you tomanage your facility with absolute pre c i s i o n .

This will allow you to improve eff i c i e n c y,m o re accurately budget for futuremaintenance and capital expenditure, andminimise your short - and long-termoperating costs.

The only purpose-built facilitymanagement software

Unlike generic maintenance software ,BEIMS is purpose-built for facility managers.

AbstractIn this paper an investigation of the characteristics of

Computerised Maintenance Management Systems (CMMSs) isc a rried out. The objectives are to highlight the need for them ini n d u s t ry and to identify their current deficiencies. Evidences of theexistence of a term coined as "black holes" are presented.

Need for information to aid maintenancemanagement

As in almost every sphere of organisational activity, moderncomputational facilities have off e red dramatic scope for impro v e de ffectiveness and eff i c i e n c y. Maintenance is one area in whichcomputing has been applied, and Computerised MaintenanceManagement Systems (CMMSs) have existed, in one form or another,for several decades. The software has evolved from relatively simplemainframe planning of maintenance activity to Windows-based, multi-user systems that cover a multitude of maintenance functions. Thecapacity of CMMSs to handle vast quantities of data purposefully andrapidly has opened up new opportunities for maintenance, facilitatinga more deliberate and considered approach to managing anorganisation's assets.

CMMS is now a central component of many companies'maintenance departments, and it offers support on a variety of levelsin the organisational hierarchy:

It can support CBM (Condition Based Monitoring) of machinesand assets, to offer insight into wear and imminent failures;

It can track the movement of spare parts and requisitionreplacements when necessary;

It allows operators to report faults faster, thus enablingmaintenance staff to respond to problems more quickly;

It can facilitate improvement in the communication betweenoperations and maintenance personnel, and is influential inameliorating the consistency of information passed betweenthese two departments;

It provides maintenance planners with historical informationnecessary for developing PM schedules;

It provides maintenance managers with information in a formthat allows for more effective control of their department'sactivities;

It offers accountants information on machines to enable capitalexpenditure decisions to be taken;

It affords senior management a crucial insight into the state ofasset healthcare within their organisation.Indeed, Labib (1998) comments that ideally a CMMS is a means to

achieving world-class maintenance, by offering a platform for decisionanalysis and thereby acting as a guide to management. CMMSpackages are able to provide management with re p o rts and statistics,detailing perf o rmance in key areas and highlighting pro b l e m a t i c

8

Department of Mechanical, Aerospace and Manufacturing Engineering, University of Manchester Institute of Science and Technology (UMIST), UKemail: ashraf.labib@umist.ac.uk

Ashraf W. Labib

ComputerisedMaintenanceManagement SystemsA black hole or a black box?

Department of Mechanical, Aerospace and Manufacturing Engineering, University of Manchester Institute of Science and Technology (UMIST), UKemail: ashraf.labib@umist.ac.uk

Ashraf W. Labib

CMMS - a black hole or a black box

issues. Maintenance activities are consequently more visible andopen to scru t i n y. Managers can rapidly discover which policies work,which machines are causing problems, where overspend is takingplace, and so on, thereby revealing information that can be used asthe basis for the systematic management of maintenance. Thus, bytracking asset 'health' in an organised and systematic manner,maintenance management can start to see how to improve the curre n tstate of affairs. However, the majority of CMMSs in the market sufferfrom serious drawbacks as will be shown in the following section.

Current deficiencies in existing on-the-shelfCMMSs

Most existing on-the-shelf software packages, especiallyComputerised Maintenance Management Systems (CMMS) andEnterprise Resource Planning (ERP) systems, tend to be 'black holes'.This term is coined by the author as an observation of data inputgreedy systems that seldom provide any output in terms of decisions u p p o rt. Companies consume a significant amount of managementand superv i s o ry time compiling, interpreting and analysing the datac a p t u red within the CMMS. Companies then encounter diff i c u l t i e sanalysing equipment perf o rmance trends and their causes as a re s u l tof inconsistency in the form of the data captured and the historicalnature of certain elements of it. In short, companies tend to spend avast amount of capital in acquisition of of-the-shelf systems for datacollection and their added value to the business is questionable.

9

... ... ... ...

Data Collection 3 3 3 3Data Analysis 3 3 3Real Time 3 3Network 3Decision AnalysisPrice Range 1k + 10k + 30k + 40k +

Figure 1: Facilities offered by commercially available CMMS packages

A BLACK HOLE

Maintenance Budgeting

Predictive Maintenance data analysis

Equipment failure diagnosis

Inventory control

Spare parts requirements planning

Material and spare parts purchasing

Manpower planning and scheduling

Work-order planning and scheduling

Equipment parts list

Equipment repair history

Preventive Maintenance planning and scheduling

70 75 80 85 90 95 100

A Black Hole

Figure 2: Extent of CMMS module usage (from Swanson, 1997)

Applications of CMMS Modules

Percentage of systems incorporating module

All CMMS systems offer data collection facilities; more expensivesystems offer a formalised module(s) for the analysis of maintenancedata; the market leaders allow real time data logging and networkeddata sharing [see Figure 1]. Yet, despite the observations made aboveregarding the need for information to aid maintenance management,v i rtually all the commercially available CMMS software lacks anydecision analysis support for management. Hence, as indicated in Thislack of decision analysis is a pronounced problem, as the key tosystematic and effective maintenance is managerial decisions beingappropriate to the particular circumstances of the machine, plant oro rganisation. This decision making process is made all the moredifficult if the CMMS package can only offer an analysis of recordeddata. As an example when one inputs a certain Pre v e n t i v eMaintenance (PM) schedule to a CMMS, say to change the oil filterevery month, the system will simply produce a monthly instruction tochange the oil filter. In other words it is no more than a diary. A steptowards decision support is to vary frequency of PMs depending onthe combination of failure frequency and severity. A more intelligentf e a t u re would be to generate and to prioritize PMs according to modesof failure in a dynamic real-time environment. PMs are usually staticand theoretical in the sense that they do not reflect shop floor re a l i t i e s .In addition, the PMs that are copied from machine manuals are notusually applicable because; a . each machine works in a diff e re n te n v i ronment and would there f o re need diff e rent PMs, b . m a c h i n e sdesigners often do not have the same experience of machines failure s ,and means of prevention, as those who operate and maintain them,and c . machine vendors may have a hidden agenda of maximizingspare parts replacements through frequent PMs.

A noticeable problem with current CMMS packages re g a rd sp rovision of decision support. Figure (2) illustrates how the use ofCMMS for decision support lags significantly behind the moretraditional applications of data acquisition, scheduling and work-ord e rissuing. While many packages now offer inventory tracking, and somef o rm of stock level monitoring, the re o rdering and inventory holdingpolicies remain relatively simplistic and inefficient. See the work ofExton and Labib (2002) and Labib and Exton (2001). Moreover, there isno mechanism to support managerial decision-making with regard toi n v e n t o ry policy, diagnostics or setting of adaptive and appro p r i a t epreventive maintenance schedules.

CMMS - a black hole or a black box

10

A c c o rding to Boznos (1998) "The primary uses of CMMS appear to be as a storehouse for equipment information, as well as a plannedmaintenance and a work maintenance planning tools". The same author suggests that CMMSs appears to be used less often as a device foranalysis and co-ordination and that "existing CMMS in manufacturing plants are still far from being regarded as successful in providing teambased functions". The author has surveyed CMMSs and investigated TPM and RCM concepts and to what extent both concepts are embeddedin existing CMMSs in the market. He has concluded that "it is worrying the fact that almost half of the companies are either in some degre edissatisfied or neutral with their CMMSs and that the responses indicated that manufacturing plants demand more user-friendly systems"(Boznos,1998). This is a further proof of the existence of a 'black-hole'.

