Available online at www.sciencedirect.

com

ScienceDirect
Procedia CIRP 30 (2015) 480 485

7th Industrial Product-Service Systems Conference - PSS, industry transformation for sustainability and
business

Planning of maintenance activities A current state mapping in industry

Maheshwaran Gopalakrishnan a*, Jon Bokrantza, Torbjrn Ylipa, Anders Skoogha
a
Chalmers Univeristy of Technology, Gteborg 41296, Sweden
* Corresponding author. Tel.: +46 72 963 6517. E-mail address: mahgop@chalmers.se

Abstract

Industrial Product Service System (PSS) thinking can be applied to production system by considering it as a product. Prior
studies show that strategic planning of the maintenance activities in manufacturing industries holds great potential to increase
productivity. Planning of maintenance activities is therefore an integral decision making aspect for maintenance engineers and it
is important to analyze how industries are currently working with planning of maintenance activities and what additional support
is needed. This paper aims at mapping the current state of the work procedures for maintenance engineers and planners in the
industry and analyzes the gap from current practices to the strategic planning which could increase productivity. The study
specifically focuses on how industries work today with finding critical resource, performing criticality analysis, and planning
maintenance. A descriptive research approach is followed, where empirical data is collected in Swedish industry through three
different data collection methods. The results show the state-of-art industrial practices and the gaps in maintenance planning.
2015 The Authors.
Authors. Published
Published by
byElsevier
ElsevierB.V.
B.V This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the International Scientific Committee of the Conference is co-chaired by Prof. Daniel Brissaud & Prof.
Xavier BOUCHER.
Peer-review under responsibility of the International Scientific Committee of the 7th Industrial Product-Service Systems Conference - PSS,
industry transformation for sustainability and business
Keywords: maintenance planning; criticality analysis; system bottleneck

1. Introduction 5% [4]. Therefore maintenance activities planning is an

integral decision making aspect for maintenance engineers,
Industrial Product Service Systems (PSS) offers business requiring support from modern methodologies, data analysis
innovation and sustainable development for industries by approaches, and Information and Communication Tools
integrating production and service for their products. (ICT). Currently, the maintenance department in industries on
Manufacturing industries have complex production systems the contrary use limited tools and analyses to assist their
producing different products. In order to produce products of decision making on an everyday basis [5].
high quality the production system needs maintenance [1]. Critical sections of the production system should be
However, overall equipment effectiveness (OEE) in effectively utilized. Dynamic decision support is needed for
manufacturing companies is about 50 55% in manufacturing maintenance and existing maintenance management systems
industries [2]. These production losses are due to direct down are insufficient [6]. There are different ways in which critical
time (failures) and system losses (blocked and idle states) of sections of the system could be classified. Failure mode effect
machines in flow-oriented production system. These result in and criticality analysis (FMECA) is the frequently method [7].
economic sustainability losses to the company. Ecologic Throughput criticality classification can be used for decision
sustainable losses occur as 30% energy losses are due to support system for planning of maintenance tasks [8] and an
system losses [3]. By considering the production system as a analytic hierarchical process (AHP) based [9] can also be used
product, the PSS thinking can be employed to the production for maintenance. Criticality analysis needs to be continuously
system. This will make the production system highly updated every day to prioritize maintenance activities [10]
productive, sustainable, and reliable. With complex production, prioritization of maintenance
A prior study shows that through strategic planning of work-orders becomes crucial and challenging [11].
maintenance activities, productivity can be increased by about Throughput improvement can be achieved through

