Sistema de Gestão de Manutenção

Prévia do material em texto

Maintenance Decision Making, Supported By Computerized 
Maintenance Management System 
Ali Rastegari, Volvo GTO 
Mohammadsadegh Mobin, Ph. D., Western New England University
Key Words: Computerized Maintenance Management System , Multiple Criteria Decision Making Data Clustering
SUMMARY & CONCLUSIONS 
This paper is written based on the need for 
Computerized Maintenance Management System’s
(CMMS) decision analysis capability to achieve world 
class status in maintenance management. Investigations 
indicate that decision analysis capability is often missing 
in existing CMMSs and collected data in the systems are
not completely utilized. How to utilize the gathered data 
to provide guidelines for maintenance engineers and 
managers to make proper maintenance decisions has 
always been a crucial question. In order to provide 
decision support capability, the aim of this paper is to 
provide and examine three different decision making 
techniques which can be linked to CMMS and add value 
to collected data. This research has been conducted within 
a global project in a large manufacturing site in Sweden to 
provide a new maintenance management system for the 
company. The data from the main studies were collected 
through document analysis complemented by discussions 
with maintenance engineers and managers at the case 
company to verify the data. Methods including a Multiple 
Criteria Decision Making (MCDM) technique called 
TOPSIS, k-means clustering technique, and one decision 
making model borrowed from the literature were used. 
The results indicate the most appropriate maintenance 
decision for each of the selected machines/parts according 
to factors such as frequency of breakdowns, downtime, 
and cost of repairing. The paper concludes with a 
comparison of results obtained from the different decision 
making techniques and also a discussion on possible 
improvements needed to increase the capability of the 
maintenance decision making models.
1. MAINTENANCE TYPES
Maintenance may be performed through various 
actions, and there are various classifications of 
maintenance types [1]. One classification of maintenance 
types and their relationships is indicated in the Swedish 
standard [2] in which maintenance is divided into two 
main actions, corrective and preventive. In various studies 
in the maintenance literature, such as [1], [3] and [4], the 
term “type” has been used similarly to other terms, such 
as “approach”, “action”, ”strategy” and “policy”. 
Corrective maintenance is also known as run-to-
failure or reactive maintenance and is a strategy that is 
used to restore (repair or replace) equipment to its 
required function after it has failed [4]. It is sometimes 
used synonymously with Breakdown Maintenance (BM), 
Failure-Based Maintenance (FBM) or Operation-To-
Failure (OTF) [1], [5] and [6]. 
Preventive maintenance can be predetermined 
(periodic) maintenance or Condition-Based Maintenance 
(CBM) [2]. The Swedish standard defines predetermined 
maintenance as follows: “Preventive maintenance carried 
out in accordance with established intervals of time or 
number of units of use such as scheduled maintenance but 
without previous item condition investigation” (p.15). It is 
sometimes used synonymously with Time-Based 
Maintenance (TBM) or Fixed-Time Maintenance (FTM), 
[3] and [6]. The Swedish standard defines CBM as 
“preventive maintenance based on performance and/or 
parameter monitoring and the subsequent actions” (p.15).
According to Kobbacy et al. [7], Design-Out 
Maintenance (DOM) can be considered as another 
maintenance policy in which the focus is to improve the 
design of production equipment to make maintenance 
easier or even eliminate it. Ergonomic and reliability 
aspects are important in this policy. Labib [6] considers 
Skill-Level Upgrade (SLU) as another policy for 
maintenance that is used to improve competence of 
operators. 
In addition, various concepts have been developed to 
increase the effectiveness of maintenance and focus on 
the maintenance activities. The two more common 
concepts are Reliability-Centered Maintenance (RCM), 
and Total Productive Maintenance (TPM). Moubray [8] 
defines RCM as “…a process used to determine what 
must be done to ensure that any physical asset continues 
to do what its user wants it to do in its present operating 
Authorized licensed use limited to: Universidad Federal de Pernambuco. Downloaded on February 21,2022 at 02:01:00 UTC from IEEE Xplore. Restrictions apply. 
context” (p.7). A Japanese concept for maintenance, 
according to Nakajima [9] who introduced the concept; 
TPM may be defined as “Productive maintenance 
involving total participation” (p.10).
