Trends and Perspectives in Industrial Maintenance Management LUXHOJ RIIS THORSTEINSSON

Prévia do material em texto

Journal of Manufacturing Systems 
Vol. 16/No. 6 
1997 
Trends and Perspectives 
Trends and Perspectives in Industrial 
Maintenance Management 
James 1". Luxhej , Rutgers University, Piscataway, New Jersey 
Jens O. Riis, Aalborg University, Denmark 
Uffe Thors te insson, Technical University of Denmark, Denmark 
Abstract 
With increased global competition for manufacturing, 
many companies are seeking ways to gain competitive 
advantages with respect to cost, service, quality, and on- 
time deliveries. The role that effective maintenance man- 
agement plays in contributing to overall organizational pro- 
ductivity has received increased attention. This paper pre- 
sents an overview of trends and perspectives in industrial 
maintenance. The results of benchmarking studies from 
Scandinavia and the United States are presented and com- 
pared. Implications of the trends and perspectives for the 
management of maintenance are highlighted. Case studies 
that examine maintenance methods, knowledge, organiza- 
tion, and information systems in three Danish manufacturing 
firms are used to motivate the discussion. 
Keywords: Maintenance Management, Benchmarking 
Introduction 
As a basis for focusing on the discussion in this 
paper, British Standard 38111 defines maintenance 
as: 
The combination o f all technical and associated 
administrative actions intended to retain an item in, 
or restore it to, a state in which it can perform its 
required function. 
In this paper, maintenance is interpreted in the 
context of industrial systems (such as production 
facilities, buildings, equipment, and so on). The 
main purpose of this paper is to present the manage- 
rial implications of industrial maintenance trends. 
Wireman, 2 in his book World Class Maintenance 
Management, refers to maintenance planning as "the 
last frontier" for manufacturing facilities. In the 
move to world-class manufacturing (WCM), many 
firms are realizing a critical need for effective main- 
tenance of production facilities and systems. With 
the trend to just-in-time (JIT) production and flexi- 
ble, agile manufacturing, it is vital that maintenance 
management becomes integrated with corporate 
strategy to ensure equipment availability, quality 
products, on-time deliveries, and competitive pric- 
ing. The changing needs of modern manufacturing 
necessitate a reexamination of the role that improved 
maintenance management plays in achieving key 
cost and service advantages and of the contribution 
that maintenance improvement can make to the 
"learning organization. ''3 
Numerous technological and managerial trends 
for the past two centuries have changed the require- 
ments and content in maintenance. For example, the 
development of the automobile led to a worldwide 
network of distribution systems for spare parts and 
maintenance services. Similar systems have later 
been designed for many other industrial products 
(for example, ship engines, airplanes, and industrial 
production). The development of air transportation, 
satellites, telecommunication systems, and so on 
have led to new standards for design of technologi- 
cal systems and for maintenance requirements based 
on 100% system functionality. The development in 
maintenance requirements has been supported by 
advances in information technology that have given 
new possibilities for maintenance performance. 
Developments in the chemical industry (such as oil 
refineries, plastic and dyestuff factories, biochemical 
companies, and pesticide manufacturers) have added 
some new aspects for maintenance: environmental and 
labor protection. Technological developments have 
also given the maintenance area new tools and meth- 
ods: nondestructive material tests, automated data col- 
lection systems, computer tools, video technologies, 
and so forth. New trends indicate more application of 
computer systems for diagnosis and analysis. 
437 
Journal of Manufacturing Systems 
Vol. 16/No. 6 
1997 
In addition to technological changes, several 
trends in management have changed the viewpoints 
on maintenance. Managerial attitudes have changed 
toward maintenance because new management 
philosophies, such as JIT, have focused on reduced 
time for delivery or enhanced quality. Trends with 
job enrichment and automation have led to embed- 
ding maintenance information technology in prod- 
ucts and production equipment, with corresponding 
changes in maintenance jobs from mechanical to 
electronic. Sociological trends, such as lack of capi- 
tal, fluctuations in currencies, competition, quality, 
and environmental consciousness, have also affected 
the need for maintenance. 
The next section presents a synopsis of two 
benchmarking studies of maintenance. Case studies 
of the experiences and challenges of three Danish 
companies with maintenance management are pre- 
sented. Managerial implications of these industrial 
perspectives on maintenance are emphasized. 
Benchmarking of Maintenance 
Thorsteinsson 4 observes that the importance of 
maintenance in society, or more specifically in 
industry, can be described in different ways by using 
various measures, such as: 
• Accounting of the total maintenance cost. 
• Percentage of maintenance cost to total produc- 
tion cost or capital cost in assets. 
• Total number of personnel working with mainte- 
nance--or percentage of maintenance personnel 
to total number of production personnel. 
• Possible consequences for lack of maintenance: 
financial, environmental, human, equipment 
damage, and so on. 
Traditionally, maintenance data have been diffi- 
cult to obtain from organizations. However, the fol- 
lowing sections summarize two maintenance bench- 
marking studies. 
Benchmarking Maintenance in 
Scandinavian Countries 
In February 1992, a EUREKA 5 project was initi- 
ated that attempted to benchmark maintenance in 
Scandinavian countries. Participating countries were 
Denmark, Norway, Sweden, and Finland. Denmark 
served as the liaison country. Some key Nordic 
maintenance societies participated, such as the 
Danish Maintenance Association, the Finnish 
Maintenance Society, the Swedish Maintenance 
Society, and the Norwegian Society of Maintenance 
Engineers. 
The aim of the benchmarking study was to estab- 
lish a trade-by-trade overview of maintenance meth- 
ods to assist companies in identifying areas for 
improving maintenance. As such, general mainte- 
nance trends may be induced from the report. It must 
be remembered that many dynamic factors make 
interpretation of the study's results among countries 
difficult. Some of these factors are due to varying 
age and quality of machinery and buildings, inter- 
pretation/use of maintenance concepts, varying 
environmental conditions, differing forms of pro- 
duction operations (number of shifts, production 
technology, and so on). 
In Denmark, the benchmarking study was based 
on an analysis of questionnaire responses from 43 
industrial companies. The companies accounted for 
approximately 12% of the total turnover in Danish 
industry and approximately 8% of industrial 
employment. The industrial sectors of chemical and 
petroleum, nonmetallic mineral products, and man- 
ufacturers of food, beverages, and tobacco account- 
ed for approximately 64% of the industry sector 
turnover in the sample. 
On average, approximately 4.9% of the compa- 
nies' turnover in 1991 was spent on maintenance, 
which was the same percentage as 10 years earlier. 
From 1981 to 1991, there were increases in mainte- 
nance costs (expressed as a percentage of capital 
value) for the overall survey average (0.6%), for pro- 
duction, transport, and storage equipment (0.9%), 
and for spare parts (0.4%). 
The "average" Danish company represented in 
the survey spent 32% of its maintenance budget on 
spare parts, 32% on salary and wages, and 31% on 
external services. In the average company, 23.8% of 
the maintenance costs was attributed to unforeseen 
repairs, 28.7% to preventivemaintenance, and 
45.5% to planned repairs. 
The survey showed that, on average, 58% of the 
companies use quantitative figures for measurement 
and control of costs, 58% use figures for availabili- 
ty, and 44% use quantitative figures for service 
level. Regarding maintenance organization, approx- 
imately 20% of the companies have a centralized 
438 
Journal of Manufacturing Systems 
Vol. 16/No. 6 
1997 
maintenance function with a maintenance manager. 
In 23% of the companies, the production manager is 
responsible for maintenance, and in only 16% of the 
companies is the production and maintenance func- 
tion integrated. In 84% of the companies, the main- 
tenance function has a medium to large influence in 
the design and specification phase. 
Approximately 39% of the time spent on mainte- 
nance is used for unforeseen repairs, 20% for pre- 
ventive maintenance, and 37% for planned repairs. 
Planning and control of preventive maintenance is 
performed in 45% of the companies. Use of the 
computer to control spare parts increased from 10% 
to 50% from 1981 to 1991, and computer usage to 
control preventive maintenance increased from 9% 
to 60% in the corresponding period. However, 25% 
of the companies do not have any inventory control 
procedures in place for spare parts. 