2212-8271 2015 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the International Scientific Committee of the 7th Industrial Product-Service Systems Conference - PSS,
industry transformation for sustainability and business
doi:10.1016/j.procir.2015.02.093
 Maheshwaran Gopalakrishnan et al. / Procedia CIRP 30 (2015) 480 485 481

prioritizing maintenance activities for the static and dynamic 2.1. Survey
bottleneck machines [4]. The highest criticality is given to the
equipment that is most important for a specific purpose, Quantitative data was collected in Swedish industry
which normally is production. The equipment with the highest through the use of a web-based questionnaire. Invitation to the
criticality gets the highest priority code and is thus scheduled questionnaire was sent by e-mail to selected respondents, and
first when performing maintenance [1]. Effectiveness is an open invitation was listed publicly on the website of
achieved through prioritizing machines criticality, and Sustainability and Maintenance Global Centre (SMGC), as
focusing on specific components [9]. well as included in an SMGC e-mail newsletter. SMGC is a
Hence there is strong motivation to understand the use of non-governmental maintenance organization with over 50
criticality classification and bottleneck detection in the member companies. A non-probabilistic judgement sample
industries in relation to planning of maintenance. In order to was used [12], where the primary target group were
do that, a current state mapping of how companies currently maintenance or production experts.
working with planning of maintenance activities and the 62 out of 82 selected respondents answered, resulting in a
additional needed support is necessary. Therefore the authors response rate of 75 percent. The open invitation resulted in 22
formulate the following research questions (RQ): additional responses. Out of the total 84 submissions, non-
RQ1: To what extent are companies working with experts were excluded, and the respondents with the highest
criticality classification? management level were chosen at plant-level for each
Finding the extent to which companies work with company. The final selection consisted of 76 responses from
criticality classification is an important starting point as this 71 companies, where the 5 duplicates represent individual
will help in identifying the critical resource of the system, as respondents from different plants within the same company,
in RQ2. but separated geographically and operating with different
RQ2: What is criticality from a maintenance perspective, management. A majority of the respondents can be classified
and how are critical resources identified? as the maintenance department. The companies represent
Criticality classification can be created in many different various production contexts such as manufacturing, energy,
ways and from different perspectives. Finding the critical nuclear, paper and food industries. The questionnaire covered
resource from a maintenance perspective will help in the topics of criticality, bottleneck detection, and maintenance
prioritizing maintenance activities, as in RQ3. prioritization. The remaining part of the questionnaire covered
RQ3: To what extent are maintenance activities prioritized, other areas such as production disturbances, tools and
and how are the criticality classification used for this purpose? methods in maintenance etc.
Maintenance activities for production system needs
effective planning. This paper will identify the extent to 2.2. Maintenance Fair
which maintenance activities are prioritized and the use of
criticality classification for the same. Seven structured interviews were conducted during one
day of Scandinavias largest maintenance fair, which was held
2. Methodology during 4 days in March in Gothenburg with 250 participating
companies. These interviews were short (less than 10
Three mixed method research questions were stipulated, minutes), and focused specifically on the topics of tools and
with the intent of increasing knowledge about how criticality methods used in maintenance planning, use of priorities,
and bottleneck detection is used from a maintenance criticality, and bottleneck detection. The interview questions
perspective in industry. A descriptive survey research were formed as a combination of closed questions with
approach was adopted [12, 13], aiming to provide additional multiple choices and open-ended question.
information about the use of these practices in industry, where
the three questions serves to explore and explain the current 2.3. Interviews
situation. The three data collection methods were used to form
empirical evidence to answer the research questions. Four semi-structured face-to-face interviews [14] were
Quantitative data was collected using a web-based conducted with personnel of the maintenance department
questionnaire survey and structured interviews during a from two of the partner companies in the research project
maintenance fair, and a combination of quantitative and StreaMod. Three maintenance managers and one
qualitative data was collected using semi-structured maintenance strategist were selected as interviewees since
interviews. Throughout the paper, the three data sets will be they represent high strategic level within large multi-national
referred to as the survey, maintenance fair, and corporations, thus indicating a specific context that could
interviews. The three data collection methods were chosen benefit from using bottleneck and criticality analysis in
in order to investigate the subject area from both a general and maintenance. The interview template was created on the basis
a specific perspective. The survey and the maintenance fair of the previous two data collection methods, and covered the
describe the general perspective since it was collected from topics of criticality and bottlenecks. The interviewees
both small and large companies in various industrial branches received information regarding the covered topics prior to the
and production contexts. In contrast, the interviews depict a interviews. The concepts were not explained in further detail
specific perspective since they were conducted in two of at this point, thus assumed to be familiar to the interviewees.
Swedens largest discrete manufacturing companies. The interviews were structured to first ask about the critical
482 Maheshwaran Gopalakrishnan et al. / Procedia CIRP 30 (2015) 480 485