2. MAINTENANCE DECISION MAKING
Maintenance decision making involves assessing and 
selecting the most efficient maintenance approach (i.e., 
strategies, policies, methodology or philosophy) [5]. It
involves determining the most appropriate maintenance 
policy to take, such as BM, TBM or CBM. The 
consequences of an inefficient maintenance policy go far 
beyond the direct costs of maintenance. BM is a strategy 
that leads to high levels of machine downtime (production 
loss) and maintenance (repair or replacement) costs due to 
unexpected breakdown [10]. A preventive maintenance 
strategy contributes to minimizing failure costs and 
machine downtime (production loss) and increasing 
product quality [11]. However, the TBM practice is not 
usually applicable when attempting to minimize operation 
costs and maximize machine performance [3]. Marquez 
[12] also states the reason that the maintenance plans 
provided by the equipment manufacturer are not 
completely reliable is that they are not aware of 
“business-related consequences of failure, safety 
considerations, regulatory requirements, the use of 
condition monitoring techniques, availability of resources 
and unique environmental conditions” (p.16). This 
statement is supported by Tam et al. [13], who note that 
preventive maintenance intervals based on Original 
Equipment Manufacturer (OEM) recommendations may 
not be optimal because actual operating conditions may 
be very different from those considered by the OEM. As 
such, actual outcomes may not satisfy company 
requirements. In addition to BM and TBM, according to 
Gupta and Lawsirirat [14], the main goal of CBM is to 
perform a real-time assessment of equipment conditions 
to make maintenance decisions, consequently reducing 
unnecessary maintenance and related costs. Rastegari and 
Bengtsson [15] emphasize that reducing the probability of 
maximal damage in production equipment and reducing 
production losses, particularly in high production 
volumes, are two potentially significant benefits of a 
proper CBM implementation.
The achievement of more efficient maintenance 
depends on the capability of the implemented 
maintenance policy to effectively provide and employ 
relevant information concerning the factors that affect the 
life of the component/equipment in question [5]. 
Providing more relevant information on component 
condition increases the ability (effectiveness) of a 
maintenance solution to avoid failures and makes the best 
possible use of the equipment/component’s effective life 
by performing replacements ‘‘just’’ before failure; thus, 
this information improves the maintenance policy’s 
accuracy [5]. Jantunen et al. [16] propose a guide for 
maintenance decision making based on component failure 
models. According to Figure 1, in case of wear models D, 
E and F, the use of CBM is not possible or sensible as 
failures can take place without a warning being registered 
by the measuring signals. In such a case, the best solution 
is to run the component until failure occurs; hence, the 
optimal maintenance type is BM. When infant mortality is 
high (A and F), TBM is not an appropriate option. Cases 
A, B and C can be monitored, but is not sensible to 
monitor the remaining three (D, E and F).
Figure 1 - Failure models,adopted from [16]
3. COMPUTERIZED MAINTENANCE MANAGEMENT
SYSTEM (CMMS)
The increased amount of information available and a 
growing need to have this information at hand and in real 
time for decision making necessitates a CMMS to aid 
maintenance management [6]. Legacy maintenance 
systems with large batch reports in which the focus was 
on data throughput are being replaced by dynamic, on-
line queries created on-the-fly with answers in seconds 
rather than days [6]. The CMMS can provide the 
following items [6]:
Support CBM
Track the movement of spare parts
Allow operators to report faults faster
Improve communication between operations and 
maintenance personnel
Historical information necessary for developing 
preventive maintenance schedules
Provide maintenance managers with information to 
have better control over their departments
Offer accountants information on machines to enable 
capital expenditure decisions to be made
Data collection and data analysis are more or less 
offered by commercially available CMMS packages, but 
CMMS packages have always been lacking decision 
analysis. “This lack of decision support is a definite 
problem; because the key to systematic and effective 
maintenance is managerial decision making that is 
appropriate to the particular circumstances of the 
machine, plant or organization” [6], (p.193). 
4. RESEARCH METHODOLOGY
The purpose of this paper is to provide models that 
can be linked to CMMS in order to add value to data 
collected in the system by providing decision making 
Authorized licensed use limited to: Universidad Federal de Pernambuco. Downloaded on February 21,2022 at 02:01:00 UTC from IEEE Xplore. Restrictions apply. 
capability. The empirical basis for the study was a single 
case study of a major manufacturing site in Sweden. The 
case company’s products are gearboxes, with a production 
volume of 95 000 products per year. This research has 
been conducted within a global project at the company to 
provide a new CMMS. The data was collected through 
document analysis complemented by discussions with 
maintenance engineers and managers at the case company 
to verify the data. Various documents, including 
Emergency Work Orders (EWO) database, maintenance 
audits, maintenance strategies and maintenance activities 
at the case company were studied. The focus of the 
research was on physical assets at the company 
(machines, motors, pumps...) and buildings, services and 
software were excluded.
The findings, presented below, are focused on the 
maintenance decision making according to key factors 
found in the case study, followed by a discussion on the 
applied decision making models.
