sustainability

Article
Machine Learning Approach Using MLP and SVM
Algorithms for the Fault Prediction of a Centrifugal
Pump in the Oil and Gas Industry
Pier Francesco Orrù 1 , Andrea Zoccheddu 1 , Lorenzo Sassu 2 , Carmine Mattia 2 ,
Riccardo Cozza 3 and Simone Arena 1, *
1 Department of Mechanical, Chemical and Material Engineering, University of Cagliari, Via Marengo 2,
09134 Cagliari, Italy; pforru@unica.it (P.F.O.); and.zoccheddu@gmail.com (A.Z.)
2 Sartec-Saras Ricerche e Tecnologie Srl, Via 2◦ Traversa Strada Est, 09032 Cagliari, Italy;
lorenzo.sassu@sartec.it (L.S.); carmine.mattia@sartec.it (C.M.)
3 Saras S.p.A.-S.S. Sulcitana n.195 Km 19, 09018 Cagliari, Italy; riccardo.cozza@saras.it
* Correspondence: simonearena@unica.it; Tel.: +39-070-6755350




Received: 25 May 2020; Accepted: 9 June 2020; Published: 11 June 2020


Abstract: The demand for cost-effective, reliable and safe machinery operation requires accurate fault
detection and classification to achieve an efficient maintenance strategy and increase performance.
Furthermore, in strategic sectors such as the oil and gas industry, fault prediction plays a key role to
extend component lifetime and reduce unplanned equipment thus preventing costly breakdowns and
plant shutdowns. This paper presents the preliminary development of a simple and easy to implement
machine learning (ML) model for early fault prediction of a centrifugal pump in the oil and gas industry.
The data analysis is based on real-life historical data from process and equipment sensors mounted on
the selected machinery. The raw sensor data, mainly from temperature, pressure and vibrations probes,
are denoised, pre-processed and successively coded to train the model. To validate the learning
capabilities of the ML model, two different algorithms—the Support Vector Machine (SVM) and the
Multilayer Perceptron (MLP)—are implemented in KNIME platform. Based on these algorithms,
potential faults are successfully recognized and classified ensuring good prediction accuracy. Indeed,
results from this preliminary work show that the model allows us to properly detect the trends
of system deviations from normal operation behavior and generate fault prediction alerts as a
maintenance decision support system for operatives, aiming at avoiding possible incoming failures.

Keywords: predictive maintenance; machine learning; artificial neural networks; oil and gas industry;
fault diagnosis

1. Introduction
In the middle of the fourth industrial revolution, industries are constantly looking for ways to
optimize production lines and, at the same time, to reduce their costs. Maintenance expenses typically
amount to more than one-third of total operating costs. Traditional maintenance techniques are based
on two different strategies: corrective maintenance and preventive maintenance [1]. The former is
performed to repair faulty systems and equipment only when a failure has already occurred and,
in such a way, the process’ direct costs increase. The latter is performed after regular time intervals
to prevent systems/equipment failures. Thus, repairs on machinery or components are carried out
when they still have an unknown remaining useful life leading to both unplanned downtimes of
machinery and an increase in the operating costs. Therefore, a proper maintenance strategy should
improve general equipment health status and, furthermore, it should reduce equipment failure rates,
minimizing maintenance costs and maximizing equipment’s useful life [2]. In this scenario, the concepts

Sustainability 2020, 12, 4776; doi:10.3390/su12114776 www.mdpi.com/journal/sustainability