resources in the production system, followed by how they In connection to the establishing of criticality levels,
work with criticality classification and bottleneck detection Figure 2 displays to what extent these levels are continuously
from a maintenance perspective. Prompts and probes were updated. It shows a similar distribution, where less than 30%
used to increase the clarity of the answers, which led to that of the companies continuously update the criticality levels to a
some answers related to critical resources developed towards relatively high or very high degree, and 65% to a relatively
discussions about bottlenecks. low degree or not at all.
A mixed method approach was also used within these
interviews, where open-ended questions were asked in 60% 36% 29%
combination with a structured questionnaire. Through the use 40% 20%
of 4-point Likert scales, the interviewees where instructed to 20% 8% 8%
assess the certainty regarding critical resources and 0%
To a very To a To a Not at all ?, N/A
bottlenecks. An additional probe followed with the intent to high degree relatively relatively
further clarify the reasoning of the chosen answer. high degree low degree

2.4. Data analysis and presentation

Fig.2. Updating of criticality levels
Following the data analysis methodology proposed by [13]
and adhering to the descriptive research approach, It is however important to note the results if the two
quantitative data were analysed using SPSS and presented questions regarding establishing and updating of criticality
with descriptive statistics in terms of frequency of responses. levels are cross-tabulated. The 4 main Likert-scale variables
For qualitative descriptive research, [15] advice to perform are recoded into 2 categories, High degree, and Low
content analysis, where the outcome is a descriptive summary degree, and the Do not know, Not applicable, and
of the data organized in a way that best fits the data. missing answers are excluded. A Fishers exact test on the
Following the criteria described by [16], the interview data recoded 2x2 matrix shows a two-tailed P value = 0.0001. This
was chosen to be analysed by hand, and presented narratively data indicates that a majority of the companies that are
with identified themes and illustrating quotes. Throughout the working with establishing of criticality levels to a high degree
paper, the data from the four interviews are referred to in also consider that they continuously update the criticality
terms of the order they were conducted, i.e. data from levels to a high degree, and vice versa.
Interviewee 1, 2, 3, and 4. The results from the three data Furthermore, Figure 3 shows the survey data regarding to
collection methods are presented in chronological order in what extent companies are working with identifying and
terms of each stipulated research question, starting with the analysing bottlenecks in production. The data indicates that a
survey and finishing with the interviews [17]. majority of the companies, 56% in total, work with bottleneck
detection to a relatively high or very high degree, and less
3. Results than 10% do not work with it at all.

All the three collected data are presented in this section. 60% 39%
Note that ? refers to the answer alternative Do not know, 33%
40% 17%
and N/A refers to the combination of the alternative Not 20% 8% 3%
applicable and missing answers. 0%
To a very To a To a Not at all ?, N/A
high degree relatively relatively
3.1. RQ1 high degree low degree

Survey: Figure 1 illustrates to what extent the companies

Fig 3. Bottleneck detection.
participating in the survey are establishing criticality levels
for maintenance of processes, equipment, or components. It Maintenance fair: 4 out of 7 respondents indicated that
shows that 35% of the companies work with criticality they work with bottleneck detection, as well as mentioned that
classification to a relatively high or very high degree, and it is detected by Value Stream Mapping (VSM), operation
55% to a relatively low degree or not at all. monitoring, OEE, availability, throughput, and buffers.
Interviews: All four interviewees indicated that they work
60% 34% with establishing equipment criticality levels. Interviewee 1
40% 26% 21% explained that they create equipment priority numbers, which
20% 9% 9%
are created and updated when installing new machines or
0% scrapping old ones. In contrast, interviewee 2, 3, and 4
To a very To a To a Not at all ?, N/A
high degree relatively relatively indicated that the classification is a part of a technical
high degree low degree specification carried out during the early project phase. When
asked about bottleneck detection; all interviewees stated that
Fig.1. Establishing of criticality levels. it is performed. Interviewee 1 analyses both short term and
long term bottlenecks based on data and meetings. Daily
shop-floor meetings are used to decide short term bottlenecks,
 Maheshwaran Gopalakrishnan et al. / Procedia CIRP 30 (2015) 480 485 483