5. MAINTENANCE DECISION MAKING MODEL 
FROM THE CASE COMPANY
Two decision making models found at the case 
company are considered in this study. One is illustrated in 
Figure 2, which is based on production output.
Figure 2 - Maintenance decision making flow chart based 
on production output
The other model is based on the technical viewpoint,
which is indicated in Figure 3.
Figure 3 - Maintenance decision making flow chart based 
on technical viewpoint
In addition, there are guidelines for selecting 
appropriate maintenance policies at the company. Here, 
for example, are points to choose CBM:
When a breakdown takes place, it will considerably 
influence production output
Irregular breakdowns often take place
Breakdowns cannot be prevented by a normal 
inspection and servicing
The time interval between two sequential
breakdowns is relatively long
Breakdowns often take place because of deterioration
The progress of deterioration is relatively slow
6. APPLICATION OF MAINTENANCE DECISION 
MAKING GRID IN THE CASE COMPANY
The model proposed by Labib [6] is called the 
Decision Making Grid (DMG) and is used in this section. 
The model is based on identifying the criteria of 
importance including downtime, frequency of failures and 
cost. The DMG then proposes different maintenance 
policies based on the state in the grid, which is indicated 
in Figure 4.
Figure 4 - DMG, adopted from [6]
The first step that must be taken to make a decision 
making grid is criteria analysis [6]. The aim of this phase 
is to establish a Pareto analysis of two important criteria:
downtime and frequency of breakdowns. Downtime and 
frequency can be substituted by Mean Time To Repair 
(MTTR), and Mean Time Between Failures (MTBF). The 
Authorized licensed use limited to: Universidad Federal de Pernambuco. Downloaded on February 21,2022 at 02:01:00 UTC from IEEE Xplore. Restrictions apply. 
purpose is to assess how bad the worst performing 
machines are for a certain period of time. 
The machines at the case company are classified in 
AA, A, B and C levels, which is called the machine 
ledger. In this classification, AA machines are the most 
critical machines. Appendix A.1 indicates criteria analysis 
on data collected from EWO database and maintenance 
engineers for AA machines. It shows that 53 percent of 
the total downtimes of the company during the six months 
of the year in addition to 54 percent of the number of 
breakdowns are due to AA machines failures. The worst 
performing machines in both criteria are sorted and 
grouped into High, Medium, and Low sub-groups. It is 
obvious that machine number 8527784 and 87843 are the 
worst machines by having highest downtime and 
frequency of breakdowns. However, doing criteria 
analysis needs to have consideration on different aspects
such as analytic hierarchy process of faults related to the 
machine system and components, as well as performing 
more mathematical analysis to reach an accurate scope for 
each level.
The next step is to place the machines in the DMG
(Figure 4) to recommend asset management decisions [6].
By locating machines in to the decision making grid 
(decision mapping), the model indicates which 
maintenance policy should be selected for each machine 
(Figure 5). For instance, machine number 87843, which 
has the highest downtime and the highest frequency of 
failure, in the top-right region, is the worst performing 
machine and the action to implement, or the rule that 
applies is DOM, accordingly the machine design should 
be modified. On the other side, in the bottom-left region, 
machine number 8528609 that has the lowest downtime 
and frequency of failures can work until a failure happens 
(BM). In the top-left region, SLU is the most appropriate 
policy, high frequency of breakdowns with low downtime 
show that a machine has been stopped many times for 
limited periods of time. Therefore, maintaining this 
machine is a rather easy task that can be performed by 
operators after upgrading their skill levels. Machines for 
which their performances are located in the bottom-right 
region are problematic machines. The low frequency of 
breakdowns shows that the machines don’t breakdown 
often, but their high downtime presents that each 
breakdown can last for long time due to a big failure. 
Therefore, the appropriate action for these machines is 
CBM to analyze the breakdowns and monitor the 
machines’ conditions. If a machine is located in a region 
with a medium downtime or a medium frequency the 
appropriate policy to take is TBM. However, sometimes 
two machines are exactly the same and they are doing the 
same work in the same environment but they are located 
in different grids. Maintenance engineers should therefore 
do more analysis according to maintenance concepts such 
as TPM and RCM in order to select the appropriate 
maintenance policy. In addition, in some cases two 
machines are located near or in the border of two different 
grids. Maintenance concepts can be helpful in these cases
to decide which decision to make. For example, when 
downtime is high but the frequency of the failure is low, 
performing some analysis such as RCM can be considered 
to decrease downtime.