In Finland, the benchmarking survey was based 
on responses from 80 companies, which accounted 
for approximately 12% of the total Finnish turnover 
and approximately 14% of industrial employment. 
On average, approximately 4.8% of the companies' 
turnover in 1991 was spent on maintenance. 
The Swedish maintenance survey was based on 
responses from 71 of 200 large and medium-sized 
companies from varied industries, such as chemical, 
paper and pulp, steel and metal works, machine and 
transport equipment, electromechanical, and food. 
The Swedish survey illustrates that despite discus- 
sions of decentralization of maintenance resources, 
in the participating organizations, the majority of 
maintenance resources used (approximately 70%) 
are organized centrally. For 62% of the companies 
surveyed there was a written policy on maintenance. 
The companies in the survey identified the highest 
priorities for improvement as the maintenance skills 
of the production staff, involvement of the produc- 
tion staff in maintenance work, continuous use of 
key figures, knowledge of maintenance throughout 
the organization, and control of the effects of main- 
tenance on production volume. Organizational 
aspects, such as skills, goals, key metrics, and so on 
were generally given high priority while administra- 
tive aspects, such as planning, control, and followup, 
were given low priority. Also, the survey indicated 
that the companies with the fewest number of shifts, 
or the shortest production time, reported a greater 
need for improvement. 
Norway received 194 responses to its mainte- 
nance benchmarking study; approximately 60% of 
the respondents were from the food, engineering, 
and chemical industries. Seventy percent (70%) of 
the companies were small and medium-sized enter- 
prises (SMEs). About 56% of the companies had no 
clear maintenance and availability objectives. Most 
of the companies had a centralized maintenance 
function. The workload in the maintenance depart- 
ment in about 45% of the companies is reported to 
be large or extremely large and is increasing in about 
22% of the companies. Forty percent (40%) of the 
companies have preventive maintenance programs. 
Off-line condition monitoring is performed in about 
60% of the companies and involves visual inspec- 
tions, vibration analysis, pollution control, and oil 
analysis. The most common reason for faults and 
failures is normal wear and tear, but incorrect usage 
of equipment during production is a considerable 
factor. In about 77% of the companies, production 
personnel perform simple maintenance tasks. Only 
about 27% of the companies use a computer-based 
maintenance system. 40% of the companies have no 
maintenance information system in place. 
In Norway, for 60% of the companies, mainte- 
nance activities are the major causes of production 
downtime and stoppages. Planned stoppages 
account for only about 30% of the total number of 
stoppages. Forty percent (40%) of the companies 
have not defined availability measures for produc- 
tion equipment. 
The common trends from this Scandinavian 
benchmarking study for maintenance suggest that 
there is a need to develop clear maintenance objec- 
tives and goals, to define key variables for measur- 
ing and controlling maintenance activities, to ensure 
better linkages between maintenance and produc- 
tion, to focus on organizational issues, to move 
toward computer-based maintenance systems, to 
decentralize some maintenance activities, to instill 
better training, and to investigate modern mainte- 
nance methods. 
Benchmarking Maintenance in the United States 
Wireman 2 conducted a similar maintenance 
benchmarking survey in the United States. Since 
1979, maintenance costs for industrial firms in the 
United States have risen 10-15% per year. 
Unfortunately, the total "waste" in excessive mainte- 
439 
Journal of Manufacturing Systems 
Vol. 16/No. 6 
1997 
nance expenditures was approximately 200 billion 
dollars in 1990, which equaled the total maintenance 
costs in 1979! 
Wireman's survey results illustrate the need for 
better maintenance planning and the need for more 
maintenance research and development in U.S. com- 
panies. Only about one third of all firms in the sur- 
vey were using a maintenance planning system. 
Most of the time, failed components were simply 
changed and seldom was any failure analysis per- 
formed. Overtime by the maintenance staff typically 
accounted for 14.1% of the total maintenance time. 
There were notable difficulties in coordinating the 
maintenance and production functions in a typical 
manufacturing organization. Also, many firms in the 
survey were carrying excessive maintenance inven- 
tories for the "just-in-case" situations. As a conse- 
quence of the above problems, it is not surprising 
that the survey average of lost production/mainte- 
nance cost is 4 to 1 and that this ratio is as high as 
15 to 1 in some cases. Such a situation is unaccept- 
able if a firm truly intends to compete in a world- 
class economy. 
The principal maintenance problems were sched- 
ule conflicts, hiring, breakdowns, training, excessive 
inventories, lack of management support, and bud- 
get control. Wireman contends that one of the pri- 
mary reasons for the failure of maintenance man- 
agement systems is that frequently the planning and 
supervising functions are mixed in an organization. 
Such a situation creates confusion and leads to a 
reactive rather than proactive approach to mainte- 
nance management. 
Benchmarking the Managerial Effort in 
Maintenance Management 
With economic support from the Nordic 
Industrial Fund and from national funds, the Nordic 
Research Program, entitled "Terotechnology-- 
Operational Reliability and Systematic 
Maintenance," was undertaken during the same time 
as the EUREKA project. Several technical universi- 
ties in the Nordic countries investigated specific 
projects, such as accelerated testing, collection of 
failure data, condition monitoring, and operational 
reliability? 
One project in the last group was entitled 
"Maintenance Profiles for Industrial Systems. ''7 
The aim of this project was to develop an analytical 
too1 tO measure the managerial effort in maintenance 
management and to complete some industrial case 
studies demonstrating the use of the new tool. The 
intent was to establish an alternative to benchmark- 
ing based exclusively on cost, because high cost may 
be a consequence of poor management. Therefore, 
the research premise was that sound benchmarking 
should include both cost studies and analysis of the 
managerial effort toward maintenance. 
The resulting analytical tool is referredto as the 
Management Radar Profile, which shows the man- 
agerial effort in 12 main control areas for mainte- 
nance. Thorsteinsson and Hage 7 develop a broad 
definition of the maintenance task based on viewing 
the maintenance system as a "production system" 
where the "products" are maintenance services. 
These authors identify 12 main maintenance tasks or 
"fields" that are grouped into three primary cate- 
gories-technical, human, and economic. The tech- 
nical category of the maintenance task is comprised 
of maintenance "products" (maintenance services), 
the quality of these products, and the methods, 
resources, materials, and control strategies for main- 
tenance. The human category of the maintenance 
task is comprised of internal relations between the 
maintenance organization and other departments 
(such as production), including corporate staff, 
external relations to local regulatory authorities, 
labor organizations, vendors, and so on, and the 
design of the maintenance organization itself (that 
is, centralized vs. decentralized, responsibilities, 
skill levels, personnel selection, and so on). The eco- 
nomic category of the maintenance task is com- 
prised of the structure of maintenance (work break- 
down, relation to accounting system), the economic 
control of maintenance (budgets, cash flow), and 
production economy (economic relationship 
between maintenance and production). Definitions 
of these 12 main maintenance fields are elaborated 
on in Figure 1. 
As shown in Figure 2, the radar diagram contains 
the 12 control areas shown as fans in a full circle. 
The diagram also includes five concentric circles 
corresponding to the levels in the control circle: for- 
mulating goals and strategies, analysis, recording 
and controlling, specifying and planning, and con- 
ducting the work. This radar diagram, based on the 
graphical concept of a radar screen for tracking 
ships or airplanes, may then be used as a diagnostic 
440 
Journal of Manufacturing Systems 
Vol. 16/No. 6 
1997 
The technical part: 
T h e 1 2 M a i n M a i n t e n a n c e M a n a g e m e n t F i e l d s 
• The maintenance "products." Specification of the different types of services and "products" from the maintenance function. 
Specification in relation to each plant system. 
• Quality of the maintenance "products." Specification of quality of the maintenance jobs. Quality reports, certification 
documents, decision about maintenance standards, etc. 
• Ml/tSnance working methods. Specification of working methods, time standards, relation between maintenance jobs, etc. 
• Maintenance resources. Equipment for maintenance, buying maintenance services, information about new equipment, capacity 
of equipment, usage control, etc. 
• Maintenance materials. Inventory planning (spare parts, etc.), warehousing, relation to vendors, etc. 