and long term bottlenecks are discussed on managerial level. classification model with a tree-structure of questions, where
Number of stops, short stops, and long stops are data the answers to redundancy, safety, productivity, environment,
considered for bottleneck detection. Interviewee 2 answered etc. generate the criticality level .When asked what the critical
that their detection is based on facts and data. Interviewee 3 resource of the production system is, the answers were
said that the production engineers are responsible for random at first. After prompting, interviewee 1 talked about
bottleneck detection, not the maintenance department and an old robot in a particular line as being critical. When
interviewee 4, said they use VSM to detect bottleneck specifically asked why, the answer was that the robot was a
conducted by production and maintenance personnel. bottleneck and it was throughput critical. However,
interviewee 1 also mentioned that the critical measure from a
3.2. RQ2 maintenance perspective is availability. Interviewee 2 said
Its really difficult for me to point out one that is critical []
Critical for me is focus. I mean its not always the machines.
Survey: Table 1 indicates that the most common primary
Its the people around it. However, when questioned further,
basis for establishing criticality levels is using an ABC-
it was indicated that bottlenecks could be critical, but the
classification. Furthermore, it shows that constructing the question of the bottleneck being the true bottleneck was
levels on other basis, such as bottleneck analysis, is not used raised. Interviewee 3 answered that if you look at the
to the same extent. Note that the respondents could only assembly line from an overall perspective, then that is very
choose one alternative. critical. If we get a stop here, it always affects the end
customer directly. They have A classified all the machines in
Table 1. Basis for establishing criticality levels.
that particular line. Interviewee 4 answered The layout is
Primary basis for criticality levels n %
very unfortunate and the parts move back and forth making it
ABC-classification 23 30% hard to understand the flow and analyse the losses. On
Operator influence 8 11% further questioning, interviewee 4 also mentioned bottlenecks
Bottleneck analysis 7 9% as critical and explained that a bottleneck machine is A
Cost-based priority 5 7%
classified.
Time of purchase 4 5%
80% 50% 75%
Other basis 9 12% 60%
40% 25% 25% 25%
Do not know/ N/A /Missing answer 20 26% 0% 0% 0%
20%
0%
Very certain Somewhat Not very Not at all
If answering other basis, the respondents were asked to certain certain certain
clarify the answer. These 9 respondents commented that their
criticality levels are established on the following basis: Critical resource Bottleneck
demand controlled, from the business system, cost of
root cause category, similar to ABC but with 1-5
Fig. 4. Interviewees perception on critical and bottleneck resources
classification where 5 is highest, product mix,
RCM/FMECA, reactor safety (nuclear), safety, Figure 4 shows the data from the Likert-scale follow up
equipment with importance for nuclear safety. questions, and it indicates that the interviewees are not always
Maintenance fair: In relation, when asked about how they certain about their bottlenecks and critical resources.
establish criticality levels, all the respondents indicated that Enquiring about bottlenecks in particular, Interviewee 1
they establish using ABC-classification. Apart from that, 2 talked about known and unknown bottlenecks, short term and
respondents answered operator influence and bottleneck and 1 long term bottlenecks, and says there are many bottlenecks in
answered for cost, upon purchase, and machine health as the different areas. Interviewee 2 talked about true bottlenecks.
basis of criticality levels. Note that the respondents could They were certain where their bottlenecks were but not
choose more than one answer alternative. convinced at all time. Interviewee 3 talked about moving
Interviews: When asked how they work with criticality bottleneck and said they have a good control over their
classification, all four interviewees said that their company
bottleneck. Interviewee 4 said that they were rather certain,
work with an ABC-type classification. Interviewee 1
but that maintenance and production can have different views
mentioned that they use a 1 5 numbered equipment priority
about bottlenecks since they measure them differently.
code for the classification, and the other three uses exactly
ABC-classification. When the interviewees were asked how
the criticality levels are set, they answered differently. 3.3. RQ3
Interviewee 1 said that the equipment priority codes are
dependent on the production set-up, e.g. single or parallel Survey: Figure 5 shows that a majority of the companies
flow. Similarly, interviewee 4 explained that they have a prioritize maintenance work orders. In detail, 67% of the
general thinking about the ABC classification where A respondents indicated that this is done to a relatively high or
single line machine, B a parallel machine, and C spare very high degree; whilst 25% answered that it is only done to
machine. Interviewee 2 explained that establish the criticality a relatively low degree or not at all. However, this data do not
classification on a component level through a risk analysis, connect to how the priorities are set.
which is based on fault frequency, mean time between failure
(MTBF), and its consequences. Interviewee 3 uses a
484 Maheshwaran Gopalakrishnan et al. / Procedia CIRP 30 (2015) 480 485