Figure 5 - Decision making grid for AA machines at the 
case company
The model proposed in this sectioncan be a solution 
to provide decision making analysis capability for the 
system by adding value to data collected in CMMS. For 
having a more logical model and using data for selecting 
maintenance policy, cost function must be considered. For 
this reason, the company needs to have historical cost data 
such as cost of failures for production and cost of 
maintenance. By considering cost and performing criteria 
analysis, the model will be in three dimensions in a fuzzy 
surface and each region indicates which maintenance 
policy should be selected (Figure 6). In this model the 
assumption is that the cost function of maintenance 
policies is linear and trails the relationship as follows: 
DOM > CBM > SLU > TBM > BM.
Figure 6 - The fuzzy decision surface showing the regions 
Authorized licensed use limited to: Universidad Federal de Pernambuco. Downloaded on February 21,2022 at 02:01:00 UTC from IEEE Xplore. Restrictions apply. 
of different strategies, adopted from [6]
In addition to this, the criteria analysis that has been 
performed in this section is at the machine level. In order 
to make a more accurate decision making model, this 
criteria analysis should be performed at the component 
level. And, the level of faults in the Analytical Hierarchy 
Process (AHP) must be prioritized and analyzed 
according to the components.
7. APPLICATION OF MCDM AND CLUSTERING 
TECHNIQUE IN THE CASE COMPANY 
In this section, the logic of the DMG (in Figure 4) is 
borrowed from Labib [6] and is used in combination with 
two mathematical methods to rank and categorize the case 
company’s machines and parts in their relevant 
maintenance policy groups.
1.1 TOPSIS
The most widely used MCDM tool called TOPSIS 
[17], is used in this section to rank the machines and
parts. The basic mechanism to this approach is to 
calculate the distance from each alternative to a Positive-
Ideal Solution (PIS) and a Negative-Ideal Solution (NIS) 
that are defined in n-dimensional space, where n
represents the number of criterion in the decision 
problem. The chosen alternative should have the smallest 
vector distance from the PIS and the greatest from the 
NIS. The TOPSIS algorithm presented in [18] and [19]
are utilized in this paper. Using TOPSIS method, 
machines are ranked based on three criteria including 
downtime, cost and frequency. For the sake of simplicity, 
the weights of criteria are considered as equal. Based on 
maintenance experts’ opinions, the ranked machines are 
divided into five categories and are presented in Figure 7 
for visualization purposes. According to these categories, 
five different maintenance policies can be considered. 
The configurations of each category are summarized 
as Appendix A.2. The total number of machines is equal 
to 540.
In order to have more investigation on data, we also 
ranked the parts based on three criteria as mentioned 
above. Applying TOPSIS method with equal weights for 
criteria, the ranks of parts are obtained (summarized in 
Appendix A.3). All parts are divided into five categories 
as presented in Figure 8:
Figure 7 - Ranked machines based on TOPSIS using 3 
criteria
Figure 8 - Ranked parts based on TOPSIS using 3 criteria
Based on the category of the machines and parts, 
maintenance engineers or managers can more easily
decide which decision is the most appropriate. For 
example a machine in the first rank which has the highest 
downtime with the highest frequency and cost, is a
problematic machine. Therefore, CBM can be a good 
action to take. However, it still needs performing more 
analysis such as failure analysis before implementation.
1.2 K-means clustering technique
The k-means clustering technique [20] is used to 
cluster the machines based on three criteria. The 
algorithm was run for a number of different means and 
the optimal number of clusters was determined to be five
according to the calculated Silhouette Index [21]. Each 
cluster is represented by a single representative, which 
reflects the characteristics of the cluster. 
Table 1 - Five clusters based on 5 downtime and 
frequency
Clusters DT Fr
1 (BM) Low (0-100) Low (1-5)
0
100
200
300
400
500
0
5
10
15
20
25
30
0
5
10
15
x 10
4
DowntimeFrequency
Co
st
Rank:1-100
Rank:101-200
Rank:201-300
Rank:301-400
Rank:401-540
0
100
200
300
400
0
2
4
6
8
10
0
1
2
3
4
5
6
x 10
4
Downtime
Frequency
Co
st
Rank:1-500
Rank:501-1001
Rank:1001-1500
Rank:1501-2000
Rank:2001-2878
Authorized licensed use limited to: Universidad Federal de Pernambuco. Downloaded on February 21,2022 at 02:01:00 UTC from IEEE Xplore. Restrictions apply. 
2 (CBM) High (100-500) Low (1-5)
3 (TBM) Medium (100-200) Medium (5-15)
4 (SLU) Low (0-100) High (6-23)
5 (DOM) High (100-500) High (6-30)
The representatives (Centroids) are obtained from the 
data based on two criteria of different maintenance policy 
(Figure 4) [6]. Based on Figure 4, the data for machines 
are divided into five categories as presented in Table 1. 