In addition, to make matters worse, it appears that there is a new breed of CMMSs that are complicated and lack basic aspects of user-friendliness. Although they emphasise integration and logistics capabilities, they tend to ignore the fundamental reason for implementing CMMSsthat is reducing breakdowns. These systems are difficult to handle by either production operators or maintenance engineers. They are moreaccounting and/or IT oriented rather than engineering based. In short they are Systems Against People that further promote the concept ofblack holes.

Several factors are deriving the need for a new paradigm to aid maintenance management through an intelligent maintenance informationsystem. Firstly, the amount of information available, even to quite modest organisations, continues to increase almost exponentially. What ismore, there is an increasing requirement to have this data and information on hand and in real-time for decision-making. Secondly, data-life-time is diminishing as a result of the shop-floor realities, which are real-time in nature, and the rapid pace of change. The initiative now is toacquire data about individual machines, based upon real interactions rather than deduced behaviour from historical data. Finally, the way thatdata is being accessed has changed. The days of legacy maintenance systems of large batch re p o rts, where the focus was on data thro u g h p u t ,are being replaced by dynamic, on-line queries, created on-the-fly, and with answers in seconds rather than days.

Results of an investigation of the existing reliability models and maintenance systems (EPSRC Grant No. GR/M35291) show that managerslack of commitment to maintenance models has been attributed to a number of reasons (Shorrocks and Labib, 2000), and (Shorrocks, 2000):i . Managers are unaware of the various types of maintenance models, ii. A full understanding of the various models and the appro p r i a t e n e s sof these systems to companies are not available and, i i i . Managers do not have confidence in mathematical models due to their complexitiesand the number of unrealistic assumptions they contain. This correlates with recent surveys of existing maintenance models and optimisationtechniques, Ben-Daya etal (2001) and Sherwin (2000) have also noticed that models presented in their work have not been widely used ini n d u s t ry for several reasons such as: i . Unavailability of data, ii. Lack of awareness about these models, and iii. Some of these modelshave restrictive assumptions. Hence, theory and implementation of existing maintenance models are to a large extent disconnected. Theyconcluded that there is a need to bridge the gap between theory and practice through intelligent optimisation systems (e.g. ru l e - b a s e dsystems). They argue that the success of this type of re s e a rch should be measured by its relevance to practical situations and by its impacton the solution of real maintenance problems. The developed theory must be made accessible to practitioners through information technologytools. Eff o rts need to be made in the data capturing area to provide necessary data for such models. Obtaining useful reliability inform a t i o nf rom collected maintenance data re q u i res eff o rt. In the past, this has been re f e rred to as data "mining," as if data can be extracted in itsd e s i red form if only it can be found.

Conclusion

The main idea is based on the fact that the 'black hole' or missing functionality in conventional CMMSs is intelligent decision analysis tools.In PART 2 we demonstrate how maintenance system can be transferred from being a black hole to a black box where the input to that box aredata and the outputs produced are decisions. We will present an industrial application of holonic concepts in manufacturing maintenance. Theproposed model provides combination features of both fixed rules and flexible strategies. PART 2 will be published in the May 2004 issue of theMaintenance Journal

References

[1]. Ben-Daya, M., S.O. Duffuaa, A. Raouf (eds), (2001) Maintenancemodelling and Optimisation, Kluwer Academic Publishers.

[2]. Boznos, Dimitrios, "The Use of CMMSs to Support Team-Based Maintenance", MPhil Thesis, Cranfield University, 1998.[3]. Exton, T. and Labib, A.W., Spare Parts Decision Analysis - The Missing Link in CMMSs (Part II), Journal of Maintenance & Asset

Management, ISSN 0952-2110, Vol 17 No1, 2002.[4]. Labib, A.W., and T. Exton, Spare Parts Decision Analysis - The Missing Link in CMMSs (Part I), Journal of Maintenance & Asset

Management, ISSN 0952-2110, Vol 16 No 3, pp 10-17, 2001.[5]. Labib, A.W., "World Class Maintenance Using a Computerised Maintenance Management System"; Journal of Quality in Maintenance

Engineering (JQME); MCB Press; Vol 4, No 1.; pp 66-75; April 1998.[6]. Sherwin, D. (2000) A review of overall models for maintenance management, Journal of Quality in Maintenance Engineering, Vol. 6 No.

3.[7]. Shorrocks, P., and A.W. Labib, "Towards A Multimedia-based Decision Support System for Word Class Maintenance", Proceedings of the

14th ARTS (Advances in Reliability Technology Symposium), IMechE, University of Manchester, November, 2000.[8]. Shorrocks, P, "Selection of the most appropriate maintenance model using a decision support framework", unpublished report UMIST,

2000.[9]. Swanson, L. (1997) Computerized Maintenance Management Systems: A study of system design and use, Production and Inventory

Management Journal, Second Quarter pp. 11-14.

CMMS - a black hole or a black box

S o f t w a re accounts now for a increasing share of the content ofm o d e rn equipment and tools, and must similarly be maintained toe n s u re its continuous operational eff i c i e n c y. Although themaintenance of equipment is discussed extensively, very little ispublished about software maintenance and how it affects us. Thispaper presents an overview of key topics of software engineeringmaintenance.

Has your production ever been stoppedbecause of a software problem?

S o f t w a re maintenance is indeed re q u i red to support many keyequipment and product lines throughout their daily operational cycles.For instance, software - related problems and modification re q u e s t sa re sent to the supplier of the product where it is logged and tracked,the impact of proposed changes is determined, software code ismodified, testing is conducted, and a new version of the softwareproduct is released. This looks quite simple, but if it were that simple,the software fix would be applied in minutes. Then why does it oftentake days, weeks and sometimes months to dot it? Because softwareis sometimes quite complex and that even a seemingly very minorchange to an element might have a very extensive impact thro u g h o u tthe whole stru c t u re, should such element be used across the stru c t u reof either the software itself or throughout the operation system linkedto such software.

S o f t w a re maintenance is often perceived merely as fixing bugs,which is reactive to errors and omissions. However, studies ands u rveys over the years have indicated that there is much more tos o f t w a re maintenance than merely fixing bug. Some studies have evenreported that corrective software maintenance represents less than20% of the maintenance workload. The consensus on the keycomponents of the software maintenance process has beendocumented in the ISO/IEC 14764 the International Standard for

S o f t w a re Maintenance. This ISO standard, although as well knownas ISO900: 2000, is important to the software maintainers and togeneral management for understanding better the services providedon the software you own and that you are about to service or modify.

The ISO/IEC standard recognizes four categories of maintenancework (see figure 1):

Corrective maintenance: Reactive modification to a softwareproduct, performed after delivery to correct the problemsidentified.

Adaptive maintenance: Modification to a software product,performed after delivery to keep a software product usable in achanged or changing environment.

Perfective maintenance: Modification to a software product,after delivery to improve performance or maintainability.

Preventive maintenance: Modification to a software product,after delivery to detect and correct latent faults in the softwareproduct before they become effective faults.

In well-managed software organizations, most of the changes tos o f t w a re are carried out to adapt such software to the changingbusiness or technology environment. Because the business contextnow moves very quickly the equipments must also improve constantly,new and better functionality need to be inserted in the existingsoftware. The software does not deteriorate physically with time anddoes not age when it operates. However, due to continuous additionsor modifications, it gets more complex and patched with the numero u schanges and progressively becomes more difficult to maintain.

These are some of the issues that have to be recognized andunderstood for maintaining management control over the maintenancebudget of both the software and of the related equipment. So nexttime you look into the list of software problems try to separate themand identify the real % of corrections. Youll see that it does notaccount to so much. Its the changes that take the most out of yourbudgets. And changes, by definition, have more to do withmaintenance projects than routine maintenance. It becomes evidentthat some of those requests will take longer because they are notroutine work!