• Controlling maintenance activities. Scheduling of maintenance jobs, progress in work, manpower planning, etc. 
The human part: 
• Internal relations in maintenance function. Relation to other departments, corporation and coordination especially to 
production. 
• External relation for the maintenance function. Relation to external parties, especially related to environment and safety. 
Contact to local authorities, press, labor organization, customer, vendors, neighbors, etc. 
• Organization of the maintenance function. Design of the organization, selection of people, relation between groups of skills, 
responsibility, and authority. 
The economic part: 
• Structure of maintenance. Work breakdown of maintenance, responsibility for work packages, area structure, relation to 
accounting system, specification base (drawings, documentation), etc. 
• Maintenance economy. Economic control of maintenance: Cost estimates, budgets, cash flow, accounting for the maintenance 
function. Plant investment and financing. 
• Production economy. Production economy versus maintenance economy, cost benefit of maintenance. 
Figure 1 
Definitions of the Thorsteinsson-Hage 712 Maintenance Tasks 
tool to evaluate the level o f managerial effort cur- 
rently being expended by an industrial company 
toward its maintenance activities. The management 
effort related to each cell in the diagram is shown 
through the percentage o f the area that is shaded. 
Twelve small circles show the amount of success 
factors related to a specific control area. I f a small 
circle is heavily shaded, one might expect that the 
managerial effort in this area should be heavily 
shaded, based on the notion that there must be some- 
thing to manage or some goals to guide the course o f 
action. 
Use o f the radar diagram leads to the develop- 
ment of"maintenance profiles" that clearly illustrate 
to management where efforts have been allocated 
with respect to the maintenance task within the orga- 
nization. Moreover, the radar diagram may also be 
used to suggest changes to management efforts and 
to evaluate the possible organizational effects o f 
new, improved maintenance management programs. 
The radar tool has been computerized, and the 
authors report positive test case results from analyz- 
ing maintenance tasks at a steelworks, a manufac- 
turer o f electrical lamps, and a vendor selling main- 
tenance services for industrial control systems. 
Management Radar Profiles for two o f the three 
case companies are reported in this paper. 
B e n c h m a r k i n g L i f e c y c l e C o s t s for 
I n d u s t r i a l S y s t e m s 
The idea o f designing for the system lifecycle has 
been gaining popularity. The l ifecycle concept 
addresses the phases o f research and development, 
production and construction, operations and mainte- 
nance, and system retirement and phaseout ) It has 
been suggested that approximately 80% of a prod- 
uct's or system's lifecycle costs are determined by 
the completion o f the design phase. 9 Poor design can 
lead to excessive manufacturing and maintenance 
costs. As a benchmark, Table i presents some main- 
tenance cost ratios for various industries. 
C o n c l u s i o n s f r o m B e n c h m a r k i n g S t u d i e s 
The common trends from Scandinavian ( E U R E - 
441 
Journal of Manufacturing Systems 
Vol. 16/No. 6 
1997 
Oualit 
Maintenanc intenance 
working meth conomy 
Maintenance Maintenance 
resources structure 
Maintenan 
material: laintenance 
rganization 
ions for 
ma=menance mam~nance function 
activities Internal relations in 
maintenance function 
Figure 2 
Thorsteinsson-Hage 7 Radar Diagram 
for Analysis of Maintenance Tasks 
KA s) and United States 2 benchmarking studies for 
maintenance suggest that there exists a need to 
develop clear maintenance objectives and goals, to 
define key variables for measuring and controlling 
maintenance activities, to ensure better linkages 
between maintenance and production, to move 
toward computer-based maintenance systems, to 
decentralize some maintenance activities, to instill 
better training, and to investigate modern mainte- 
nance methods. 
Two major trends in the development of mainte- 
nance management research identified by the bench- 
marking studies may be succinctly stated as: 
1. emerging developments and advances in mainte- 
nance technology, information and decision 
technology, and maintenance methods; and 
2. the linking of maintenance to quality improve- 
ment strategies and the use of maintenance as a 
competitive strategy. 
The next section of the paper presents further 
elaboration of the issues associated with the first 
Table 1 
Representative Maintenance Cost Ratios 
Source: D. Remer, J. Sherif, and H. Buchanan, "Determining 
Maintenance Cost Ratios" Cost Engineering (v34, n7, July 1994), pp9-13 
Item Industry Annual Maintenance Comment 
Cost, % 
1 Communications and 
data processing 
Communications 
(network tracking) 
Federal Aviation 
Adminstration (FAA) 3.2 A 
Communication 
Satellite Corp. 
(COMSAT) 4 A 
European SpaceAgency (ESA) 8 - 12.5 A 
Software 
GTE 10 B 
GM 15 B 
USAF 14 B 
Software (in 
general) 8- 16 B 
Computers 
Large machines 4 - 7A 
Minicomputers 
(contract) 10 A 
Workstations 
(a) Contract 15 A 
(b) 24-hr service >20 A 
Avg. All computer 
industry 10 A 
2 Industrial chemical 
companies (Dow, Union 
Carbide, Ethyl, etc.) 3.3 - 13.5 C 
3 Specialty chemical 
companies (Coming, 
4 
A 
B 
C 
Petrolite, Loctite, etc.) 2.3 - 9.8 C 
Petroleum 5 C 
Annual maintenance cost as a percentage of total 
equipment costs (replacement) 
Annual maintenance cost as a percentage of total software 
lifecycle (assumes five-year life) 
Annual maintenance cost as a percentage of current plant 
and equipment cost 
major trend and summarizes the impact that artifi- 
cial intelligence techniques, such as expert systems 
and neural networks, are having on the formation of 
maintenance knowledge in industrial organizations. 
The second major trend is typified by the emergence 
of total productive maintenance (TPM), which is a 
relatively new approach to the development of main- 
tenance systems./°,ll TPM is a lifecycle and employ- 
ee involvement approach to maintenance manage- 
ment that: 
1. Tries to maximize overall equipment effective- 
ness and overall efficiency. 
442 
Journal of Manufacturing Systems 
Vol. 16/No. 6 
1997 
2. Develops a preventive maintenance program for 
the lifecycle of the equipment. 
3. Uses team-based concepts. 
4. Involves operators in maintaining the 
equipment. 
5. Uses motivational management (autonomous 
small groups) to promote preventive 
maintenance. 
In the TPM framework, the goals are to develop 
a "maintenance-free" design and to involve the par- 
ticipation of all employees to improve maintenance 
productivity. A metric, termed the "overall equip- 
ment effectiveness (OEE)," is defined as a function 
of equipment availability, quality rate, and equip- 
ment performance efficiency. 
If a company has an OEE of 85% or more, then 
it is considered to be a world-class company. TPM 
also provides a systematic procedure for linking 
corporate goals to maintenance goals. This proce- 
dure considers external and internal corporate 
environments, development of a basic maintenance 
policy, identification of key points for maintenance 
improvement, and finally the definition of target 
values for maintenance performance. The TPM 
metrics offer a starting point for developing quan- 
titative variables for relating maintenance measure- 
ment and control to corporate strategy. 
TPM focuses on eliminating major "losses" 
encountered in production activities. These losses 
are decomposed into major losses obstructing 
equipment efficiency, manpower (personnel) effi- 
ciency, and efficiency of material and energy. 
Based on their links to achieving corporate goals, 
targets to eliminate or reduce these losses are 
developed directly for the maintenance task in a 
production system. Just as in activity-based costing 
(ABC), where cost drivers are identified, the objec- 
tive here is to identify variables that can demon- 
strate causal relationships that maintenance activi- 
ties have on production system performance. Each 
of the major equipment losses are functionally 
related to availability, performance efficiency, 
and/or quality rate. Thus, the improvement from a 
maintenance system can be measured by its impact 
on the OEE. The concept of total productive main- 
tenance is gaining wider acceptance in industrial 
firms. 11 
The major challenge of industrial maintenance 
management research is to develop new models 
and methods that integrate both major trends. 