Sustainability 2020, 12, 4776 2 of 15

of sustainable energy and environmental management are framed in many industrial activities,
since they provide better accountability, better control and costs’ allocation, improved performance
and waste material reduction [3]. To this end, reliable and efficient maintenance strategies and
good practices have to be considered indispensable to improve sustainability, for they increase
plant efficiency aiming at both reducing downtime, waste of energy and materials and enhancing
socioeconomic well-being significantly [4]. Enhanced by these factors, Predictive Maintenance (PdM)
has become one of the hottest topics of the Industry 4.0 era, being based on high quality and
optimally scheduled sustainable maintenance practices by integrating physical and digital systems
of production environments. Technological developments in recent years have led to a reduction
in expenses related to monitoring the health of machinery and to the acquisition and storage of
huge amounts of data. The application of cutting-edge analytical models to data provides valuable
information and knowledge from manufacturing processes, production systems and equipment to
support strategic decision-making.
PdM’s goal is to predict a future failure event allowing to react proactively and plan the most
appropriate maintenance solution at the right time by using data and algorithms. Thus, through
advanced condition monitoring techniques, this approach assures high-value machinery management
optimization in highly competitive fields, such as manufacture, transport and energy production.
Quite recently, advances in computational performance have allowed the application of machine
learning (ML) algorithms which are able to find correlations and to identify patterns over extremely
complex and large amounts of data [5,6]. The basic idea is to use machines’ data aiming at processing
the training database, in order to provide reliable and efficient fault diagnostic models. To date, a large
amount of research and studies have been conducted on predictive maintenance using ML algorithms
applied to industrial applications. Carvalho et al. [2] presented a Systematic Literature Review (SLR)
focused on machine learning methodologies applied to predictive maintenance techniques. Lei et al. [7]
presented a SLR on the applications of machine learning to machine fault diagnosis and provided a
roadmap for this field. Del Ser et al. [8] presented a detailed review of the recent developments in data
analysis strategies and machine learning algorithms for data-driven prognosis within the Industry 4.0
paradigm according to a specific classification in descriptive, predictive and prescriptive models.
According to the results reported in these mentioned SRLs, the number of publications on
machine fault diagnosis by using machine learning has rapidly increased since the 2010s. In particular,
Del Ser et al. [8] reported more than twenty works on predictive prognostic models but only one of
them is based on machine learning modeling focused on oil and gas. In this paper, Costello et al. [9]
present a data-driven model for health monitoring of a gas circulator units using real vibration data.
Despite a detailed literature review on machine learning modeling for industrial applications
(manufacturing, transportation, energy production, heavy industry among others) it has been difficult
to make use of past studies, carried out on real raw data from operating machinery [10–14]. Recently,
Hajizadeh [15] and Hanga and Kovalchuk [16] reported the efforts to date of applying artificial
intelligence in fault detection in the oil and gas industry, both analyzing its benefits and suggesting ways
to ensure its greater adoption for strategic management and technology enablement. Indeed, due to the
lack of willingness on the side of companies to make public data ownership, particularly in the field of
petrochemical applications, most studies rely on data retrieved from simulations or experimental tests
carried out in a laboratory, with ideal setups been used [5,17–19]. In recent years, different ML methods
applied to rotating machinery fault diagnosis have been developed. Qian et al. [20,21] proposed a
transfer learning method for fault diagnosis of these components under variable working conditions.
Su et al. [22], Weiwei et al. [23], Zhang et al. [24], and Lei et al. [25] presented different approaches
to develop machine learning algorithms to roller bearing fault diagnosis. Bilski [26] presented
a Support Vector Machines (SVM) model to monitor the health state of the induction machine.
Jirdehi and Rezaei [27] proposed two methods based on artificial neural networks (ANN) and adaptive
neuro-fuzzy inference system (ANFIS) to estimate the induction motor parameters in the single-cage
and double-cage models. Romeo et al. [28] proposed an innovative Design Support System (DesSS)
Sustainability 2020, 12, 4776 3 of 15