The cost criterion is also considered in the clustering 
algorithm, which is assumed as the average cost of 
machines in each cluster. The representatives of each
cluster are presented in Table 2.
Table 2 - K-means clusters representatives (Centroids)
Clusters DT Fr Cost
1 (BM) 16.55 1 4555.41
2 (CBM) 173.99 3 14509.26
3 (TBM) 141.58 7 39036.57
4 (SLU) 50.62 9 20755.24
5 (DOM) 266.4382 16 68378.02
The result of the clustering technique is presented in 
Figure 9. This figure indicates that each cluster can be 
assigned to a specific maintenance policy.
Figure 9 - Clustered machines based on k-means 
clustering technique
8. CONCLUSIONS
The aim of this paper was to provide and examine 
three different decision making techniques that can be 
linked to CMMS and add value to collected data. Methods 
including DMG, TOPSIS and clustering techniques 
borrowed from the literature were used in the case 
company. As a result of this study, the necessary 
information for proper maintenance decision making and 
the decision making models are identified, and the details
of utilizing them are described. The results indicate the 
most appropriate maintenance decision that suits each of 
the selected machines according to factors including
frequency of breakdowns, downtime, and cost of 
repairing. 
Preparing the decision making models for the case 
company by use of collected data from CMMS indicated 
that the models can be feasibly and practically applied. 
Comparing the results with maintenance engineers’ ideas 
revealed that utilizing various mathematical decision 
making tools increases the decision analysis capability, by 
helping maintenance engineers and managers to have 
more appropriate maintenance decisions. However, the 
suggested tools require further practical testing in 
different potential applications. Based on the empirical 
findings, it can also be concluded that the results of the 
decision making models cannot be autonomously used, 
rather they should be further analyzed in terms of 
prioritizations and characterization of different failure 
types and main contributing components.
9. ACKNOWLEDGEMENTS
This research work has been funded by the KK-
foundation (the INNOFACTURE research school), 
VINNOVA through the “FFI – Hållbar 
produktionsteknik” research programme, and Mälardalen 
University. The research work is also a part of the 
initiative for Excellence in Production Research (XPRES) 
which is a cooperation between Mälardalen University, 
the Royal Institute of Technology, and Swerea. XPRES is 
one of two governmentally funded Swedish strategic 
initiatives for research excellence within Production 
Engineering.
REFERENCES
1. S. Hess, W. Biter, S. Hollingsworth, “An Evaluation 
Method for Application of Condition Based 
Maintenance Technologies”, Annual Reliability and 
Maintainability Symposium, USA, Philadelphia, pp. 
240-245, 2001.
1. J. H. Shin, H. B. Jun, “On condition based 
maintenance policy”, Journal of Computational 
Designand Engineering, 2(2), pp. 119-127, 2015.
2. Swedish Standards Institute, “Maintenance 
Terminology”, SS-EN 13306, 2001.
3. A. Rosmaini, Sh. Kamaruddin, "An overview of 
time-based and condition-based maintenance in 
industrial application", Computers & Industrial 
Engineering, 63(1), pp. 135-149, 2012.
4. B. S. Blanchard, D. Verm, E. L. Peterson,
“Maintainability: A key to effective and maintenance 
management”, NY: John Wiley & Sons, 1995.
5. B. Al-Najjar, I. Alsyouf, "Selecting the most efficient 
maintenance approach using fuzzy multiple criteria 
decision making", International journal of 
production economics, 84(1), pp. 85-100, 2003.
Authorized licensed use limited to: Universidad Federal de Pernambuco. Downloaded on February 21,2022 at 02:01:00 UTC from IEEE Xplore. Restrictions apply. 
6. A. Labib, "A decision analysis model for 
maintenance policy selection using a CMMS",
Journal of Quality in Maintenance Engineering, 
10(3), pp.191–202, 2004.
7. K. A. H. Kobbacy, D. P. Murthy, “Complex System 
Maintenance Handbook”, Springer, 2008.
8. J. Moubray, “Reliability-centered Maintenance”. 
Industrial Press Inc., Second Edition, New York, 
1997.
9. S. Nakajima, “Introduction to TPM – Total 
Productive Maintenance”, Productivity Press, 
Cambridge, 1998.
10. A. H. C. Tsang, “Condition-based maintenance tools 
and decision making”, Journal of Quality in 
Maintenance Engineering, 1(3), pp. 3–17, 1995.