12

What Do You Need ToKnow About SoftwareMaintenanceEcole de Technologie Superieure, Montreal, Canada,aapril@ele.etsmtl.ca , aabran@ele.etsmtl.ca

Alain April, A. Abran

Otto von Guericke University of Magdeburg, Germany,dumke@ivs.cs.uni-magdeburg.de

Reiner R. Dumke

Correction EnhancementProactive Preventive PerfectiveReactive Corrective Adaptive

Figure 1: ISO14764 software maintenance categories

What Do You Need To Know About Software Maintenance

But why does it take so long to change thesoftware?

A number of very complex issues must be dealt with to ensureadequate maintenance of software systems. For example, it is mostchallenging for software maintainers to analyze 500,000 or 2,000,000lines of code software system that the maintainer did not develophimself to find a hidden defect or to identify where a specific changemust be implemented. Furthermore, software engineering is far froma mature engineering discipline and unlike mechanical engineering,still too little is provided in terms of professional and accre d i t e dtraining to the software maintainers. Instead it is often observed thats o f t w a re maintainers have a limited understanding of the pro d u c t sthey must maintain. Over the years, both practitioners and re s e a rc h e r shave reported that up to 60% of time spent on maintenance is indeeddevoted to developing a good understanding of the software to bemodified, prior to initiating any change to it. This of course leavesmuch less time to carry out the change and to test it extensively.

If its not the maintainers fault whose is it?Why is software so hard to maintain? Many of such difficulties can

be traced back to the software development process itself which oftendoes not take into account the maintainability re q u i rements: too oftens o f t w a re is developed in uncontrolled environments (e.g. re a d'hackers style') and not to professional engineers standards. Inindustry, if a product or an equipment is not built to the proper qualitys t a n d a rd, and without adequate maintenance documentation, thenmaintenance will be abnormally high when compared to products orequipments developed to the highest quality standards.

Because software is often embedded (that is, hidden) into industrialp roducts it lacks visibility, suffers from lack of management attention,then from lack of re s o u rces, which often leads to lower quality: forinstance, when a trade-off must be made in a situation of schedulec o m p ression, then software is often where development cuts happenrather than on the more visible hard w a re related components. Ofcourse, software vendors are part of the problem; in making youbelieve that their software is better, faster and maintainable (but withlittle supporting evidence).

The software maintainability issue is often quoted in theI n f o rmation Technology industry. The IEEE Computer Society[IEEE610.12] defines "maintainability" as the ease with which softwarecan be maintained, enhanced, adapted, or corrected to satisfyspecified requirements. ISO/IEC defines maintainability as one of themain quality characteristics of a software [ISO9126].

Maintainability of software must be specified, reviewed andc o n t rolled during the software development activities if we wish toever properly manage the maintenance process and subsequentlyreduce the maintenance costs. If this is done successfully, the qualityof maintenance of the software (its maintainability) will likely improve.

U n f o rt u n a t e l y, there is often a lack of attention to maintainabilityduring the software development process. Often software disre g a rd sthis key business and engineering requirement. Time and time agains o f t w a re is implemented and sent to the operations without adequatemaintenance documentation, unduly adding later on considerablemaintenance cost wherever a software change must be implemented.

Arent there best practices for improvingsoftware maintenance?

The need and benefits of mature engineering processes is welldocumented, including for software development. Similarly, there isa well recognize link between the levels of maturity and related costssavings in software maintenance. For example, the SoftwareMaintenance Capability Maturity model (SMC M M) identifies the bestpractices associated with the software processes unique to amaintainer. SMCMM was designed as a customer-focused benchmarkfor either: Auditing the software maintenance capability of a software

maintenance service supplier or outsourcer; or

Internal software maintenance organizations.

Some of the activities unique to software maintainers are:

Transition: Is a controlled and coordinated sequence of activitiesduring which a system is transferred progressively from thedeveloper to the maintainer;

Service Level Agreement (SLAs) & specialized maintenancecontracts: Maintainers negotiate SLAs and domain specificcontracts;

Help Desk handling of modification requests (MRs) and problemreports (PRs): Maintainers use a problem handling process toprioritize, document and route the requests they receive;

Acceptance/rejection of MRs: Maintainers will not acceptmodification requests work over a certain size/effort/complexityand will reroute these requests to a developer;

Impact Analysis Regression Testing: Maintainers need to perform regression

tests on the software so that the new changes do not introduceerrors into the parts of the software that were not altered.The SMCMM has been developed from a customer perspective. The

ultimate objective of software maintenance improvement pro g r a m sinitiated as a result of a SMC M M assessment is increased customer (ands h a reholder) satisfaction, rather than rigid conformance to such amodel.

A higher capability level, in the SMC M M context, means, for customerorganizations:

a)Reaching the target service levels and delivering on customerpriorities;

b)Implementation of the best practices available to softwaremaintainers;

c)Obtain transparent software maintenance services and at coststhat are competitive;

d)The shortest possible software maintenance service lead times.

For a software maintenance organization, achieving a highercapability can result in:

a)Lower maintenance and support costs;

b)Shorter cycle time and intervals;

c)Increased ability to achieve service levels; and

d)Increasing ability to meet quantifiable quality objectives at allstages of the maintenance process and services.

In the SMC M M model, the key software maintenance processes havebeen grouped into three classes (Figure 2).

a)Primary processes (operational);

b)Support processes (supporting the primary processes); and

c)Organizational processes offered by the IT unit or by otherdepartments of the organization (for example: finance, humanresources, purchasing, etc.).

The key operational processes (also called primary processes) thata software maintenance organization uses must be initiated at thes t a rt of software project development and then maintaineds u b s e q u e n t l y, beginning with the transition process. The Tr a n s i t i o np ro c e s s e n s u res that the software project is controlled and that as t ru c t u red and coordinated approach is used to transfer the softwareto the maintainer. In this process, the maintainer will focus on themaintainability of this new software.

Once the software has become the responsibility of the maintainer,the Issue and Service Request Management process handles all thedaily issues, problem re p o rts, change requests and support re q u e s t s .These are the daily services that must be managed eff i c i e n t l y. Thefirst step in this process is to assess whether a request is to bea d d ressed, re - routed or rejected (on the basis of the serv i c e - l e v e l

14

What Do You Need To Know About Software Maintenance

a g reement and the nature of the request and its size). Acceptedrequests are documented, prioritised, assigned and processed in oneof the service categories: 1) Operational Support process (whichtypically does not necessitate any modification of software); 2)Software Correction process; or 3) Software Evolution process.

It is to be noted that a number of service requests do not lead toany modification to the software. In the SMC M M model, they are re f e rre dto as operational support activities, and these consist of: a)answering questions; b) providing information and counselling; andc) helping customers to better understand the software, a transactionor its documentation.

The last two main operational processes are the Ve r s i o nManagement pro c e s s, to move items to production in a contro l l e dfashion, and the Production Surveillance process, which will ensurethat the operational environment has not been degraded. Maintainersmust also monitor the behaviour of the operational system and itse n v i ronments for signs of degradation. They will quickly warn others u p p o rt groups when something unusual happens (operators,technical support, scheduling, networks and desktop support) andjudge whether or not it is an instance of service degradation, whichneeds to be investigated.

A process which is used, when re q u i red, by an operational pro c e s s

is said to be an operational support process. In this classification, weinclude: a) the many maintenance planning processes; b) them a i n t a i n e r s education and training; c) the maintenance enviro n m e n t sand testing; d) management of the contractual aspects and serv i c elevel agreements; e) rejuvenation or re t i rement of software; and,f i n a l l y, f) resolution of problems. These are all key activities, whichare available to support some operational process activities.

O rganizational processes are typically off e red by the IT depart m e n tand by other departments in the organization (for example: humanresources, finance, quality assurance and ISO9001). While they areimportant to measure and assess, it is often easier for the maintainerto start defining the operational and operational support processes.

This generic model should help understand and position thevarious key software maintenance processes. What is important, isthat these processes be explicitly listed and classified based on theirtype (operational, support or organizational). The SMC M M w a sdeveloped in an industrial environment with practices recognised asuseful.