Trends in Maintenance Knowledge 
Maintenance modeling is inherently evolutionary 
in nature. As equipment complexity increases, and 
as the need for high equipment availability becomes 
paramount in today's complex, dynamic systems, 
there has been a corresponding increase in mainte- 
nance modeling sophistication. The idea of reac- 
tionary corrective maintenance progressed to prede- 
termined preventive maintenance, then to large- 
scale industrial maintenance, to condition-based 
maintenance determined by inspection, to expert 
maintenance systems, and now toward a futuristic 
view of intelligent or "self maintenance." 
Blanchard lz and Lyonnet ~3 provide overviews of 
the evolving maintenance categories. Corrective 
maintenance involves all unscheduled maintenance 
actions performed as a result of system/product fail- 
ure to restore the system to a specified condition. 
Corrective maintenance includes failure identifica- 
tion, localization and isolation, disassembly, item 
removal and replacement or repair in place, 
reassembly, and checkout and condition verification. 
Preventive maintenance includes all scheduled 
maintenance actions performed to retain a system or 
product in a specified condition. These actions 
involve periodic inspections, condition monitoring, 
critical item replacements, and calibration. 
Predictive maintenance is a relatively new con- 
cept in maintenance planning. This category of 
maintenance occurs in advance of the time a failure 
would occur if the maintenance were not performed. 
The time when this maintenance is scheduled is 
based on data that can be used to predict approxi- 
mately when failure will occur if certain mainte- 
nance is not undertaken. Data such as vibration, 
temperature, sound, color, and so on have usually 
been collected off-line and analyzed for trends. 
With the emergence and use of programmable 
logic controllers (PLCs) in production systems, 
equipment and process parameters can now be con- 
tinually monitored. With condition-based mainte- 
nance, the PLCs are wired directly to an on-line 
computer to monitor the equipment condition in a 
real-time mode. Any deviation from the standard 
normal range of tolerances will cause an alarm (or a 
repair order) to be automatically generated. 
443 
Journal of Manufacturing Systems 
Vol. 16/No. 6 
1997 
Installation costs for such a maintenance system can 
be high, but equipment service levels can be signif- 
icantly improved. 
Intelligent maintenance, or self-maintenance, 
involves automatic diagnosis of electronic systems 
and modular replacement units. ~4 Sensor data from 
remote facilities or machines would be provided on 
a continuous basis to a centralized workstation. 
From this workstation, the maintenance specialist 
could receive intelligent support from expert sys- 
tems and neural networks for decision-making tasks. 
Commands would then be released to the remote 
sites to begin a maintenance routine that may 
involve adjusting alarm parameter values, initiating 
built-in testing diagnostics, or powering standby or 
subsystems, for instance. The U.S. Federal Aviation 
Administration (FAA) is developing the Remote 
Maintenance Monitoring System (RMMS) that is an 
example of the future direction in maintenance 
automation. ~5 In some cases, robotics may be used 
for remote modular replacements. 
Emergence of New Maintenance Methods 
Developments in the area of artificial intelligence 
(AI) have led to the emergence of expert systems and 
neural networks. These solution techniques have 
found numerous applications in maintenance plan- 
ning. Milacic and Majstorovic 14 report on a survey that 
identified a list of 60 different expert maintenance sys- 
tems as of 1987. Frequently, the reasons for the use of 
expert systems in maintenance are the increasing com- 
plexity of equipment, the interdisciplinary nature of 
modern maintenance problems, the departure of main- 
tenance expertise from an organization due to retire- 
ments, the reduced training time of novice technicians, 
and consistently good decisions. ~6 Spur, Specht, and 
Gobler 17 discuss two general categories of expert 
maintenance systems: associative diagnosis and 
model-based diagnosis. In the former, conclusions are 
reached based on an analysis of fault possibilities that 
are verified by testing. The search tree uses codedknowledge from domain experts. In the latter, the real 
performance of equipment is compared with the sim- 
ulated performance of a computer model, and faults 
are inferred from the differences between the two. 
The applications of expert systems in mainte- 
nance are quite diverse. Representative industries 
include automotive, aerospace, electronics, process, 
computers, and telecommunications. CATS is an 
expert maintenance system with a knowledge base 
of 550 rules to detect sudden failures in diesel-elec- 
tric locomotive systems. IN-ATE is an expert system 
used for electronic circuit diagnosis. FSM is an 
expert system that Boeing uses for continuous con- 
dition monitoring of aircraft alarms. Lockheed 
developed RLA, an expert system for repair-level 
analysis for major parts in an aerospace system, is 
Bajpa119 uses an expert system architecture to trou- 
bleshoot general problems with machine tools in 
manufacturing industries. Bao z° develops an expert 
system to assist in the manufacturing and maintain- 
ability of surface-mount technology (SMT) printed 
circuit board (PCB) assembly. Khan et al) ~ discuss 
GEMS-TTS, an expert system used by AT&T main- 
tenance specialists to isolate faults in communica- 
tion links. Corn et al. 2z describe TOPAS, an expert 
system that diagnoses transmission and signaling 
problems in real time that may arise on switched cir- 
cuits. One of the most successful expert systems is 
"CHARLEY" which was developed by General 
Motors and is based on the knowledge of Charley 
Amble, an experienced maintenance engineer) 3 This 
expert system is used to diagnose problems with 
broken machine tools and to instruct less experi- 
enced individuals by providing explanations. GM 
reports that "CHARLEY" has reduced training costs 
by as much as $500,000 per year per plant. 
Although the idea of utilizing expert systems in 
maintenance held early promise, the use of rule- 
based programming has led to practical problems in 
implementation. For example, XCON, an expert 
system developed by Digital Equipment Corp. for 
product configuration, has more than 10,000 rules. 
Issues such as maintainability of the knowledge 
base, testability of the program, and reliability of the 
advice have limited the practical use of most expert 
systems in maintenance. 24 Other approaches, such as 
constraint-based reasoning, are being developed as 
alternatives to rule-based systems) 5 Also, the recon- 
sideration of Bayesian theory to support probabilis- 
tic reasoning and maintenance diagnostics is being 
reexamined. 26 
Neural networks are computing systems that 
incorporate a simplified model of the human neu- 
ron, organized into networks similar to those found 
in the human brain. 27 Instead of programming the 
neural network, it is "taught" to give acceptable 
results. The ability of artificial neural networks to 
444 
Journal of Manufacturing Systems 
Vol. 16/No. 6 
1997 
capture complex trends has been researched and 
documented in a significant number of research 
papers since the "rebirth" of neural networks in 
1982 when researchers rediscovered their important 
characteristics) s-3° The large number of research 
papers available on these characteristics prohibits 
their documentation here, but as an indication of 
their diverse cognitive powers, there have been 
applications of neural networks in varied areas from 
stock market price prediction and credit rating 
approval to engineering applications such as pat- 
tern/image recognition, digital signal processing, 
and automated vehicle guidance. 31 
Luxhoj and Shyur, 32 Luxhoj, Williams, and 
Shyur, 33 and Shyur, Luxhoj, and Williams 34 report on 
the use of artificial neural networks to capture and 
retain complex underlying relationships and nonlin- 
earities that exist between an aircraft's maintenance 
data and safety inspection reporting profiles. Neural 
networks will be used to implement condition-based 
maintenance because real-time sensor data can be 
trended to predict out-of-tolerance conditions for 
critical equipment parameters. Maintenance actions 
can then be initiated for an adaptive response to 
these anticipated system perturbations. An oil and 
gas company in Denmark is examining the use of 
artificial neural networks to predict the meter factor 
(pulses/unit volume) or k factor for turbine flow 
meters. By predicting the k factor in future time 
periods, significant deviations from the usable flow 
range can be anticipated so that maintenance techni- 
cians can make adjustments and prevent the expen- 
sive shutdown of a turbine for pumping oil or gas. 
Although the use of neural networks in maintenance 
will undoubtedly increase in the future, their solu- 
tion potential is constrained by current understand- 
ing of human reasoning capabilities and the limits of 
available computing power. 
Maintenance Improvement Connected to 
Organizational Learning 
In the quest toward world-class manufacturing, 
many industries are appreciating the need for effi- 
cient maintenance systems that have been effective- 
ly integrated with corporate strategy. New produc- 
tion management techniques, such as JIT and flexi- 
ble and agile manufacturing, demand quick response 
times and equipment availability. Riis and 
Neergaard 3s argue that there is a need to link manu- 
facturing planning to the perspectives of individual 
behavior, decision support, management systems 
and organizational structure, and corporate culture. 