11. S. J. H. Usher, A. Kamal, W. H. Syed, “Cost optimal 
preventive maintenance and replacement scheduling”
IEEE Transactions, 30, 1121–1128, 1998.
12. A. Marquez, “The Maintenance Management 
Framework: Models and Methods for Complex 
Systems Maintenance”, Springer, 2007.
13. A. S. B. Tam, W. M. Chan, J. W. H. Price, “Optimal 
maintenance intervals for a multi-component system”
Production Planning & Control, 1–11, 2006.
14. A. Gupta, C. Lawsirirat, “Strategically optimum 
maintenance of monitoring-enabled multi-component 
systems using continuous-time jump deterioration 
models” Journal of Quality in Maintenance 
Engineering, 12(3), pp. 306–329, 2006.
15. A. Rastegari, M. Bengtsson, “Cost Effectiveness of 
Condition Based Maintenance in Manufacturing”, 
IEEE 61st Annual Reliability and Maintainability 
Symposium, Florida, USA, 2015.
16. E. Jantunen, A. Arnaiz, D. Baglee, L. Fumagalli, 
“Identification of wear statistics to determine the 
need for a new approach to maintenance”, 
Euromaintenance, 5-8, 2014.
17. C. L. Hwang, A. S. M. Masud, "Multiple objective 
decision making-methods and applications", Berlin: 
Springer-Verlag, vol. 164, 1979.
18. M. Mobin, M. Dehghanimohammadabadi, C. 
Salmon, “Food Product Target Market Prioritization 
Using MCDM Approaches”, Industrial and Systems 
Engineering Research Conference (ISERC), 
Montreal, Canada, 2014.
19. C. Salmon, M. Mobin, A. Roshani, "TOPSIS as a 
Method to Populate Risk Matrix Axes" Proceedings 
of the 2015 Industrial and Systems Engineering 
Research Conference, 2015.
20. B. Dash, M. Debahuti, A. Rath, M. Acharya, "A 
hybridized K-means clustering approach for high 
dimensional dataset" International Journal of 
Engineering, Science and Technology, 2(2), pp. 59-
66, 2010.
21. B. Mokhtarpour, J. Stracener, "Application of a 
clustering technique in identifying the “best” System 
of Systems (SoS) during development." IEEE 
International Conference on Systems, Man and 
Cybernetics (SMC), 2014.
BIOGRAPHIES
Ali Rastegari, Maintenance Eng. & PhD Candidate
Operational Technical Support
Volvo Group Trucks Operations Powertrain Production
Dept BE62143, Köping 
SE-731 80, Sweden
e-mail: ali.rastegari@volvo.com
Ali Rastegari is an industrial PhD candidate at Mälardalen
University since September 2012. He is employed as a 
maintenance engineer at Volvo Group Trucks Operation. 
Ali received his M.Sc. from Mälardalen University in the 
area of Product and Process Development – Production 
and Logistics and his B.Sc. from Tehran Azad University 
in Mechanical Engineering. His background includes 
work as a mechanical and maintenance engineer in 
manufacturing industries. His research interest lies in the 
area of strategic maintenance development focusing on 
use of condition based maintenance in the manufacturing 
industry. He is a member in IEEE Reliability Society.
Mohammadsadegh Mobin, PhD Fellow
Western New England University
Springfield, MA, 01119
e-mail: mm337076@wne.edu
Mohammadsadegh Mobin is a PhD fellow in Industrial 
Engineering and Engineering Management at Western 
New England University. He holds a Master degree in 
Operations Research (2011) and a bachelor degree in 
Industrial Engineering (2009). His research interests lie in 
the areas of reliability engineering and operations 
research.
Authorized licensed use limited to: Universidad Federal de Pernambuco. Downloaded on February 21,2022 at 02:01:00 UTC from IEEE Xplore. Restrictions apply. 
Appendix A.1 - Criteria analysis on data collected from 
EWO database
AA 
Machines' 
Name
Downtime 
(hrs)
AA 
Machines' 
Name
Frequency 
(No. of 
breakdowns)
H
ig
h
8527784 179,18 86301 13
H
ig
h
87843 131,83 88157 12
86588 104,31 8527784 12
86300 103,36 86588 12
M
ed
iu
m
8531809 69,99 87843 12
86301 66,42 8531809 11
88157 57,41 8528618 10
8527783 57,36 87842 7
M
ed
iu
m8531801 50,55 8527783 6
Lo
w
87842 41,46 86587 5
8528618 36,78 86300 5
86587 34,77 8531802 4
Lo
w8531802 23,94 8531801 2
8528609 6,12 8528609 2
Sum of AA 
Machines 963,48 Sum of AA 
Machines 113
Sum of All 1818,13 Sum of All 208
Percentage 53% Percentage 54%
Appendix A.2 - The summary of categories of machines
Category Rank
min
DT Cost Fr.