In summary software maintenance is not all that simple and amaturity model might be helpful at assessing your suppliersmaintenance maturity. Its proven in engineering, better design formaintenance leads to lower operational costs.

15

OperaionalSupportService

CorrectiveService

EvolutiveServices

MaintenancePlanning

MaintenanceTraining

Environment,Verification- Validation

SLAandSupplier

Management

SoftwareRejuvenation

andRetirement

CasualAnalysis and

ProblemResolution

ConfigurationManagementand document

control

ReviewProcess

MeasurementInternal AuditAnd QualityAssurance

Purchasingand HumanResources

ProcessImprovement

Issue andRequest

ManagementTransition Production

Surveillance

Version MngmtRestart and

Upgrade

Figure 2: A classification of the Software Maintainers Key Processes

What Do You Need To Know About Software Maintenance

16

Winning the battlewith downtime using RBI

Tony Musgrave

ABB Eutech Ltd, UKtony.musgrave@gb.abb.com

In continuous manufacturing 90 percent of all potentialf a i l u res are likely to be caused by just 10 per cent of the installedequipment. When planning preventive maintenance, it there f o remakes good business sense to focus on this high-risk group. RiskBased Inspection (RBI) lets you do exactly that.

ABB Eutech1 has developed a Risk Based Inspection (RBI) methodthat enables companies to substantially increase plant re l i a b i l i t y,reduce the number of plant failures and cut the time re q u i red forregular inspection/maintenance.

Recent results with four customers in the chemical andp e t rochemical industry provide compelling evidence of major savings.The total cost of Iinspection costs, for example, could be reduced by49 to 80 percent. Average inspection intervals were increased bybetween 35 and 57 percent, with an average increase of 44 per cent.And more than half the plant equipment could be removed from theinvasive inspection programs.The bottom line is that RBI re d u c e sdowntime, planned or unplanned, for a saving in maintenance costs.The time and capacity/availability that is released as a result has adirect, positive effect on a plants output.

The principles of RBI

RBI is a knowledge-based method that uses risk as a basis forprioritizing and managing an inspection program. In this definition,risk is seen as the result of the probability of future failure and thelikely consequence of that failure.

Generally speaking, the level of risk associated with the differentpieces of equipment that make up a plant is variable. However, thisfact is seldom reflected in the inspection routines applied across theinventory of equipment. Risk based inspections focus the inspection

and maintenance eff o rt on those areas where the risk and its potentialeffects are greatest.

The prime objectives of an RBI program are to:

Focus effort on identifying and reducing the safety and business risks.

Achieve increased plant availability by ensuring that outages only take place for essential inspections.

Reduce the maintenance costs associated with n e c e s s a ry or excessive dismantling and pre p a r a t i o n .

Improve safety by getting rid of hazards associated withpreparing for inspections.

RBI is becoming the pre f e rred tool by which good engineeringpractice is measured. Its predictive approach is designed to eliminateexcess and inadequacy by concentrating inspection eff o rt on re a lrisks (Figure 1). Use of RBI has shown to be effective in reducing thenumber of unforeseen breakdowns.

The ABB approach to RBI

A B B s RBI process is built around an asset care strategy that isdesigned to monitor the plant throughout its lifecycle. This involvesthe acquisition of detailed knowledge and re q u i res a goodunderstanding of the behavior of every component in the plant underits current duty conditions. Multidisciplinary in its approach, it looksat parameters such as the design/construction quality,inspection/maintenance history, and the service conditions, including

1 ABB Eutech is the process solutions center of excellence within ABBs petroleum, chemical, life science and consumer goods industries business area.

Winning The Battle With Downtime Using RBI

n o rmal and abnormal excursions. The review identifies all failuremechanisms and associated risks.

This accumulated knowledge and experience helps engineers todecide what equipment needs to be inspected and when, as well asto establish where failure would be least acceptable and cause themost problems. This makes it easier to see where eff o rt has to befocused in order to maximize the return. It facilitates the optimizationof examination intervals whilst at the same time identifying equipmentfor which non-invasive examinations would be equally effective.

ABB has refined its proven RBI approach into an eff i c i e n t ,s t reamlined solution called Risk Based Inspection(Figure 2) Anexample of the success of Risk Based Inspection+ implemented inclose partnership with a customer is a project ABB undertook re c e n t l yfor Victrex plc, a UK polymer producer.

RBI delivers for VictrexVi c t rex manufactures PEEK polymer, a high-perf o rm a n c e

t h e rmoplastic. Vi c t rex is in the fortunate position of enjoying very highdemand for its product. Its major challenge is to increase productionwhile controlling overhead and investment costs.

Working with ABB, Victrex engineers were aiming to break out ofthe vicious loop that had always linked higher output to investment innew plant. The general rule was that an annual increase of 100 tonsin output always required an investment of $1.5million.

In recent years output has been raised from 1000 tons per year to2000 tons in two jumps of 500 tons, first in 1996 and then again in 2000.In each case almost $10 million was invested in new plant.

Saving time, saving costThe inspection and maintenance routines were a prime target for

close examination. It was seen that time and cost savings could makea vast difference to output, efficiency and profitability of the plant.

B e f o re RBI the inspection regime was prescriptively invasive onall items, regardless of the risks associated with them.

It is a fact that, very often, extensive inspection during theshutdown periods reveals no deterioration. However, there areoccasions when unexpected problems are found, and this can leadto unplanned repairs that increase the outage time.

All inspection and maintenance work on the main pre s s u revessels must comply with Pre s s u re Systems Safety Regulations2000. These look mainly at how the safety perf o rmance andcondition of the vessels reduce workplace risks as far as isreasonably practical.

As a Vi c t rex engineer puts it, The total cost of inspection includesdecommissioning, decontamination and preparation for intern a linspection. All this time and eff o rt just to find that, more often thannot, everything is OK! This led us to question our approach.

It was this questioning that steered the company toward aknowledge-based approach to understanding the behavior of the plant.

The RBI objective was to raise plant availability fro m80 to 85 per cent - from 290 days to 310 days per year -which is the world-class standard for similar batchspecialty chemical processes.

It was concluded that the main contribution toi n c reasing plant availability would come from areduction in the annual shutdown time. This would needto be cut from 35 days to 21 days.

The potential financial benefit of saving a larg eamount of downtime was estimated to be almost $10million. Such an amount would result from ani m p roved gross margin and reduced short - t e rminvestment needs.

None of this means that future plant expansion isruled out, but Vi c t re x s immediate priority was to get themost return from its existing facility before embarkingon further major investment.

For both ABB and Vi c t rex there were some essential steps inensuring that RBI delivered the desired benefits. Above all else therewas the need to harness all existing in-house knowledge andexperience and link it to ABB expertise. As a starting point, the in-house team members were to pool all their knowledge andexperience. This data, both qualitative and quantitative, togetherwith historical re c o rds of previous inspections and a thoro u g hanalysis of the causes of lost output in the last 12 months, pro v i d e dthe basis for the knowledge base on which the RBI plan would bed e v e l o p e d .

A full understanding of the results of past inspections of pressurevessels during shutdowns was vital, since preparing for and carryingout invasive inspections accounts for much of the time and eff o rtinvolved in a shutdown. Knowing and understanding the failuremechanisms each item is susceptible to, and where to look for them,helped the team decide on the adequacy of non-invasive inspection.This proved to be a major benefit.

T h o roughly gathering and assessing salient historical data, plusthe use of ABB advanced software tools, enabled the team todevelop an inspection routine that prioritizes and responds to risksand foreseen failure scenarios. It also minimizes the risk ofunexpected failure .

Having carried out the full review and created an RBI program, theresults have proved immediate and impressive.

The first shutdown period to be affected, in October 2001, re q u i re donly nine days. The necessary engineering and maintenance workwas completed in four days. The total cost of the exercise was halfthe amount budgeted.