These authors develop a multifactor "organizational 
learning model" as a new manufacturing paradigm. 
This model, illustrated in Figure 3, includes both 
formal and informal dimensions in a company. 
As new methods and technologies evolve for 
maintenance planning, such as expert systems, con- 
straint-based reasoning, belief networks, and artifi- 
cial neural networks, there is an important need to 
develop a broader framework to ensure that these 
methods become integrated into the modus operan- 
di of an organization. Otherwise, there exists the real 
possibility that the new methods and technologies for 
maintenance management may become "islands of 
automation" that will not provide any meaningful feed- 
/ / 
Management 
systems and 
organizational 
structure 
perspective 
Individual 
behavior 
perspective 
Corporate 
culture 
perspective 
Decision 
support 
perspective 
Individual Collective 
dimension dimension 
Formal 
dimension 
Informal 
dimension 
Decision 
support 
perspective 
Individual 
behavior 
perspective 
Management 
systems and 
organizational structure 
perspective 
Corporate 
culture 
perspective 
Figure 3 
35 Riis-Neergaard Model of Organizational Learning 
445 
Journal of Manufacturing Systems 
Vol. 16/No. 6 
1997 
back to improve organizational performance. New 
methods for reliability prediction and maintenance 
diagnoses coupled with the use of maintenance 
automation technology offer significant promise in 
helping to meet the global demands of competitive 
pricing, quality, and on-time deliveries. 
If a maintenance system is to have a successful 
implementation, the additional learning perspectives of 
the Riis-Neergaard model need to be considered during 
the system design process. A Danish-Norwegian sur- 
vey 36 reports that most enterprises tended to introduce 
technological means first and afterward adjust the 
imbalance by applying organizational means. 
To better ensure that technology, organizational 
development, and corporate strategy are coupled, the 
Riis-Neergaard model can be used to ask key questions 
during the maintenance system development process, 
such as: 
• What will be the consequences for the four types of 
learning by introducing the new maintenance sys- 
tem for a company? 
• Which changes are required in a company's man- 
agement systems, information technology, and 
organizational structure and culture as well as indi- 
vidual behavior to fully utilize the new mainte- 
nance system? 
Such an approachwill prevent the situation where a 
new maintenance system is developed in isolation and 
views the system development as an interactive, col- 
laborative process with due regard to the mutual inter- 
relationships to the other three learning perspectives. 
This integrative approach encourages the development 
of individual qualifications for maintenance techni- 
cians rather than relying extensively on the procedures 
embedded in a computerized maintenance manage- 
ment system. The approach also places awareness on 
collective learning processes as implemented by for- 
mal organizational structure and management systems 
for production planning and control and recognizes the 
role that informal systems play in operationalizing 
maintenance concepts. Such an understanding is essen- 
tial for a successful implementation of new mainte- 
nance models and methods that are more quantitative 
and technology-based. 
Case Examples of Maintenance 
Management in Danish Industry 
Case examples from Danish industry will be used 
[ 
Tasks ] ~-- 
\ 
People 
Technology 
I~ Structure 
/ 
J, 
Figure 4 
Leavitt's 37 Model of Organizational Change 
to illustrate different approaches to maintenance 
management. These descriptions focus on the "as is" 
condition of maintenance planning within these 
companies. The five primary maintenance cate- 
gories of corrective maintenance, preventive mainte- 
nance, predictive maintenance, condition-based 
maintenance, and intelligent or self-maintenance as 
defined earlier will be used to classify the mainte- 
nance management systems. It must be remembered 
that there will be overlap among the maintenance 
categories and that the primary purpose of this clas- 
sification scheme is to illustrate the distinctions 
among the different types of maintenance. 
To enrich the classification of maintenance man- 
agement systems in a study of three Danish compa- 
nies, Leavitt's modeP 7 will be used. As depicted in 
Figure 4, this model includes the factors of tasks, 
people, technology, and structure in examining orga- 
nizations. Leavitt's model is primarily used to under- 
stand the interdependencies of these four elements 
but will be included in this section to add more 
dimensions to the maintenance categories. Leavitt's 
sociotechnical model elements have been renamed 
as Maintenance Tasks, Structure and Management 
Systems, Technology, and People. Table 2 presents a 
summary of the three case studies. Management 
radar diagrams depicting the managerial effort in 
maintenance are presented for Companies B and C. 
Company A 
This Danish company produces high-end audio 
446 
Journal of Manufacturing Systems 
Vol. 16/No. 6 
1997 
Table 2 
Overview of Case Companies 
Organizational Structure Maintenance 
Company Maintenance Task and Management Systems Technology People Performance 
A --identical machines --generally, decentralized --varied technology --no training on how --maintenance is in a 
--simple technology maintenance --some registration of to use new systems development phase 
--manufacturer of -<luality products in a short - - in assembly, more time used for different --maintenance staff --90% of maintenance 
high-quality delivery time centralized and formal maintenance jobs has good knowledge jobs completed within 
audio/visual equipment --goal of no inventory of --hierarchical structure of existing systems 30 minutes 
- -8 factories with 7 finished goods --group leaders --lack of systematic --metrics needed 
assembly/production --maintenance is divided --primarily, corrective collection of --poor support from 
sections into mechanical and maintenance with some maintenance management 
--factory under study is electrical tasks preventive maintenance knowledge --plans to involve 
an assembly plant --informal inventory control of --good motivation of maintenance staff in 
spare parts maintenance staff development projects 
--poor use of information ---organic culture of 
system maintenance 
--informal planning of 
preventive maintenance 
B --accurate on-time JIT ----centrally organized --statistical process 
delivery --task divided into eontrol (SPC) used to 
--manufacturer of --production is in large operators, setting-up, measure machine 
industrial products batches and maintenance workers conditions and plan 
--factory under study --39 different product --20 maintenance workers preventive 
produces mechanical variants --mechanical/electrical maintenance and 
steering units for heavy --20 major nonparallel sections overhauls 
machinery and is machines --traditional centralized --central quality 
connected to the ----complex, CNC machines information system use function to take 
mobile hydraulic --average capacity to manage spare parts and measurements 
division utilization is 65% plan preventive maintenance --possibility of 
--high demands for -- log books at machines not outsourcing for 
accurate processes regularly used special measurements 
--strong demarcations 
within the 
maintenance 
organization 
--skilled staff 
normally qualified 
in a few machines 
--no coordination with 
production 
--no metrics 
--possible to follow up 
on maintenance costs 
in the EDP system 
--management focus is 
on capacity 
utilization 
C --pull-oriented demand --decentralized production --simple measuring 
--few processes and centers instruments 
--manufacturer of machines ~da i ly maintenance (vibrations, 
medical products --most machines are performed by production temperature, 
--factory under study is company-designed and operators resistance, etc.) 
a subsupplier to other constructed --maintenance staffreports to --special technology is 
factories in the --complex machinery plant manager outsourced 
company --high-quality demands --no EDP system 
from regulatory --manual planning of 
agencies preventive maintenance 
--only record completion 
of preventive 
maintenance jobs 
-- log books at machines 
--no economic metrics 
--education plan for 
each maintenance 
worker 
--~no cross training of 
skills 
--minimal 
coordination with 
production 
--no metrics 
- n o clear maintenance 
goals for production 
--no economic analysis 
of maintenance 
--10-15% of the time is 
used for preventive 
maintenance, less 
than 10% for 
corrective 
maintenance 
management focus 
views maintenance as 
a necessity, and one 
that should be 
minimized and 
invisible 
and video equipment, such as televisions, loud- 
speakers, CD players, tape recorders, and integrated 
music, audio, and video systems. Customers are 
located in upper social levels in numerous countries 
throughout the world. The company is divided into 
eight factories with seven assembly/production sec- 
tions and one administrative section. 
Table 2 provides a maintenance profile of Company 
A. There are a number of identical machines available. 
The technology in use involves simple assembly tools 
and fixtures in the audio section and integrated com- 
puter-controlled assembly lines in the video section. 