for the prediction and estimation of machine specification data. Giantomassi et al. [29] investigated
the fault detection and diagnosis of an electric motor by using an estimation algorithm of the
probability density function of current signals. Paolanti et al. [30] presented a ML architecture based
on Random Forest approach for fault diagnosis of a cutting machine. Yang et al. [31] presented a
fault diagnosis scheme combined with hierarchical symbolic analysis (HSA) and convolutional neural
network (CNN). They focused the analysis on experimental testing of a centrifugal pump and a motor
bearing. Pang et al. [32] proposed a new fault diagnosis method based on multiple-domain data
by using experimental testing dataset of gearbox, rotor and engine rolling bearing. Chen et al. [33]
proposed a novel fault diagnosis approach integrating CNN and Extreme Learning Machine (ELM)
by using experimental dataset of gearbox and motor bearing. Zhang et al. [34] presented an ant
colony algorithm for synchronous feature selection and parameter optimization for support vector
machine in fault diagnosis of rotating machinery by the experiment of rotor system and locomotive
roller bearings. Wang et al. [35] developed a novel ensemble ELM network for compound-fault
diagnosis of rotating machinery by using real-world database. Panda et al. [36] focused on the
vibration-based condition monitoring and fault diagnosis of centrifugal pumps using the Support
Vector Machine (SVM) algorithm. Wang et al. [37] proposed a novel fault diagnosis method that
utilizes ensemble learning with differentiated probabilistic neural networks by using an experimental
dataset of rotary actuator systems and plunger hydraulics. Liu et al. [38] presented a comprehensive
review of artificial intelligence algorithms in rotating machinery fault diagnosis, from both the views of
theory background and industrial applications. Tang et al. [39] proposed a novel adaptive learning rate
deep belief network combined with Nesterov momentum for rotating machinery fault diagnosis. The
authors implemented the proposed model on datasets from gearbox and locomotive bearing test rigs.
Yu et al. [40] proposed a probabilistic neural network algorithm to predict oil-immersed transformer
internal faults. Zenisek et al. [41] reported a machine learning-based approach for detecting drifting
behavior to identify wear and tear, and consequent malfunctioning by analyzing real time condition
monitoring data. Li et al. [12] proposed a machine learning approach to predict impending failures
and alarms of critical rail car components aiming at both driving proactive inspections and repairs
and reducing operational equipment failure. Lee et al. [42] presented an algorithm based on artificial
intelligent for predictive maintenance to monitor two critical machine tool system elements, i.e., the
cutting tool and the spindle motor. Guedes et al. [43] investigated the evaluation of the electrical
insulation of the stator of three-phase induction motors and the classification of the failure mechanism
using artificial neural networks (ANNs). The literature review presented above highlights that
machine learning has attracted both academic and practitioners in the field of industrial maintenance
management, emerging as a powerful tool for intelligent predictive algorithms development in many
applications. ML approaches are able to find patterns of interest based upon the data monitored from
the process or asset and transform raw data into features spaces by means of mathematical algorithms.
To this end, the recent development in technological fields—such as Cyber-Physical systems, Internet of
Things (IoT) and Big Data—has provided a large amount of data extracted from production processes,
allowing ML techniques to face a wide variety of problems including prediction, anomaly detection,
classification, regression, or forecasting [44]. On the other hand, as reported in [45], the implementation
of ML techniques in real factories remains rather challenging. Major challenges are:

• Data acquisition and storage [46]. A large amount of data is used in the process of training
and learning. In general, more data leads to more reliable models and, consequently, better results.
However, data should be representative of the analyzed process. Therefore, the acquisition of
relevant data has a strong influence on the ML algorithm performance. Moreover, the processing
of large repositories of time series data capturing during the process is necessary to handle them
aiming at extracting valuable knowledge and information;
• Selection and design of the ML algorithm [44]. The ML model must be able to estimate machinery
condition in a short time interval, aiming at performing proper and agile decision-making.
As said before, the growing interest in the field of ML in manufacturing has led to the development
Sustainability 2020, 12, 4776 4 of 15

of a large number of different ML algorithms or at least variations to adapt their operation