60% 46% have pretty good awareness of the equipment, and they know
40% 21% 20% whats critical and not. So thats pretty much how we control
20% 5% 8%
0% and plan. Instead, they use it more for managing the
To a very To a To a Not at all ?, N/A equipment and try to make them less critical. Interviewee 3
high degree relatively relatively explains: we find a way to attack our already critical
high degree low degree
equipment, make them less critical and the most important.
The interviewees were also asked on how bottleneck
Fig. 5. Prioritizing work orders.
detection is used for maintenance planning. Only interviewee
1 clearly described how bottlenecks are directly used for this
Maintenance fair: When questioned how reactive and
purpose. It was explained that they have identified different
preventive maintenance work orders are prioritized 4 out of 7
long term bottlenecks for which they have decided to spend
answered fixed priorities, 2 out of 7 answered continuously
more maintenance hours, and that their ABC alarms are based
updated priority, only one answered first-come-first-served
on the current short term bottlenecks in the factory. The
basis, and one did not answer the question.
VSMs that are done according to interviewee 4 are used to
Interviews: When asked in general about the priorities they
ensure that they are spending maintenance work in the right
set for their maintenance activities, interviewee 1 descried that
place. However, the role of the maintenance department in
an equipment priority routine is followed in their company by
resolving bottlenecks varies depending different issues.
meeting with people from different departments. It is also
Interviewee 2 said, they do not work with bottlenecks directly
noted that the maintenance engineer sets the equipment
for maintenance planning, and instead emphasised that they
priority based on the routine that is created in the meetings
are focusing on what is necessary for their line. Interviewee 3
along with different things. Interviewee 2 explained that they
explains that bottlenecks are more production related, and
decide on a department priority in a factory meeting, where
thus not performed by the maintenance department.
the priority is fixed and based on what is crucial for us right
now. Interviewee 4 says that the logistics department sets a
4. Discussion
plant or line priority based on a delivery perspective.
When asked in particular how reactive maintenance is
RQ1: Criticality classification is used for different
prioritized, all interviewees said they are largely situation
maintenance related purposes [7, 9]. Instantaneous bottleneck
dependent. They react to alarms, and sometimes morning
detection which updates bottleneck every day is used in
meeting results in deviations from the priority plan.
maintenance decision support system [8]. However, the data
Interviewee 1 said that the maintenance technician combines
from the survey and maintenance fair showed that few
how the situation looks in the factory with the equipment
companies work with criticality classification for maintenance
priority number, and then sets the priority. They also have
purposes to a high degree. Criticality analysis needs to be
bridges with alarm operators, where the severity of the alarm
continuously updated every day from maintenance
decides if an operator is called. Interviewee 3 also said they
perspective [10], and the companies working with criticality
have similar alarms, but similarly indicated that the person
classification also update their classification to a high degree.
who creates the work order finally decides which machines
Majority of the companies works with bottleneck detection.
get what alarm Interviewee 1s statement sums this up: For
All the interviewees who belong to major organizations told
reactive maintenance work orders, it is up to each
that they worked with criticality classification and bottleneck
maintenance technician to prioritize.
detection. This lays the foundation to find what critical
Regarding preventive maintenance priorities, interviewee 1
resource of the system is (RQ2) is.
clarified that they combine the equipment priority a type of
RQ2: There are different methods in which criticality
maintenance priority to prioritize preventive maintenance
classification of resources can be done. There are system
(e.g. preventive maintenance for safety reasons vs. availability
value based, AHP based [9], and FMECA based [7] criticality
reasons). In contrast, interviewee 2, 3, and 4 answered
analyses. All the methods are not based on equipment level.
typically: We have special windows within production where
Instead it is from an overall systems perspective including all
we stop the production. Interviewee 2 said that they focus on
equipment in the system. In this study, all three data sources
getting all planned maintenance done during the allotted time.
indicate that criticality classifications are mostly established
Interviewee 3 indicated that they use a 24 hour / 1 week / 2
by an ABC-type classification. From the interviews, which
week priority for allocating the work during the windows, but
were big industries, the classification found to be based on
clarified yet again that it is the person who schedules the work
production layout such as parallel flow, single flown as well
who sets the final priority. Interviewee 4 said they find the
as redundancy, safety, productivity, environment, etc. Again
activity that are closest in time and then work upwards.
these are not equipment specific strategies but a system level.
From three of the interviewees, it is observed that the
Despite this, the term critical resource was ambiguous initially
criticality classification is not directly used for prioritization
to all the participants. There was no clear connection with
of maintenance activities, despite the fact that the criticality
criticality classification. Only on further explanation with
levels for the equipment are printed on each work order. The
examples, the interviewees mentioned what they perceive
priorities are instead set based primarily on the personnels
critical from maintenance perspective. Two of them
own experience and knowledge. Interviewee 3 exemplifies
mentioned bottlenecks could be perceived critical. Evidently
this: If we use the criticality classification for prioritizing?
the interviewees were not 100% certain about which were
Hmm, I dont know The people who are running around
 Maheshwaran Gopalakrishnan et al. / Procedia CIRP 30 (2015) 480 485 485