The philosophy of the production control system is to 
be able to deliver customer-specified products in a 
short delivery time, while at the same time to have no 
inventory of finished goods. Maintenance is divided 
into mechanical and electrical tasks. 
Maintenance is decentralized and informal. 
However, at the assembly factory, maintenance is 
somewhat centralized both formally and informally 
and placed as a subfunction known as a "production 
service" function. There is some hierarchical inter- 
nal maintenance where maintenance is divided into 
a mechanical section and an electrical section. 
The assembly factory is characterized by almost 
all breakdown or corrective maintenance. Rare pre- 
ventive maintenance focuses on checkups when 
machinery is not in use. Maintenance tasks involve 
447 
Journal of Manufacturing Systems 
Vol. 16/No. 6 
1997 
simple changing of familiar parts and lubrication to 
complicated searches for faults. Occasionally, eas- 
ily changedmodules are prepared. There are infor- 
mal rules of how to prioritize maintenance jobs, 
and informal rules to fix "flow/output critical" 
problems first. 
Possibilities to coordinate spare parts inventory 
are not used because the inventory control of spare 
parts is mainly informal. The information system is 
poorly used. It only collects time spent on each job. 
Rarely is the job type and cause of the breakdown 
recorded. There is informal planning of preventive 
maintenance; breakdown jobs are carried out at 
once if there is free capacity. The maintenance 
technology in use varies just as much as the pro- 
duction technology. There is some registration of 
repair times used at different jobs. 
There is no systematic training on how to main- 
tain new systems. The new maintenance staff is 
trained by older members, is well skilled in different 
maintenance jobs, and possesses a good knowledge 
of existing systems. However, new systems are not 
properly introduced, which causes problems. 
The poor use of information systems results in 
lack of collection of knowledge. The knowledge 
from the maintenance staff is not used by other cor- 
porate functions. The culture may be characterized 
as organic; that is, the maintenance staff suspects 
other functions regard them as a necessary evil. 
The maintenance staff desires to change this per- 
ception. Sometimes an attitude of laissez faire can 
be detected among the maintenance staff. 
The maintenance work at Company A is still in a 
development phase. The usage of information sys- 
tems is still poor. The maintenance staff is general- 
ly motivated to be more involved in various tasks 
and motivated to supply other functions with 
knowledge. There exists poor coordination between 
maintenance and production planners/leaders as 
well with systems development functions. Ninety 
percent (90%) of the maintenance jobs are com- 
pleted within 30 minutes. 
The maintenance staff is well aware of the 
importance of its own function and would like to 
see a company-wide recognition of their role. 
Metrics are needed to prove relevance of the 
maintenance function. There is poor support of 
the maintenance function from management. 
Plans for the future involve establishing metrics 
and involving the maintenance staff in various 
development projects. 
Company B 
As profiled in Table 2, Company B is divided into 
divisions within different product areas. The factory 
that was studied is connected to the mobile hydraulic 
division that produces mechanical steering units for 
heavy machinery. The case only covers a part of the 
production process and the description below con- 
centrates on this part. 
The factory under study is a subsupplier to 
assembly factories within the tractor industry, which 
requires accurate and on-time JIT delivery accord- 
ing to customer demand given by their assembly 
plan. The production philosophy is production to 
buffer stock (push) in the production area with three 
weeks of throughput time. The production is in large 
batches and the process flow can nearly be charac- 
terized as a production line. After the analyzed area, 
there is a pull-oriented assembly area with produc- 
tion directly to customer orders. The product pro- 
gram consists of approximately 39 different product 
variants based on nearly the same components, with 
minor differences. 
In the analyzed part of the factory there are 
approximately 20 major machines, with almost no 
parallel machines. The machines can be character- 
ized as complex CNC machines, transfer lines, and 
machine centers of which the major part of the 
machines were implemented in the early 1970s. The 
machines are normally running in three shifts; some 
machines run in four and five shifts. The average 
capacity utilization is approximately 65%. Due to 
the product tolerances and specifications, there are 
high demands for accurate processes. 
Maintenance is organized centrally under the pro- 
duction manager and can be characterized as mech- 
anistic-oriented. The maintenance tasks are divided 
into operators, setting-up workers (fitters), and 
maintenance workers. The operators should perform 
daily cleaning, but this does not work well. The fit- 
ters only perform minor maintenance tasks, but 
there are some problems with qualifications and 
coordination. Maintenance workers perform the 
remainder of the jobs. 
There are approximately 20 employees in the 
maintenance function (covering the whole factory). 
Both the maintenance tasks and the workers are 
448 
Company B 
Present Managerial Effort, Maintenance ManaGement 
Maintenance "products" Maintenance "products" 
Figure 5 
Radar Diagram Showing Managerial Effort for Company B 
divided into mechanical and electronics. 
Approximately 30% of the maintenance tasks are 
preventive maintenance, the rest corrective mainte- 
nance. The maintenance jobs are given a priority 
depending on the urgency. Each worker is responsi- 
ble for his or her own machine. 
The company has a traditional centralized infor- 
mation system for maintenance. The system is used 
for managing the stock of spare parts and for plan- 
ning preventive maintenance. The information sys- 
tem is used to coordinate the stocks of spare parts 
across the factories in the company. In addition, 
there are log books placed at each machine, but they 
are not used regularly. 
The maintenance function uses SPC (statistical 
process control) to measure the machine conditions 
and the need for preventive maintenance or major 
overhauls. A central quality function takes calibra- 
tions and minor machine tests (such as oil tests), 
which are carried out regularly. 
There is poor, informal coordination in planning 
between production and maintenance. There are no 
performance measures connected to daily mainte- 
nance, and the only measurement connected to the 
machine conditions is SPC measurement. In the 
EDP (electronic data processing) system, it is pos- 
sible to periodically follow up on maintenance 
costs by machine. The management focus on 
Company B 
Present Managerial Effort, Maintenance ManaGement 
Maintenance 
resources 
Journal of Manufacturing Systems 
Vol. 16/No. 6 
1997 
Maint~ 
Maintenanc 
work metho 
Maintenance 
materials 
Mainte 
Internal relations 
Figure 6 
Radar Diagram Showing Managerial Effort Compared to 
Goals for Company B 
maintenance is based on capacity utilization in the 
production area. 
Maintenance Managerial Effort 
Figure 5 illustrates the main results from the 
Management Radar Profile Analysis for Company 
B. The figures portray a managerial effort that is 
very unequal in each of the 12 main maintenance 
control areas. Much effort is made in the economic 
and organizational area (right side), but effort is very 
weak in the technological areas (left side). The 
analysis places focus on a weak point: How is the 
company able to control the maintenance economy 
when the efforts are so weak in controlling resources 
and materials? Also, the work methods for mainte- 
nance may be weak (that is, little specification of 
maintenance work methods). Figure 6 shows the 
managerial effort profile with indication of how dif- 
ficult the maintenance task is in relation to each of 
the 12 control areas. Overall, the managerial efforts 
in the right-side control areas are too strong com- 
pared with the actual maintenance tasks. Internal 
relations seem to have too low an effort. Figure 7 
shows the managerial effort profile by indicating the 
importance of success factors related to each of the 
12 control areas. Two control areas might be critical: 
quality of maintenance and internal relations. In 
both areas, the managerial efforts are low compared 
449 
Journal of Manufacturing Systems 
Vol. 16/No. 6 
1997 
Company B 
Present Managerial Effort, Maintenance Management 
Maintenance "products" 
Mainl ;onomy 
Maintenar~ ntenance 
work methc ;onomy 
Maintenance 
resources 
Maintenance 
structure 
Maintenanc(laintenance 
materials rganization 
Maint( 
Internal relations 
Figure 7 
Radar Diagram Showing Difficult Management Areas for Company B 
with the importance that the company places on suc- 
cess factors related to these two areas. 
Company C 
Company C manufactures products for the hospi- 
tal sector, such as protheses and other devices to 
assist people recovering from operations. The case 
study as presented in Table 2 focused on the factory 
that produces parts for the other factories in the 
company. 