according to different situations requiring deep skills and expertise. Although the core principles
of ML are now accessible to a wider audience, adequate knowledge and deep technical know-how
Sustainability 2020, 12, x FOR PEER REVIEW 4 of 15
are required.
led to the development of a large number of different ML algorithms or at least variations to
Theadapt
foregoing discussion suggests that there is a need to address these challenges in order to
their operation according to different situations requiring deep skills and expertise.
enable anAlthough
effectivethe
implementation
core principles of MLare
of ML techniques aiming
now accessible to at leading
a wider to a good
audience, maintenance
adequate knowledge strategy
for an accurate fault diagnosis. In this work,
and deep technical know‐how are required. a fault prediction technique based on a ML supervised
method using data from actual working operating conditions is presented. The aim is to develop a
The foregoing discussion suggests that there is a need to address these challenges in order to
ML model that is ready to implement. The motivation behind this approach is based on the need for
enable an effective implementation of ML techniques aiming at leading to a good maintenance
companies
strategytofor
havean an accurate,
accurate faulteasy and responsive
diagnosis. In this work,toola in order
fault to predict
prediction the failures
technique basedofon one
a MLor more
system equipments.
supervised method This need
using is not
data fromeasy to working
actual meet, especially
operatingwhen it comes
conditions to complex
is presented. Theproblems
aim is to that
require the handling
develop a ML modelof large number
that is ready toofimplement.
computations. The motivation behind this approach is based on
theselect
To need for companies
a simple to have an
and accurate MLaccurate,
model easy and responsive
to failure prediction, tool in order to predict
a comparison between thetwofailures
different
of one or more system equipments. This need is not easy to meet, especially
algorithms, the Support Vector Machine (SVM) and the Multilayer Perceptron (MLP) is performed. when it comes to
complex problems that require the handling of large number of computations.
The open-source software KNIME platform is adopted to set-up the model workflow. The data used
To select a simple and accurate ML model to failure prediction, a comparison between two
in this study come from a real oil and gas industry operating centrifugal pump inserted inside the
different algorithms, the Support Vector Machine (SVM) and the Multilayer Perceptron (MLP) is
production line of the SARLUX refinery (Sarroch, Italy). The data analysis is complicated by difficulties
performed. The open‐source software KNIME platform is adopted to set‐up the model workflow.
linked
Theto data
the production
used in this processes,
study comesuch fromasa the
realnon-rigorous recording
oil and gas industry of events
operating by fieldpump
centrifugal operators
and sensors’ malfunction.
inserted inside The following
the production chapters
line of the SARLUX arerefinery
structured as follows:
(Sarroch, Section
Italy). The data 2analysis
illustrates is the
methodologies
complicated by difficulties linked to the production processes, such as the non‐rigorous recording of SVM
applied to the case study. Particularly, the different steps performed to develop
and ANN
eventsmodels
by field and selection
operators of the most
and sensors’ suitableThe
malfunction. input variables
following are briefed;
chapters Section
are structured as 3follows:
reports the
Section
results 2 illustrates
achieved for boththethe
methodologies applied toand,
selected algorithms the case study.
finally, Particularly,
in Section the differentand
4, conclusions steps future
performed
research to develop SVM and ANN models and selection of the most suitable input variables are
are depicted.
briefed; Section 3 reports the results achieved for both the selected algorithms and, finally, in Section
4, conclusionsApplied
2. Methodology and future
toresearch
the CaseareStudy
depicted.

Machine learning
2. Methodology is a subfield
Applied of computer
to the Case Study science which can be defined as the process of solving
a practical problem by gathering a dataset and algorithmically building a statistical model based
Machine learning is a subfield of computer science which can be defined as the process of
on that dataset to extract information about it [47]. Learning can be supervised, semi-supervised,
solving a practical problem by gathering a dataset and algorithmically building a statistical model
unsupervised
based on and that by reinforcement.
dataset to extract Ininformation
this paper, about
a supervised
it [47]. learning
Learning technique has been used,
can be supervised,
sincesemi‐supervised,
every sample ofunsupervised
the dataset has been labeled to train a model that takes feature
and by reinforcement. In this paper, a supervised learning techniquevectors as input
and outputs
has beeninformation that sample
used, since every allows oftothe
deduce the
dataset haslabel
beenoflabeled
new collected
to train a data.
model that takes feature
vectors
The as input Maintenance
Predictive and outputs information that allows
(PdM) oriented to deduce
pipeline thethis
built for label of new
study collected
relies data. learning
on machine
techniques applied to data analysis collected upon operations of a centrifugal pump. machine
The Predictive Maintenance (PdM) oriented pipeline built for this study relies on The pipeline
learning
has been techniques
structured appliedin
as shown toFigure
data analysis
1. The collected
reportedupon operations
pipeline of a centrifugal
represents pump.
the schematic The
diagram of
pipeline has been structured as shown in Figure 1. The reported pipeline represents the schematic
the ML technique which involves some steps: (i) data acquisition step, in which data are collected
diagram of the ML technique which involves some steps: (i) data acquisition step, in which data are
and stored aiming at achieving relevant data to equipment health; (ii) data processing step, in which
collected and stored aiming at achieving relevant data to equipment health; (ii) data processing step,
collected data have been labeled and cleaned in order to be efficiently processed by the ML model.
in which collected data have been labeled and cleaned in order to be efficiently processed by the ML
Furthermore, feature engineering
model. Furthermore, has been performed
feature engineering to extract to
has been performed new features
extract new and the relevant
features and the ones
haverelevant
been selected
ones haveto train
beenthe model;
selected to (iii)
trainmodel selection,
the model; training
(iii) model and validation
selection, step,
training and in which the
validation
selection
step, of
in the
whichproper ML model
the selection is selected,
of the proper ML trained
modeland then validated
is selected, trained andby maximizing
then validatedthebymore
maximizing the more
appropriate evaluation metrics.appropriate evaluation metrics.