their critical and bottleneck resources. However, the term critical resource is vague in industries at the moment,
maintenance activities need prioritization (RQ3). and not connected to the critical classification. Most industries
RQ3: Prioritizing maintenance work orders is crucial for prioritize maintenance activity. However, criticality
handling product variety [11]. Maintenance operation classification and bottleneck detection are seldom used to
efficiency is improved by prioritizing tasks [6]. The survey make priorities.
and maintenance fair also showed that vast majority of the
companies prioritize maintenance work orders. The Acknowledgements
interviewees as well agreed on some sort of prioritization for
preventive and reactive maintenance work orders. According The authors thank the StreaMod research project, its partners,
to [1], equipment with the highest criticality should get the and funding agency VINNOVA for their support.
highest priority code and thus be scheduled first when
performing maintenance. However in companies interviewed References
the priority for the work order doesnt come from criticality
classification or the critical resource. Instead the priorities are [1] Smith, R. & Hawkins, B. 2004, Lean Maintenance: Reduce costs, Improve
set by the person who creates work orders. The criticality Quality, and Increase Market Share, Elsevier, Amsterdam; Boston.
[2] GoodSolutions AB .2012. Good Solutions customer survey. Presented at
classifications are used for asset management purposes. the FFI production research conference (In Swedish): Mtesplats fr
Interviewee 1s work procedure was a good example as they framtidens framgngsrika verkstder, Katrineholm, Sweden, 22-24 May
use the criticality classification in setting priorities for 2012.
maintenance work orders. Also interviewee 1 strongly pointed [3] Skoogh, A., B. Johansson, and L. Hansson. 2011. Data Requirements
out the connection between priorities and bottlenecks. The and Representation for Simulation of Energy Consumption in Production
Systems. In: Proceedings of the 44th CIRP Manufacturing Systems
results of all 3 research questions are summarized in table 2. 2011.
The results show the gap in industries practices with [4] Gopalakrishnan, M., Skoogh, A., and Laroque, C. 2013: Simulation-
respect to criticality classification and maintenance planning. Based Planning Of Maintenance Activities In The Automotive Industry
Even in literature the connection between them are not always In: Proceedings of the 2013 Winter Simulation Conference.
clear. One reason could be the criticality classification is not [5] Bokrantz, J., Skoogh, A., and Ylip, T. 2014. Lean principles and
engineering tools in maintenance organizations a survey study, paper
made for planning. There is also difference between criticality presented to the Swedish Production Symposium, Gothenburg.
classification and the term critical resources. Dynamic [6] Ni, J., and X. Jin. 2012. Decision support systems for effective
decision support is needed for effective maintenance and it maintenance operations. In: CIRP Annals - Manufacturing Technology
can be achieved through criticality classification and using it 61:411414
for planning maintenance. Effective planning will lead [7] Bevilacqua, M., and Braglia, M. 2000. The analytic hierarchy process
applied to maintainence strategy selection. Reliability Engineering and