The next shutdown period, a little less than a year later, was cutf rom 35 days to 20 days. The team expects shutdowns to be kept tothis level in future years as some of the equipment gets older and wearand tear increases.

Besides the welcome time saving it provides, RBI has met andexceeded expectations. During a recent period of high demand, theg ross margin has been improved by more than $9 million over eighteenmonths of operation and other savings on capital investment andassociated depreciation exceed $3 million. Year on year shutdownsavings are running at almost $100,000.

Vi c t rex engineers admit that their initial desire to apply RBI wasan act of faith. They saw the opportunity and believed that the goalswould be achieved, but they had no previous experience to refer to.As Andrew Anderson, Engineering Manager of Vi c t rex says, The pastexperience, professionalism and expertise of ABB were vital elementsin steering the work forw a rd and in providing the confidence in theachievability of our objectives.

The Vi c t rex experience shows quite clearly that Risk BasedInspections can provide the benefits businesses need to save onpreventive maintenance while minimizing the risks of failure.

18

1 RBI risk matrix moving the operating risk envelope to a safer region

A No special care required

B Hazard study

C Periodic inspection

D Full registration procedureConsequence

Winning The Battle With Downtime Using RBI

19

ABB RBI+

ABB inspection serviceInspection life cycle database

Identifydeteriorationmechanisms

Datacollection &

analysis

Repairs &modifications

Execution ofinspection

Developinspection

plans

Reductionof

actions

Riskanalysis

Materials/NDTexpertise

Designexpertise

Inspectiontechnology

Operationalknowledge

Maintenanceexpertise

Auditing

New assets

Existingassets

Training

Defectassessment

Independentdesign review

Asset lifeplanning

Asset lifeplanning

Fitnessfor purposeassessment

Registration Support

2 Inspection life cycle database

ABB Risk Based Inspection+ in brief

Approach What - equipment to inspect, where failure would be

unacceptable Where - to focus effort How - which techniques to use When - optimize the examination interval

Benefits Increased knowledge of the risk of operating assets Significant reduction in the total cost of inspection Longer turnaround intervals, and reduced turnaround

durations Improved equipment availability and reliability, helping to

maximize uptime Increased confidence in equipment integrity Established path to regulatory approval

3 The victrex plan

IntroductionThis article describes the application of Reliability-centre d

Maintenance (RCM) to Naval assets and the re v o l u t i o n a ry changesbeing made through its' application.

Over the years, several myths and misunderstandings have arisenabout RCM: what it is; whether it consumes too much re s o u rc e ,whether it can be applied to all types of naval assets includingstructures; whether the ends justify the means. Following the recentvisit to Australia by Commander Nigel Morris RN, Head of the RoyalNavy's Warship Support Agency RCM Group it is apparent that theRN has made enormous pro g ress over the past five to six years inimplementing RCM-based maintenance programmes to the Hunt ClassMCMVs, Type 23 Frigates and other platforms and that the benefits ofRCM no longer need justification in the naval context. The RAN(through ANZAC) and the US Navy's Naval Air Warfare Center havealso started using RCM to review maintenance policies across a rangeof systems.

Against a background of having to do more with less, RCM offersa proven and robust means for the Navy to obtain maintenance 'valuefor money' and, in parallel, improve operating safety, system re l i a b i l i t yand platform availability.

The need for changeN o w h e re was the need for change in maintenance thinking and

practices been more necessary than in the aviation industry duringthe late 1950s and early 1960s. In this post-war period, new airc r a f ttypes were being brought into service with new technologies (gre a t l yincreased numbers of hydraulic, pneumatic, electro-mechanical ande l e c t ronic systems) placing new and unforeseen demands onoperators and maintainers

Up to this time aviation equipment had been much simpler and lesss t ressed, with underlying maintenance policy being based on thebelief that components and equipment displayed a 'useful life' afterwhich failures would accelerate in fre q u e n c y. In re s p o n s e ,maintenance policies were developed to change or overhaul items asthey approached this perceived 'life'. Graphically, the conditionalp robability of failure for equipment types was believed to incre a s earound a specific number of operating periods as shown in Figure 1.

As the new generation of aircraft entered service, aviationaccidents associated with equipment failure were becoming moref requent to the point at which the US Federal Aviation Authorityundertook a fundamental review of aircraft maintenance and safety.A major finding was that failure was considerably more complex thanhad previously been thought. There were in fact not one, but sixpatterns governing equipment failure, as we see in Figure 2.

Fig 2: The six failure patternsRecognition of these patterns heralded a revolution in the world of

aviation maintenance and equipment design. Patterns A, B and Csupported the existence of age-related failure, but only in a relativelysmall percentage of cases (11%), whereas Patterns D, E and F werenot age related and constituted the vast majority of failures (89%). Inthe case of Pattern F, scheduled overhaul activities associated withthe traditional Pattern B introduced infant mortality and contributedto early life failures and the appalling accident rate at the time of 60crashes per million take-offs (approximately 40 of these being due toequipment failure).

The low percentage of age-related failures (Patterns A, B and C)and the preponderance of failure patterns D, E and F in automatedsystems shifted the aviation world towards an emphasis on condition-based maintenance with a corresponding move away from scheduled

20

Revolutionising Naval Maintenance With RCM

The Asset Partnership

By Dr Alun Roberts

Pattern A: The Bathtub CurveHigh infant mortality, then a lowlevel of random failure, then awear out zone.

Pattern B: The Traditional ViewA low level of random failure, thena wear out zone.

Pattern C:A steady increase in theprobability of failure.

Pattern D: A sharp increase in the probabilityof failure settling down to randomfailure.

Pattern E: The Random FailureNo relationship at all between howold it is and how likely it is to fail.

Pattern F: The Reversed J curveHigh infant mortality, then randomfailure.

ConditionalProbabilityof Failure

Operating Periods

Fig 1: The traditional view of equipment failure.

Revolutionising Naval Maintenance With RCM

o v e rhaul activity as the primary means of asset care. This shift inmaintenance focus has been at the root of a 120-fold improvement inaircraft safety due to equipment failure since the mid 1960s.

The history and development of RCMRecognition of the complex nature of aviation equipment failure

culminated in a new approach to the development of airc r a f tmaintenance programmes which was first trialed on the Boeing 747in the late 1960s. This methodology, known as MSG-1 recognised that:

scheduled overhaul had little effect on overall reliability of a

complex item unless there was a dominant age related failuremode;

the intrusive nature of the overhaul activity itself was the causeof unreliability; and

there are many items and failures for which there is no effectiveform of scheduled preventive and/or predictive maintenance.

Over the subsequent decade, the ru d i m e n t a ry MSG-1 appro a c hwas further developed and towards the end of the 1970s, bothc o m m e rcial airline and defence aviation safety and reliability had been

21

Fig 3: RCM development since 1965

our clients winm a i n t e n a n c ee n g i n e e r i n ge x c e l l e n c ea w a r d s . . .

want one too?www.assetpartnership.comAustralia New Zealand

Revolutionising Naval Maintenance With RCM

t r a n s f o rmed. In 1974, the US Department of Defense commissionedUnited Airlines to prepare a report on the processes used by the civilaviation industry to pre p a re maintenance programmes for airc r a f t .The resulting document, entitled 'Reliability-centred Maintenance' byits' authors, Stan Nowlan and Howard Heap, heralded the birth of anew era in maintenance programme development. Reliability-centre dMaintenance is there f o re the specific maintenance developmentprocess described in this report.

In the early 1980s, RCM was first applied to non-aviation assets,primarily in the South African mining industry through John Moubray.T h rough this pioneering work, Moubray discovered that the originalNowlan and Heap concept, though sound, needed furt h e rdevelopment for non-aviation use and also re q u i red a compre h e n s i v etraining process to underpin it. In 1990, Moubray developed theindustrial version of RCM known as RCM II which over the pastdecade or so has become the standard approach adopted thro u g h o u tthe world. The principal developments of RCM over a thirty year periodare shown below (Figure 3).