1 1-100 28.53 6558 1
2 101-200 3.89 980 1
3 201-300 4.05 29.5 1
4 301-400 2.2 163.3 1
5 401-540 0.31 32.67 1
All 1-540 0.31 29.56 1
Max
Category Rank DT Cost Fr.
1 1-100 469.83 149086 30
2 101-200 296.21 102501 26
3 201-300 275.31 47162 11
4 301-400 108.01 31890 12
5 401-540 47.74 8695.8 5
All 1-540 469.83 149086 30
Average
Category Rank DT Cost Fr.
1 1-100 205.24 57390 14.1
2 101-200 95.80 20783 7.64
3 201-300 45.44 10371 4.33
4 301-400 18.33 5757.8 2.59
5 401-540 5.36 1389.4 1.25
All 1-540 68.95 17823 5.64
Appendix A.3 - The summary of categories of parts
Category Rank
min
DT Cost Fr.
1 1-500 1.36 163.4 1
2 501-1000 0.08 163.3 1
3 1001-1500 0.50 29.56 1
4 1501-2000 0.10 1 1
5 2001-2800 0.19 1 1
All 1-2878 0.10 1 1
Max
Category Rank DT Cost Fr.
1 1-500 382.5 59083 9
2 501-1000 59.17 12395 2
3 1001-1500 29.58 6209.6 1
4 1501-2000 13.24 2629.7 1
5 2001-2800 7.330 1423.2 1
All 1-2878 382.5 59083 9
Average
Category Rank DT Cost Fr.
1 1-500 66.45 13426 2
2 501-1000 15.90 4180.9 1.5
3 1001-1500 10.53 3084.4 1
4 1501-2000 5.093 1664.0 1
5 2001-2800 2.413 749.21 1
All 1-2878 16.55 4042.9 1.3
Authorized licensed use limited to: Universidad Federal de Pernambuco. Downloaded on February 21,2022 at 02:01:00 UTC from IEEE Xplore. Restrictions apply. 
false
 /PreserveOverprintSettings true
 /StartPage 1
 /SubsetFonts false
 /TransferFunctionInfo /Remove
 /UCRandBGInfo /Preserve
 /UsePrologue false
 /ColorSettingsFile ()
 /AlwaysEmbed [ true
 /Arial-Black
 /Arial-BoldItalicMT
 /Arial-BoldMT
 /Arial-ItalicMT
 /ArialMT
 /ArialNarrow
 /ArialNarrow-Bold
 /ArialNarrow-BoldItalic
 /ArialNarrow-Italic
 /ArialUnicodeMS
 /BookAntiqua
 /BookAntiqua-Bold
 /BookAntiqua-BoldItalic
 /BookAntiqua-Italic
 /BookmanOldStyle
 /BookmanOldStyle-Bold
 /BookmanOldStyle-BoldItalic
 /BookmanOldStyle-Italic
 /BookshelfSymbolSeven
 /Century
 /CenturyGothic
 /CenturyGothic-Bold
 /CenturyGothic-BoldItalic
 /CenturyGothic-Italic
 /CenturySchoolbook
 /CenturySchoolbook-Bold
 /CenturySchoolbook-BoldItalic
 /CenturySchoolbook-Italic
 /ComicSansMS
 /ComicSansMS-Bold
 /CourierNewPS-BoldItalicMT
 /CourierNewPS-BoldMT
 /CourierNewPS-ItalicMT
 /CourierNewPSMT
 /EstrangeloEdessa
 /FranklinGothic-Medium
 /FranklinGothic-MediumItalic
 /Garamond
 /Garamond-Bold
 /Garamond-Italic
 /Gautami
 /Georgia
 /Georgia-Bold
 /Georgia-BoldItalic
 /Georgia-Italic
 /Haettenschweiler
 /Impact
 /Kartika
 /Latha
 /LetterGothicMT
 /LetterGothicMT-Bold
 /LetterGothicMT-BoldOblique
 /LetterGothicMT-Oblique
 /LucidaConsole
 /LucidaSans
 /LucidaSans-Demi
 /LucidaSans-DemiItalic
 /LucidaSans-Italic
 /LucidaSansUnicode
 /Mangal-Regular
 /MicrosoftSansSerif
 /MonotypeCorsiva
 /MSReferenceSansSerif
 /MSReferenceSpecialty
 /MVBoli
 /PalatinoLinotype-Bold
 /PalatinoLinotype-BoldItalic
 /PalatinoLinotype-Italic
 /PalatinoLinotype-Roman
 /Raavi
 /Shruti
 /Sylfaen
 /SymbolMT
 /Tahoma
 /Tahoma-Bold
 /TimesNewRomanMT-ExtraBold
 /TimesNewRomanPS-BoldItalicMT