The production philosophy in the factory is pull- 
oriented demand from the other factories, with 
approximately two weeks of buffer stock on most 
products. There are local stocks of finished goods in 
each sales company. 
The production process consists of few processes 
and machines. The first part of the production is 
characterized as a production line with production in 
large batches. The last part of production involves 
two minor process centers/machines. Most 
machines are company-designed and constructed. 
The production equipment is complex and sensible 
according to quality, process parameters, and toler- 
ances. Production occurs in three shifts during the 
first three days in the week. The remaining two days 
are used for maintenance and other tasks. 
The organizational culture is very organic. The 
company has a mission to minimize documentation 
and formal business processes. The production is 
organized as local production centers. Many support 
functions are integrated into the production centers. 
Daily maintenance tasks are assigned to a mainte- 
nance operator at each production center. The main- 
tenance function is organized as a staff function to 
the plant manager. The function has seven employ- 
ees, of which one to three are working with mainte- 
nance. The rest are working with building new 
machines and construction changes on existing 
machines. The preventive maintenance jobs use one 
employee. Each maintenance worker has an assigned 
machine/production area of responsibility. 
An operator at each production center has the 
responsibility for maintenance and changeover. 
This operator has a technical education and per- 
forms minor tasks. The company uses no EDP- 
based information system for mintenance. Some 
PC-based programs are used in planning preventive 
maintenance, but most plans are made manually. 
The only recording of maintenance is that the pre- 
ventive maintenance jobs have been completed. 
Some of the registrations are made in a log book at 
each machine. Maintenance (repairs) have first pri- 
ority, and a new project has second priority. There 
are no economic valuations of each machine's 
importance to production. 
The maintenance function has only simple mea- 
suring instruments, for vibrations, temperature, resis- 
tance, and so on. All special technology is hired from 
external service companies. 
To ensure an educated maintenance staff, there is 
an education plan for each maintenance worker. The 
education is connected to the tasks in the production 
area where the worker is responsible for mainte- 
nance. Maintenance workers cannot operate the dif- 
ferent production machines and do not have any 
knowledge of the production flow. 
There is no formal cooperation between mainte- 
nance and production planning. There are no mea- 
sures---only that the preventive maintenance jobs 
have been completed. One of the reasons is that the 
maintenance function does not have clear goals relat- 
ed to production. Also there is no follow-up on main- 
tenance costs. Most maintenance costs (salaries) are 
hidden in the production centers and in new projects. 
All machines have the same ranking according to 
maintenance. Ten to 15 percent (10% to 15%) of the 
time is used for preventive maintenance, less than 
450 
10% for corrective maintenance. The remaining time 
is used for projects. Management focus is that the 
maintenance function is a necessity but needs to be 
minimized and invisible. 
Maintenance Managerial Effort 
The managerial effort for Company C is depicted 
in Figure 8. The main efforts are in the following 
Company C 
Present Managerial Effort, Maintenance Management 
Maintenance "products" 
Maintenance 
resources 
Journal of Manufacturing Systems 
Vol. 16/No. 6 
1997 
Mainl 
Maintenan( 
work meth~ 
Maintd 
Maintenanc~ 
materials 
Figure 8 
Radar Diagram Showing Managerial Effort for Company C 
Company C 
Present Managerial Effort, Maintenance Management 
Maintenance "products" 
Maintenance 
resources 
Maint 
Maintenanc 
work methc 
Maintenance 
resources 
Maintenanc{ 
materials 
Maint( 
Internal relations 
Figure 9 
Radar Diagram Showing Managerial Effort Compared to 
Goals for Company C 
control areas: maintenance organization, mainte- 
nance products, internal relations, and maintenance 
structure. Other areas might be unbalanced (mainte- 
nance quality, maintenance resources, and mainte- 
nance materials) with no connection between the 
level with goals and strategies and the level with 
recording. Figure 9 shows the managerial effort pro- 
file with success factors related to the 12 control 
areas. Some equality of importance of the success 
factors seems to exist (indicated in small circles), 
but the managerial efforts differ for the 12 control 
areas. The reason for this is not clear, calling for fur- 
ther analysis. Figure 10 shows the managerial effort 
profile with factors that may give difficulties when 
managing the maintenance tasks. 
General Observations from 
Case Studies 
A number of general observations may be made 
from analyzing these three case studies of Danish 
companies. Many of the observations are consistent 
with the maintenance survey results contained in the 
EUREKA 5 report. The general observations are 
summarized below: 
• An analysis of the case studies suggests that for- 
mal systems will deteriorate if not meaningfully 
reinforced. Maintenance workers must be moti- 
vated and top management must see a clear link 
Company C 
Present Managerial Effort, Maintenance Management 
Maintenance "products" 
Main1 
Maintenanq 
work meth~ 
Maintenanc, 
materials 
Maint( 
Internal relations 
Figure 10 
Radar Diagram Showing Difficult Management Areas for Company C 
451 
Journal of Manufacturing Systems 
Vol. 16/No. 6 
1997 
between maintenance activities and corporate 
objectives. The case studies reveal that the main- 
tenance sections were not involved in production 
systems development. This observation is con- 
sistent with the EUREKA report, which indicated 
a general lack of integration between the mainte- 
nance and production functions. 
• The case studies reveal a lack of maintenance 
resources in the companies. As is frequently the 
case, the concern for the execution of daily 
operations diminishes the capability for long- 
range maintenance planning. Also, the mainte- 
nance planning and supervision functions were 
mixed. As Wireman 2 observes, such a situation 
leads to more supervising than planning. In the 
companies studied, the motivation of production 
personnel to perform basic maintenance was 
weak. This observation is also consistent with 
the EUREKA report. 
• A key observation emerging from the case stud- 
ies is that the knowledge base for maintenance 
was not systematically organized or structured. 
This organization of maintenance knowledge is 
a prerequisite for improving maintenance per- 
formance. All the case companies were strug- 
gling with the quantification of maintenance 
metrics. This observation is very consistent with 
the EUREKA survey. 
• The case companies were interested in develop- 
ing differentiated repair/replacement strategies 
for specialized electrical and mechanical equip- 
ment. The case studies illustrate the need to bet- 
ter integrate maintenance planning with the pro- 
duction function and highlight the need for a 
broader framework for maintenance systems 
development. The survey also illustrates theneed to develop maintenance systems that may 
be used with different organizational s~uctures 
and may include combinations of centralized 
and decentralized maintenance activities as well 
as outsourcing of certain maintenance functions. 
The fundamental managerial implications for 
maintenance from analyzing the case companies 
may be succinctly stated as follows: 
A situational approach to maintenance systems 
development is needed due to varying company 
dynamics, products, processes, technologies, 
and market influences. 
• The maintenance function needs to be integrated 
with production systems development, organiza- 
tional structures, people, and information tech- 
nology issues. 
• Current maintenance systems have poor links to 
corporate strategy, which leads to a deterioration 
of formal systems due to lack of meaningful 
reinforcement. 
• The knowledge base in a company for mainte- 
nance is typically not well organized, structured, 
or current. 
Concluding Remarks 
Despite advances in computer technology and 
manufacturing techniques, benchmarking studies of 
actual maintenance performance signal the need for 
new, improved methods for analyzing and designing 
maintenance systems that are consistent with novel 
developments in information technology and man- 
agement systems. This paper highlighted some of 
the managerial implications for future maintenance 
management by examining recent trends in mainte- 
nance methods, knowledge, organization, and infor- 
mation systems. 
References 
1. British Standard 3811 (1984). 
2. T. Wireman, World Class Maintenance Management (Industrial Press, 
1990). 
3. P.M. Senge, The Fifth Dimension (Random House, 1990). 
4. U. Thorsteinsson, "Importance of Maintenance" (Denmark: Institute 
of Production Management and Industrial Engineering, Technical 
University of Denmark, 1995). 
5. EUREKA: European Benchmark Study on Maintenance, EBSOM-EU 
724 (1993). 
6. K. Holmberg and A. Folkeson, eds., Operational Reliability and 
Systematic Maintenance (Elsevier Applied Science, 1991). 