In the mid 1990's, the Royal Navy developed a naval version ofMoubray's RCM II known originally as Naval Engineering Standard45, since issued as Defence Standard 02-45. Both RCM II and Def Stan02-45 approaches are fully compliant with a new SAE Standard SAEJA 1011 (Evaluation Criteria for Reliability-centred Maintenanceprocesses).

So what is RCM?RCM is a process used to determine the maintenance re q u i re m e n t s

of any physical asset so that it fulfils its intended functions over thelife cycle and in its' operating context. The RCM process mustt h e re f o re start by defining user re q u i rements or 'functions'. This initself is usually something of a challenge for most org a n i s a t i o n s .Unless this user re q u i rement is understood, it is hardly surprising thatoperators and maintainers have difficulty agreeing and communicatingon when equipment failure has occurred.

SAE JA 1011 compliant RCM asks the following seven questionsf rom which a comprehensive approach to failure for the asset can bedeveloped:

What are the functions of the asset in its present operatingcontext?

How can the asset fail to fulfil each function? What would cause each functional failure? What happens when each failure occurs? In what way does each failure matter? What can be done to predict or prevent each failure? What should be done if no suitable proactive task can be found?

The first four questions develop a functional Failure Modes andE ffects Analysis (FMEA) and the last two define the appropriate failuremanagement policy. The vitally important fifth question determ i n e show we should react to the failure in relation to whether the failureis 'Hidden' or 'Evident' and whether Safety, the Environment orOperations (Mission in the naval sense) are affected. These sevenquestions can only sensibly be answered by people who know theasset best; this includes maintainers and operators, supplemented byrepresentatives from OEMs. The group (a typical example of which isshown in Figure 4) is guided through the RCM process by a competent'Facilitator' who is an expert in the RCM process and its application

rather than the system expert.RCM is an integral part of the Integrated Logistic Support (ILS)

process as defined in Def Stan 00-60 (and highlighted in Def Stan 02-45). The process falls squarely within the Logistic Support Analysis(LSA) process and provides inputs which are needed for a rationalapproach to spares, tools and skills determination.

Application of RCM produces a 'safe minimum' maintenanceprogramme which includes:

A comprehensive range of failure management tasks formaintenance and operations staff (incorporating predictive,preventive, detective maintenance as well as the foundation forthe development of all likely cor rective tasks);

Mandated and recommended redesigns of either the asset orthe way it is operated or maintained; and

Recommendations for 'no scheduled maintenance' or run-to-failure which require the development of strategies to deal withsuch failures as they occur.

The process has been applied widely to mechanical, electrical ande l e c t ronic systems as well as to platform stru c t u res. John Moubray'sbook 'Reliability-centred Maintenance' and Def Stan 02-45 bothprovide comprehensive details of how this is achieved.

The RN RCM ProgrammeFollowing a Strategic Defence Review in the early 1990s, a decision

was made by the Royal Navy to use RCM to address excessivemaintenance manpower and re s o u rce costs and to develop a rationalapproach to risk management and extension of upkeep cycles. SinceRCM was a 'new' technology for naval platforms and systems,although well established in many other applications, fullimplementation was to be dependent of the success of a trial to beconducted on the Hunt Class MCMVs. If successful, application wouldp roceed across the fleet to include frigates, auxiliaries andsubmarines.

At the outset, the decision was made to apply a 'Whole of Platform 'approach, using the RCM process as a means of not only reviewingmaintenance on specific systems, but also providing a solid foundationfor optimising upkeep cycles.

Hunt Class TrialThe Hunt Class trial began in 1996 and used four vessels as a

c o n t rol group for comparison with non-RCM maintained vessels.Selection was based on:

22

FACILITATOR

MaintainerOperator

OperationsSupervisor

MaintenanceSupervisor

Technical Specialist

Fig 4: Typical RCM review groupFig 5: Hunt Class MCMV

Revolutionising Naval Maintenance With RCM

Hunt possessing the majority of the significant functions oflarger vessels (Float-Move-Fight);

Systems being sufficiently complex to test the RCM processcomprehensively; and

Low risk to fleet operations if the trial was not successful.

The RCM analysis took place in 1996 and 1997 following a perioddeveloping Naval Engineering Standard 45 and an associateddatabase and maintenance management software package (muchlike AMPS). After completing a functional model of the vessel, aro u n d15,000 failure modes were run through the RCM decision-makingprocess using teams comprising experienced staff who had workedclosely with the systems in service and a range of OEMrepresentatives.

Of the 15,000 failure modes examined, it was found that just 20%responded to some form of condition-based maintenance, whilst'traditional' scheduled restoration or discard was only appropriate forjust 6% of all failure modes. A further 18% of the failure modes wereassociated with protective devices which re q u i red detectivemaintenance or 'failure finding' as this is known in RCM. Two thirdsof the failure modes did not respond to any form of pro a c t i v emaintenance activity. These findings lined up closely with theexperiences of the aviation industry and reflected the complexity ofsystems installed.

The first ship to go to sea with an RCM-based programme wasHMS ATHERSTONE in April 1998 followed by BROCKLESBY, LEDBURYand QUORN. All MCMVs are expected to be operating on RCM-basedmaintenance programmes within the next year to eighteen months.

Results from the trials can be divided into three categories:operational costs, availabilities and impact on HQ and shore - b a s e dsupport organisations.

Operating costs for the RCM control group showed a 19%reduction over the period of the trial amounting to 0.5M pervessel per annum, despite a stores supply problem whichpushed up spares cost for the control group. Discounting thiseffect, savings in maintenance effort for the control group were33% lower.

Availabilities of the control group and the non-RCM vessels werecomparable, although the control group was negativelyinfluenced by the stores problem above. Control groupavailabilities have improved as stores processes have beentackled.

The most significant changes have been a shift in maintenanceeffort from non-Fleet to Fleet time, and the removal of large workpackages associated with 'Refits'. All this has required thedevelopment of a more dynamic relationship between ship andshore. The associated review of the Upkeep Cycle hasrecommended slightly more frequent, but shorter dockings - withan associated increase in availability to the extent that only 9 ofthe 10 MCMVs are likely to be needed for current tasking levelsyet under the previous regime, all 10 MCMVs would have beenneeded. The potential impact on support costs is clear if thiswere to be followed through to the logical conclusion.

Overall costs for running the RCM programme on Hunt amountedto 2M. Across the Class, savings of around 5M per annum areexpected, providing a Return on Investment measured in months.

The encouraging results of the Hunt Class trial were sufficient toh a rness support for an extensive programme to apply RCM-basedmaintenance regimes to a range of other platforms, a process whichis now well and truly underway.

Implementation to Type 23 FrigateThe RCM analysis and implementation on HMS Lancaster was

completed in mid-2002, providing confidence that the RCM pro c e s scan be applied to a major warship. Over 31,000 failure modes had beenexamined over 377 separate systems with once again, a major shiftt o w a rds condition-based maintenance with only 14% of failure modesresponding to some form of scheduled restoration or discard(overhaul) activity. Benefits obtained so far include:

Removal of maintenance with no value;

Less overhaul and reduced requirement for docking whencompared with the former maintenance cycle. Reduction in thetime needed for maintenance is flowing through to shorterperiods available for refit which in turn is forecast to improvecontractor efficiency;

Reduced time needed for testing and tuning;

On-line fault diagnostics available through the RCM analysis andassociated database and maintenance management softwarepackages; and

The acquisition of failure rate data for specific equipment failuremodes which is better enabling the determination of sparesrequirements and supply chain location.

An overall maintenance cost reduction estimated to be 15M perannum for the Class.