 /TimesNewRomanPS-BoldMT
 /TimesNewRomanPS-ItalicMT
 /TimesNewRomanPSMT
 /Trebuchet-BoldItalic
 /TrebuchetMS
 /TrebuchetMS-Bold
 /TrebuchetMS-Italic
 /Tunga-Regular
 /Verdana
 /Verdana-Bold
 /Verdana-BoldItalic
 /Verdana-Italic
 /Vrinda
 /Webdings
 /Wingdings2
 /Wingdings3
 /Wingdings-Regular
 /ZWAdobeF
 ]
 /NeverEmbed [ true
 ]
 /AntiAliasColorImages false
 /CropColorImages true
 /ColorImageMinResolution 200
 /ColorImageMinResolutionPolicy /OK
 /DownsampleColorImages true
 /ColorImageDownsampleType /Bicubic
 /ColorImageResolution 300
 /ColorImageDepth -1
 /ColorImageMinDownsampleDepth 1
 /ColorImageDownsampleThreshold 1.50000
 /EncodeColorImages true
 /ColorImageFilter /DCTEncode
 /AutoFilterColorImages false
 /ColorImageAutoFilterStrategy /JPEG
 /ColorACSImageDict >
 /ColorImageDict >
 /JPEG2000ColorACSImageDict >
 /JPEG2000ColorImageDict >
 /AntiAliasGrayImages false
 /CropGrayImages true
 /GrayImageMinResolution 200
 /GrayImageMinResolutionPolicy /OK
 /DownsampleGrayImages true
 /GrayImageDownsampleType /Bicubic
 /GrayImageResolution 300
 /GrayImageDepth -1
 /GrayImageMinDownsampleDepth 2
 /GrayImageDownsampleThreshold 1.50000
 /EncodeGrayImages true
 /GrayImageFilter /DCTEncode
 /AutoFilterGrayImages false
 /GrayImageAutoFilterStrategy /JPEG
 /GrayACSImageDict >
 /GrayImageDict >
 /JPEG2000GrayACSImageDict >
 /JPEG2000GrayImageDict >
 /AntiAliasMonoImages false
 /CropMonoImages true
 /MonoImageMinResolution 400
 /MonoImageMinResolutionPolicy /OK
 /DownsampleMonoImages true
 /MonoImageDownsampleType /Bicubic
 /MonoImageResolution 600
 /MonoImageDepth -1
 /MonoImageDownsampleThreshold 1.50000
 /EncodeMonoImages true
 /MonoImageFilter /CCITTFaxEncode
 /MonoImageDict >
 /AllowPSXObjects false
 /CheckCompliance [
 /None
 ]
 /PDFX1aCheck false
 /PDFX3Check false
 /PDFXCompliantPDFOnly false
 /PDFXNoTrimBoxError true
 /PDFXTrimBoxToMediaBoxOffset [
 0.00000
 0.00000
 0.00000
 0.00000
 ]
 /PDFXSetBleedBoxToMediaBox true
 /PDFXBleedBoxToTrimBoxOffset [
 0.00000
 0.00000
 0.00000
 0.00000
 ]
 /PDFXOutputIntentProfile (None)
 /PDFXOutputConditionIdentifier ()
 /PDFXOutputCondition ()
 /PDFXRegistryName ()
 /PDFXTrapped /False
 /CreateJDFFile false
 /Description 
 /CHT 
 /DAN 
 /DEU 
 /ESP/FRA 
 /ITA (Utilizzare queste impostazioni per creare documenti Adobe PDF adatti per visualizzare e stampare documenti aziendali in modo affidabile. I documenti PDF creati possono essere aperti con Acrobat e Adobe Reader 5.0 e versioni successive.)
 /JPN 
 /KOR 
 /NLD (Gebruik deze instellingen om Adobe PDF-documenten te maken waarmee zakelijke documenten betrouwbaar kunnen worden weergegeven en afgedrukt. De gemaakte PDF-documenten kunnen worden geopend met Acrobat en Adobe Reader 5.0 en hoger.)
 /NOR 
 /PTB 
 /SUO 
 /SVE 
 /ENU (Use these settings to create PDFs that match the "Required" settings for PDF Specification 4.01)
 >>
>> setdistillerparams
> setpagedevice