DATA FEATURE DATA MODEL ML MODEL

ENGINEERING TRAINING
LABELLING AND FUTURE CLEANING SELECTION AND
EXTRACTION EVALUATION

Figure 1. Generic schematic diagram of the proposed method for machinery fault diagnosis
Figure 1. Generic schematic diagram of the proposed method for machinery fault diagnosis.

Knime platform is used to design the workflow of centrifugal pump fault diagnosis, as reported
Knime platform is used to design the workflow of centrifugal pump fault diagnosis, as reported
in Figure 2. In accordance with the logic of this workflow, the data analysis and processing will be
in Figure 2. In accordance with the logic of this workflow, the data analysis and processing will be
Sustainability 2020, 12, 4776 5 of 15

briefly described in the following. A supervised predictive maintenance-oriented problem can mainly
be addressed by means of two approaches:
Sustainability 2020, 12, x FOR PEER REVIEW 5 of 15

• A classification approach aims at predicting whether the centrifugal pump will fail or not within a
briefly described in the following. A supervised predictive maintenance‐oriented problem can
future time interval, called failure prediction horizon T in this paper, whose length is determined
mainly be addressed by means of two approaches:
by various factors, such as management needs and possibilities, domain expert decisions, etc.
 A classification approach aims at predicting whether the centrifugal pump will fail or not
The parameter T is defined as the minimum lead time required to replace components before a
within a future time interval, called failure prediction horizon T in this paper, whose length is
problem occurs and
determined by its valuefactors,
various is defined
suchinassuch a way that
management it would
needs be possibledomain
and possibilities, to apply proactively
expert
maintenance to avoid
decisions, etc. Thethe problem.
parameter T is defined as the minimum lead time required to replace
• A regression approach
components before aaims at predicting
problem occurs and how much
its value time is
is defined in left
suchbefore
a way the
that next failure
it would be event
possible to apply proactively maintenance
(RUL, Remaining Useful Life prediction) [48]. to avoid the problem.
 A regression approach aims at predicting how much time is left before the next failure event
(RUL,applies
This study Remaining Useful Life prediction)
a classification technique.[48]. The parameter T has been set to be a one-week
time interval, since domain experts have identified it as the Tminimum
This study applies a classification technique. The parameter has been settime
to berequired
a one‐week to apply
maintenance to thesince
time interval, pump before
domain a failure
experts have event.
identifiedInit order
as the to perform
minimum timeclassification, we need to
required to apply
maintenance
categorize to the pump
each timestamp beforeusing
in a class a failure event. labels.
arbitrary In order Twoto perform classification,
classes have we need to
been identified to define a
categorize each timestamp in a class using arbitrary labels. Two classes have been identified to
categorical variable named “Current pump operating status”: a positive instance class “1” referring to
define a categorical variable named “Current pump operating status”: a positive instance class “1”
the samples included inside the failure prediction horizon and a negative instance class “0” referring to
referring to the samples included inside the failure prediction horizon and a negative instance class
normal “0”
operating
referringconditions. Therefore,
to normal operating in order Therefore,
conditions. to correctly in classify
order to the failureclassify
correctly prediction horizon and
the failure
act before it happens,
prediction 168and
horizon timestamps before
act before it each168
happens, failure event before
timestamps have been
each labeled as “1”,
failure event havethe remaining
been
labeled
as “0”. It as “1”, the
is important remaining
to notice thatasthe
“0”. It is important
definition of thetoparameter
notice that T the definition
implies theof the parameter
conservative T
hypothesis
implies
that in that the conservative
period, a change hypothesis
from normal that in that period,
operating a change from
conditions normal
of the pump operating
happens.conditions
According to
of the pump happens. According to this consideration, data immediately prior to T could be seen as
this consideration, data immediately prior to T could be seen as noise, since it would be difficult to be
noise, since it would be difficult to be considered belonging to one of the classes, and will be
considered
removedbelonging
from thetoanalysis
one offurther
the classes,
on. and will be removed from the analysis further on.