24

Fig 6: Type 23 Frigate Vanguard SSBN HMS Ocean

Revolutionising Naval Maintenance With RCM

Future RN Programme and BenefitsAnalysis of several more platform classes is currently underw a y

with implementation due to be complete across the majority of surf a c evessels and submarines by around 2009. These include Va n g u a rd andTrafalgar Class submarines, HMS Ocean Amphibious Assault Ship,and several Royal Fleet Auxiliaries.

Beyond this point, the RCM activity will be largely associated withon-going review and application of the process to new assets forwhich RCM to Def Stan 02-45 is an endorsed activity within ILS. Totale x p e n d i t u re on RCM activities will be roughly 40M by this time(covering all training, software development, staffing, and contractors u p p o rt) with ongoing savings of 50M per annum by 2009/10,excluding savings on spares. Additional benefits are anticipated withavailabilities expected to increase by around 10-15%. From ac o m m e rcial perspective, the RCM project internal rate of re t u rn isabout 80% over this time period at current interest rates.

Fig 7: Forecast RN annual cost savings through RCM application

RAN and other Navy experience

It is not only the Royal Navy which has been busy re v i e w i n gmaintenance re q u i rements using RCM. Over the past eighteen monthsor so, the Royal Australian Navy has applied both Def Stan 02-45 andRCM II versions of RCM to a number of assets including three onHMAS MANOORA and two on ANZAC.

P a rticularly noteworthy is the ANZAC Ta rget Indication Radarwhich was analysed in its' entirety using a team comprising RAN staffand re p resentatives from the OEMs. As with the RN eff o rt, re d u c t i o n sin the level of scheduled discard and replacement were highlightedand increased use of condition-based maintenance techniques wererecommended. This has the effect of securing the maximum servicefrom expensive system components which permit the RAN to reducelife cycle support costs for the Radar by potentially several millionsof dollars without increasing operational risk.

Not surprisingly, faced with this type of saving, disbelievers in theRCM process argue that reductions in the level of pro a c t i v emaintenance will be quickly followed by increases in re a c t i v emaintenance. The US Navy, through the Naval Air Wa rf a re Center(Navair) has been monitoring whether this happens in practice acro s sa number of assets which have received RCM analysis. For thesesystems, proactive maintenance eff o rt has been reduced by between55 and 80% and corresponding reactive maintenance has in factdeclined, initially by 5% and now running in excess of 10%.

For more information, contact Dr Alun Roberts, Dire c t o r, The AssetPartnership. (See their advertisement in this issue - page 19)

25

P roduct ofthe year 2003

EuropeanElectronics

Industry Awards

Contact MSc for further details on our range of infraredCamera Systems or list of our authorised distributers.

27 Research Drive, Croydon VIC 3136ph 03 9761 5088 fax 03 9761 5090email: sales@maintsys.com.auweb: www.maintsys.com.au

Infrared Thermography.... find temperature problemsbefore they find you!

Affordable Portable Thermal Imaging camerasand software start from under Aud $5,000.

Revolutionising Naval Maintenance With RCM

AbstractThis is a Case Study on the successful introduction of TPM3 ( To t a l

Productive Maintenance) to an Australian underground zinc mine. The paperis about the experiences of a small group of miners taking their first steps ontheir TPM3 J o u rney to Excellence. It is not just about maintenance; it is aboutleadership, empowerment, learning by doing, ownership, perf o rm a n c em e a s u rement, working together and culture change.

What is TPM3?TPM is not a new concept, it was developed by Toyota in the 1970's

to lift equipment reliability to the higher levels re q u i red by theintroduction of the Toyota Production System and Lean Production. Ithas evolved over the years from just being equipment focused (1s t

Generation), to process focused (2nd Generation), and then companyfocused (3rd Generation).

T P M3 is an enhanced and expanded Australasian version of 3rd

Generation TPM. It is a company wide equipment managementi m p rovement strategy involving all employees aimed at significantlyimproving capacity, productivity, quality, delivery, safety, morale andbottom-line results.

T P M3 is not a simple maintenance program and it cannot beimplemented by a handful of people. It requires the co-operation andinvolvement of all levels of the company, the breaking down oftraditional attitudes towards specialization (I operate, you fix) and theestablishment of educational systems designed to upgrade skill levels

of all employees, especially maintenance and production personnel.

The Centre for TPM (Australasia) is a membership-basedo rganisation providing companies with TPM3 Training, Navigation,R e s e a rch and Networking. Since it's inception in 1996, the Centre hasdeveloped TPM3, an enhanced and expanded Australasian version of3rd Generation TPM. TPM3 is now being introduced thro u g h o u tAustralasia at over 40 sites involving some 14 industry groups.

What is World Class EquipmentPerformance?

So what is World Class and how do you know when you get there ?The TPM3 Methodology utilizes the holistic measure ' O v e r a l lEquipment Effectiveness" (OEE) to measure progress towards WorldClass Equipment Perf o rmance. The OEE incorporates three traditionalmeasures; Availability (impacted by break downs and set-ups) Performance Rate (impacted by slow running, idling and minor

stops)

Quality.Companies are recognized as having achieved World Class

Equipment Performance when they have achieved and sustained anOEE in excess of 85%. Most companies find that the OEE on their keyitems of plant is less than 50% at the start of their TPM3 J o u rney toExcellence. See figures 1 and 2.

Pasminco and TPM3

Pasminco first trailed TPM3 in 1998 at their smelter at Port Pirie inSouth Australia. The initiative was extremely successful, lifting keyplant OEE's towards the World Class levels and yielding significantcost savings with the added benefit of a greatly improved safetyp e rf o rmance. Pasminco operates a number of zinc mines in Australia,but it is the Rosebery Mine that is leading the way in the TPM3 J o u rn e yto Excellence. Rosebery is a medium sized underground base metalsmine located in Tasmania. The mine is over 100 years old but is quicklybecoming regarded as one of the most innovative and profitable zincmines in the country. The site employs some 220 people and mines

over 750,000 tonnes of ore per annum.Under the leadership of General Manager, Brett Fletcher, the

Management Team had been looking at TPM3 for some time as a

26

A TPM3 Journey To ExcellenceCase StudyThe Introduction of TPM3

into a Zinc Mine

General Manager Mining & Resources, The Centre for TPM (Australasia)

Keith Saul

A TPM3 Journey To Excellence Case Study

27

OverallEquipment

Effectiveness

Target

Zero

Minimise

Zero

Zero

Zero

Minimize

Overall Equipment Effectiveness Modelfor the Mining Industry

Six Big Losses

Availability

PerformanceRate

QualityRate

Set-up andRepositioning

ReducedSpeed

Idling andMinor Stoppage

Downgrade andReprocessing

YieldLoss

BreakdownFigure 1

Defining Equipment Losses and OEEFigure 2

Total Elapsed Time

Scheduled Production Time

Available Run Time

Reported Operating Time

Net Operating TimeEfficientNet Operating Time

Value AddingTime

Planned Downtime

Reduced Speed

Rejects andRework

Set-upTime

UnplannedRecordedStoppages

MinorUnrecordedStoppages

Overall Equipment Effectiveness (OEE):

Equipment Effectiveness (EE):

Value Adding Time

Elapsed Time

Value Adding Time

Scheduled Production Time

A TPM3 Journey To Excellence Case Study

Making it easier to work smarterWith simple methodology and powerful software.

Interest for Reliability / Maintenance Training in 2004Please fill in the following MJ form to register interest for Reliability / Maintenance

Training in 2004

Name.............................................................................................................................................

Organization.................................................................................................................................

Position..........................................................................................................................................

Phone ...................................Fax.......................................Email ..................................................

Address..........................................................................................................................................

State .....................................Country ...............................Postcode..............................................