7. U. Thorsteinsson and C. Hage, "Maintenance Management Profiles 
for Industrial Systems," Operational Reliability and Systematic 
Maintenance, K. Holmberg and A. Folkeson, eds. (Elsevier Applied 
Science, 1991), pp283-303. 
8. W.J. Fabrycky and B.S. Blanchard, Life-Cycle Cost and Economic 
Analysis (Englewood Cliffs, NJ: Prentice-Hall, 1991). 
9. D.E. Whitney, "Manufacturing by Design," Harvard Business Review 
(July-Aug. 1988). 
10. S. Nakajima, Introduction to Total Productive Maintenance (TPM) 
(Productivity Press, 1988). 
11. M. Tajiri and E Gotoh, TPM Implementation: A Japanese Approach 
(NewYork: McGraw-Hill, 1992). 
12. B.S. Blanchard, Logisfics Engineering and Management (Englewood 
Cliffs, NJ: Prentice-Hall, 1992). 
13. P. Lyonnet, Maintenance Planning: Methods and Mathematics 
(Chapman & Hall, 1991). 
14. V.R. Milacic and J.E Majstorovie, "The Future of Computerized 
Maintenance," Diagnostic and Preventive Maintenance Strategies in 
Manufacturing Systems, V. Milacic and J. McWaters, eds. (North Holland, 
1988). 
452 
Journal of Manufacturing Systems 
Vol. 16/No. 6 
1997 
15. J.T. Luxhej, C.J. Theisen, and M. Rao, "An Intelligent Maintenance 
Support System (IMSS) Concept," Proceedings of the 26th Symposium of 
the Society of Logistics Engineers (Fort Worth, TX, Aug. 1991 ), pp 117-126. 
16. R.G. Bowerman and D.E. Glover, Putting Expert Systems into 
Practice (Van Nostrand Reinhold, 1988). 
17. G. Spur, D. Specht, and T. Goblel, "Building an Expert System for 
Maintenance," Diagnostic and Preventive Maintenance Strategies in 
Manufacturing, V. Milacic and J. Majstorovic, eds. (North-Holland, 1988). 
18. J,E Hayes, E BeBalogh, and E. Turban, "Lockheed Aeronautical 
Systems Company's Program for a Competitive Edge in Expert Systems for 
Logistics," Expert Systems for Intelligent Manufacturing, M. Oliff, ed. 
(North-Holland, 1988). 
19. A. Bajpul, "An Expert System Model for General-Purpose 
Diagnostics of Manufacturing Equipment," Manufacturing Review (vl, n3, 
1988), pp180-187. 
20. H.P. Bao, "An Expert System for SMT Printed Circuit Board Design 
for Assembly," Manufacturing Review (vl, n4, 1988), pp275-280. 
21. N. Khan, P. Callahan, R. Dube, J. Tsay, and W. van Duesen, "An 
Engineering Approach to Model-Based Troubleshooting in Communication 
Networks," IEEE Journal on Selected Areas in Communications (v6, n5, 
1988), pp792-799. 
22. P. Corn, R. Dube, A. McMichael, and J. Tsay, "An Autonomous 
Distributed Expert System for Switched Network Maintenance," 
Proceedings of the IEEE Global Telecommunications Conference (1988), 
pp46.61-46.68. 
23. A.B. Badiru, Expert System Applications in Engineering and 
Manufacturing (Englewood Cliffs, N J: Prentice-Hall, 1992). 
24. X. Li, "What's So Bad About Rule-Based Programming?" IEEE 
Software (1994), pp 103-105. 
25. H.J. Skovgaard, "A New Approach to Product Configuration," 
Technical Working Paper (Struer, Denmark: Bang & Olufsen Technology 
A/S, 1994). 
26. E. Horvitz, C. Ruokangas, S. Srinivas, and M. Barry, "Project Vista: 
Display of Information for Time-Critical Decisions," Proceedings of the 4th 
Rockwell International Conference on Control and Signal Processing 
(Anaheim, CA, Jan. 1992). 
27. P.D. Wasserman and T. Schwartz, "Neural Networks, Part 1: What Are 
They and why Is Everybody So Interested in Them Now?" IEEE Expert 
(v2, n4, 1987), ppl0-13. 
28. W S. McCulloch and W Pitts, "A Logical Calculus of Ideas Immanent 
in Nervous Activity," Bulletin of Mathematical Biophysics (v5, 1943), 
pp115-133. 
29. J.J. Hopfield, "Neural Networks and Physical Systems with Emergent 
Collective Abilities," Proceedings of the National Academy of Science 
(v79, 1982), pp2554-2558. 
30. J.J. Hopfield, "Neurons with Graded Response Have Collective 
Computational Properties Like Those of Two-State Neurons," Proceedings 
of the National Academy of Science (v81, 1984), pp3088-3092. 
31. A.J. Maren, C.T. Harston, and R.M. Pap, Handbook of Neural 
Computing Applications (Academic Press, 1990). 
32. J.T. Luxhnj and H. Shyur, "Reliability Curve Estimation for Aging 
Helicopter Components," Reliability Engineering and System Safety (v48, 
1994), pp229-234. 
33. J.T. Luxhnj, T.E Williams, and H. Shyur, "Comparison of Regression 
and Neural Network Models for Prediction of Inspection Profiles for Aging 
Aircraft," IIE Transactions on Scheduling and Logistics (v29, n2, 1997), 
pp91-101. 
34. H. Shyur, J.T. Luxh~j, and T.P. Williams, "Using Neural Networks to 
Predict Component Inspection Requirements for Aging Aircraft," 
Computers and Industrial Engineering (v30, n2, 1996), pp257-267. 
35. J.O. Riis and C. Neergaard, "The Learning Company: A New 
Manufacturing Paradigm," Proceedings of the Annual CIM Europe 
Conference (Copenhagen, 1994). 
36. J. Frick, E Gersten, P.H.K. Hansen, J.O. Riis, and H. Sun, 
"Evolutionary CIM Implementation--An Empirical Study of 
Technological-Organizational Development and Market Dynamics," 
Proceedings of the 8th CIM Europe-Annual Conference (Birmingham, UK, 
1992). 
37. H.J. Leavitt, "Applied Organizational Change in Industry: Structural, 
Organizational, and Humanistic Approaches," Handbook of Organizations, 
J.G. March, ed. (Rand McNally, 1965). 
Authors' Biographies 
James T. Luxhoj is an associate professor of industrial engineering at 
Rutgers University. He received his BS, MS, and PhD in industrial engi- 
neering and operations research from Virginia Polytechnic Institute and 
State University. He serves as a department editor for IIE Transactions and 
an associate editor for the Journal of Engineering Valuation and Cost 
Analysis and Engineering Design and Automation. Professor Luxh~j's 
research interests include production economics, maintenance and logis- 
tics, and decision support systems. He is the recipient of a 1989 SAE Ralph 
R. Teetor Award for Engineering Education Excellence. His work has 
appeared in several publications, including lie Transactions on Schedulingand Logistics, International Journal of Production Research, International 
Journal of Production Economics, International Journal of Operations and 
Production Management, International Journal of Quality and Reliability 
Management, and Computers and Industrial Engineering. 
Jens O. Riis holds a master's degree in mechanical engineering from the 
Technical University of Denmark and a PhD in operations research from 
the University of Pennsylvania. Professor Riis has a chair in industrial man- 
agement at the engineering school of the University of Aalborg (Denmark) 
and has published articles and books in the field of project management, 
industrial management, design of production management systems, and 
management of technology. He is currently heading a Danish government 
supported research program on integrated production systems. Professor 
Riis is a member of the Academy of the Technical Sciences in Denmark and 
is a member of the editorial board and referee for the International Journal 
of Project Management and Production Planning & Control. 
Uffe Thorsteinsson is an associate professor of production management 
and industrial engineering at the Technical University of Denmark. He 
obtained his MS in chemical engineering from the Technical University of 
Denmark. Professor Thorsteinsson is participating in a Danish government 
supported research program in integrated production systems, His research 
interests include maintenance management systems and project manage- 
ment. He has published numerous articles and chapters in books in the 
fields of plant design, maintenance, logistics, and project management. He 
is vice president of the West European Group of University Teachers for 
Industrial Management. 
453