towards productivity and reliability for the production system. System Safety, 70 (2000), pp 71-83.
[8] Lin, L., and N. Jun. 2009. Short-term decision support systems for
Table 2. Summary of results. maintenance task prioritization. International journal of production
RQ Result economics, 121 (2009), pp 195 -202.
[9] Labib, A.W., OCorror, R.F., and Williams, G.B. 2006. An effective
RQ1 According to the survey and maintenance fair, majority of the
maintenance system using the analytic heirarchy process. Integrated
companies work with bottleneck detection to high degree. Companies
Manufacturing System, Vol. 9 Iss 2 pp. 87 98.
do not work with criticality classification to a high degree, however
[10] Moore, W.J., Starr, A.G. An intelligent maintenance system for
the ones using it updates it. All 4 companies from interview data
continuous cost-based prioritisation of maintenance activities. Comput
worked with criticality classification and bottleneck detection.
Ind [special issue on e-maintenance] 2006;57(6):595606.
RQ2 According to all three data sources, the most common criticality [11] Guo, W., Jin, J., and Hu, S.J. 2013. Allocation of maintenance resources
classification is ABC-type. During interviews the term critical in mixed model assembly systems. Journal of Manufacturing Systems,
resource resulted in random answers, and not everyone was sure 32 (2013), pp 473 479.
what are critical and bottleneck resources. [12] Forza, C. 2002. Survey research in operations management: a process-
RQ3 According to the survey and maintenance fair, the vast majority based perspective, International Journal of Operations & Production
prioritize maintenance work orders, and all interviewed companies Management, vol. 22, issue 2, pp. 152-194.
does it. From the interviews, prioritization of reactive maintenance is [13] Owneugbuzie, A.J., and Leech, N.L. 2006. Linking Research Questions
largely situation-dependent, criticality classification is not always to Mixed Method Data Analysis Procedures. The Qualitative Report, vol.
used to prioritize maintenance, and bottlenecks are not extensively 11, no. 3, pp. 474-498.
used in maintenance work. [14] Tong, A., Sainsbury, P., Craig, J. 2007. Consolidated criteria for
reporting qualitative research (COREQ): a 32-item checklist for
interviews and focus groups, International Journal for Quality in Health
5. Conclusion Research, vol. 19, no. 6, pp. 349-357.
[15] Sandelowski, M. 2000. Whatever happened to Qualitative Description?.
Research in Nursing & Health, issue 23, pp. 334-340.
This paper describes the current state practices of [16] Auld, G.W., Diker, A., Ann Bock, M., Boushey, C.J., Bruhn, C.M.,
manufacturing industries towards planning maintenance Cluskey, M., Edlefsen, M., Goldberg, D.L., Misner, S.L, Olson, B.H.,
activities. That includes finding critical resource, how are they Reicks, M., Wang, C., and Zaghloul, S. 2007. Development of a
classified, and the gap between criticality classification and Decision Tree to Determine the Appropriateness of NVivo in Analyzing
maintenance planning in industries. Overall, criticality Qualitative Data Sets, Journal of Nutritional Education Behaviour, vol.
38, pp. 37-47.
classification is not extensively used in industry however [17] Mertens, D.M. 2011. Publishing Mixed Methods Research. Journal of
bottleneck detection is widely used. The industries that use Mixed Method Research, vol. 5, pp. 3-6.
criticality classification have an ABC-type approach. The