I am interested in attending. Please send me more information & training dates;

q Root Cause Analysis 2 Day Practitioners Course

q Root Cause Analysis 1 Day Participant Course

q Understanding Reliability

q Introduction to RCM

q Advanced RCM Skill Building Using RAMS Analysis

q QRA studies using Faultree Plus

q Lifecycle Costing

q System Availability Modeling

q Weibull Analysis

q Performing Hazard and Operability Studies with Hazop Plus

q Reliability 2004

q On-site training (Please specify area of interest)

q MELBOURNE q BRISBANE q PERTH q GLADSTONE q NEWCASTLE

q ADELAIDE q GOLD COAST q AUCKLAND q NEW PLYMOUTH q SYDNEY

q OTHER PLEASE INDICATE

SUGGESTED DATES ...........................................................................................................................

SUGGESTED NUMBER OF ATTENDEES .............................................................................................

PLEASE COMPLETE & FAX TO MICHELLE: 03 5255 5778

ARMS RELIABILITY ENGINEERS, PO BOX 501, OCEAN GROVE VIC 3226

* PH: 03 5255 5357 * FAX: 03 5255 5778 *

* EMAIL: arms@reliability.com.au * WEB: www.reliability.com.au *

RELIABILITY 2004 - The Must Attend event on your 2004 Reliability Calendar

This annual event is designed around the needs of our users. Hear about the latest developments in RAMS software and learn howothers are getting results. Knowledge sharing is a focus each year with case studies presented by users of RAMS software andARMS methodology. Listen and learn from other national & international organisations from a large range of industries about thechallenges and benefits of Reliability Engineering. Round table sessions, discussion forums and social activities ensure aninteractive networking pro g r a m .

A P O L LO ROOT CAUSE ANALYSIS - Finding effective solutions that prevent recurrence and remove blame

Apollo RCA at any level is suitable for anyone whose job includes problem solving. There are 3 Apollo courses to choose from: TheHalf Day Managers course: 1 Day (Participants Course); or 2 Day (Practitioners Course). The one-day participants course pro v i d e sattendees with the skills necessary to find effective solution and understand the benefits of problem definition, creating a causeand effect chart and finding a range of possible solutions before deciding which are the best. The two day program looks furt h e rinto what makes a successful program, and focuses on facilitation of Root Cause Analysis sessions. The half day managers pro g r a mdeals with what is necessary to manage a successful Apollo RCA program and institutionalize it in an organization.

U N D E R S TANDING RELIABILITY Using Powerful Technology and Simple Methodology to achieve world

class Reliability Practices

This 2 day seminar is aimed at Managers, Planners and Practitioners who are involved in managing the reliable perf o rmance ofassets. The objective is to provide attendees with an awareness of how to improve asset perf o rmance using Reliability EngineeringMethods. Background theory is combined with workshops using the latest RAMS technology to enhance skill levels. Case studieswill also be presented using RCMCost and AVSIM Plus for Maintenance Optimization and System Availability Analysis. All attendeeswill gain an appreciation and understanding of the benefits of using reliability methods including, reduction in cost of capital fornew projects, improved quality and safety, increased equipment re l i a b i l i t y, reduced maintenance and operating costs and pro a c t i v emaintenance management.

RCM SKILL BUILDING USING RCMCOST Maintenance Optimisation for Reliability Engineers, Planners or

Personnel involved in improving their RCM Program

RCM Skill Building Using RCMCost is a 2 day workshop, designed to provide an understanding of the RCMCost method ofmaintenance task optimization. The workshop will be of practical benefit to technicians, engineers, planners and managers,p roviding the information that allows them to assess the cost effectiveness of existing maintenance plans, and know how to selectoptimum tasks and frequency of maintenance tasks for both new and existing equipment. This course also covers the essentialdata analysis skills necessary to be able to utilise failure re c o rds and equipment history. Instruction is through a series of targ e t e de x e rcises to ensure all participants leave with practical hands on knowledge.

PLANT AVAILABILITY MODELING Evaluating & Improving System Performance through Availability

S i m u l a t i o n

This course provides the knowledge and skills necessary to perform system analysis using Reliability Block Diagrams andAvailability Simulation. This is a skill building course designed for those wishing to improve the availability of systems and re d u c eplant downtime. Practical examples will be worked on using AvSim+ availability simulation program.

PERFORMING HAZOP Studies Identifying hazards for New and Existing Plant

Identifying hazards through HAZOP studies is the first step in many risk management programs. This course teaches the basicsinvolved in understanding Risk and how to use the Hazop Technique to identify Hazards, rate the hazard and identify which hazard sneed to be addressed to reduce unacceptable exposure. A systematic approach is essential for an efficient Hazard identificationand risk reduction program. This course introduces both spreadsheet and database applications that can be employed.

QRA STUDIES USING FAU LTREE PLUS

Quantified Risk Analysis can be used to identify the critical events carrying the highest risk and evaluate the effectiveness of riskreduction options. Used for Risk Management, Level of Protection Analysis, Safety Integrity Level Determination, Isographs Faultre ep rogram is a world leader. This day is a familiarization day with one of the worlds leading programs to perf o rm Faultree and EventTree analysis.

possible vehicle to involve everyone in continuous improvement.

In a move that signalled his commitment to TPM3, Brett Fletchertook 20 of his key people to a 2 day Introduction to TPM3 Workshop.

TPM3 Introduction StrategyA TPM3 Leadership Team was established comprising the General

M a n a g e r, Mine Manager, Technical Services Manager, Metallurg yM a n a g e r, Metallurgy Superintendent, and the two TPM3 C o - o rd i n a t o r s .A half-day TPM3 I n t roduction Strategy Workshop resulted in theselection of the first two pilot improvement teams, one from thes u rface (Mill) operations and the other from underg round (Mine)operations. The teams were each given a mandate, which was toidentify all of the losses associated with their selected pilot area, toi n c rease the Overall Equipment Effectiveness (OEE) by 25%, makerecommendations for further loss improvement activities, and tocomplete the project within 12 weeks.

The Journey - Team Awareness TrainingAll team members attended a one day TPM3 Aw a reness Wo r k s h o p .

The majority of the team members had never worked at another siteand with many long serving employees amongst them, there wasseveral hundred years of experience in the room.

Team Kick -Off The CycleThe first of 12 team meetings was a half-day kick-off workshop to

f o rm the team through a series of team building exercises and to guidethem through the stru c t u red step-by-step approach that is an integralpart of the TPM3 Methodology. (Figure 3)

Going SoloFor a number of team members this was a major learn i n g

experience wrestling with issues such as how to be a team leader /team member, how to plan meetings around shift work, how to carryout surveys and how to analyse data with the added pressure of a 12week time limit to achieve a significant improvement in performance(25% increase in OEE). The team followed a 9-step process, with eachstep clearly outlined in their TPM 3 Team Member Manuals.

Analysing the Current SituationThe Mine Team, which focused on Mine Development, spent the

next 3 weeks busily analyzing the current situation, determining theOEE for the Drilling Rig and identifying the OEE losses, the strategicintent and expected life of the equipment; the baseline performanceof drilling and blasting, the preventive/ predictive maintenancecapability, the status of drawings, operating procedures, spares etc.

This was complemented with a survey of their work mates tod e t e rmine how operators and maintainers viewed the equipment inre g a rds to ease of operation, re l i a b i l i t y, process capability,housekeeping, safety, environment, rework / rehandle, andmaintenance practice.

The results provided the team with valuable information about theequipment and processes and a better understanding of issuesaffecting the rate of development. The survey involved all other plantoperators in the project and provided some interesting feedback onthe WIIFM (what's in it for me) issues that need to be addressed ino rder to reduce operator frustration and gain commitment andownership to any changes in operating procedures.

30

1. IdentifyOpportunities

QualityAssurance

Presentation

Act Plan

DoCheckPresentation

6. Pilot ProposedSolutions

7. Refine and ImplementSuccessful Solutions

2. Form Teamand

Scope Project

3. AnalyseCurrentStation

4. Develop Visionof Improved

Performance

5. Identify PossibleRoot Causes

and Solutions

8. Measure Prog