Figure
Figure 2. Knime
2. Knime workflowbuilt
workflow built for
for fault
faultdiagnosis
diagnosisof aofcentrifugal pump.pump.
a centrifugal

2.1.Acquisition
2.1. Data Data Acquisition
StepStep

Dataset2.1.1. Dataset
The dataset used in this work refers to centrifugal pump operations‐related sensor readings
The dataset used in this work refers to centrifugal pump operations-related sensor readings which
which were collected between January 2010 and May 2014 since, on that date, the pump was
were collected
replaced.between January 2010 and May 2014 since, on that date, the pump was replaced.
The dataThefor PdM
data for are
PdMin are
timeinseries format.
time series Dataset
format. includes
Dataset a timestamp,
includes a set
a timestamp, ofof
a set sensor
sensorreadings
readings
collected at the collected
same timeat the
assame time as timestamps,
timestamps, and deviceandidentifiers.
device identifiers.
Measurements from eight different sensors have been included in the analysis and used to build the
features, i.e., the predictive attributes the model is built on: one of flow rate, two of bearings vibration,
two of axial displacement and three of motor coil temperature.
Sustainability 2020, 12, 4776 6 of 15

Sampling has been executed with a frequency of one sample per hour, in order to deal with a
reasonable amount of data.
By referring to machine’s historical events register, it has been possible to identify eight major
failure events which occurred during the considered period, four of which were related to seal leakages
and four to ambiguously registered failure events. The lack of standardized recording and reporting
failure methodologies leads to the need to reject such events due to uncertain data interpretation. Thus,
in this study, only four seal leakages failure events have been included in the dataset.

2.2. Data Processing Step

2.2.1. Missing Values Treatment

For each feature, it has been necessary to fill gaps between data samples due to missing values.
Missing values are a common issue related to on-field operating machinery measurements, since
many ML algorithms are not efficient in dealing with datasets affected by missing data. This has
been performed by means of the “Missing Values” node in KNIME (Figure 2), using the linear
interpolation technique, which replaces each missing value with new ones obtained by linear
interpolation between the previous and next non-missing value encountered for each feature column.
Linear interpolation was chosen, among different imputation methods, to be simple and effective for
time series data [49].

2.2.2. Feature Engineering

The correlation between a data sample and preceding samples in chronological order is one of the
most important aspects of a time series type dataset. It is appropriate to perform a feature engineering
in order to introduce this prior knowledge in the ML algorithm. Thus, eight new features have been
created by means of backward simple moving average calculation, based upon a ten hours window
over each data point for each feature. This step, achieved with the “Moving Average” node in KNIME,
guarantees to take the correlation existing between a sample and the preceding ones of the time series
into consideration and ensures noise attenuation.

2.2.3. Data Cleaning

It has also been necessary to filter out the data related to downtime periods, in which the
machine has been shut down due to maintenance operations or other issues, as well as data related
to start-up periods. This has been done by setting up a threshold over the flow rate sensor, based
upon domain knowledge of the process the pump is inserted in: every sample corresponding to a
flow rate below the threshold has been filtered out by means of a “Rule Based Filter” node. Model’s
performance would have been highly penalized by including this set of data in the study, since it does
not refer to neither a normal operating condition nor a condition included in the failure prediction
horizon interval.
It has been considered appropriate to eliminate all the data following a pump substitution, which
may have caused variations of normal operating conditions behavior from the ones of previously
installed machine.
A further data filtration has been performed in order to wipe out samples included in the time
interval between one week and three weeks prior to each failure event, due to the high level of
uncertainty related to labeling this sets of data either as normal operating conditions or pre-failure
ones and in order to get a more class-balanced dataset.

2.3. Model Selection and Validation

The machine learning model hyperparameters have been optimized by means of a grid search
and k-fold cross validation technique. The results have been validated by means of the k-fold
cross validation technique to assess how accurately the model would perform in practice and to
Sustainability 2020, 12, 4776 7 of 15

avoid overfitting. The input dataset is partitioned over a loop in k equal-sized subsets: of all subsets, a
single one is retained as validation set and the remaining as training for the machine learning algorithm.
The loop executes k iterations, each one referring to a different validation set and, accordingly, a
different training set. The scoring metrics of the model are given by the average of the scoring metrics
of all iterations [50].
The value of k parameter has been chosen in such a way that every fold of the k equal-sized
subsets includes at least one failure event and data samples preceding a failure event are not split
between folds. Since the dataset’s failure events are not distributed uniformly, an optimal value for k
has been determined. In the case study, a value of three for k ensures a good representation of the
input dataset by the training and validation subset groups.
Since the algorithm’s performance would have been penalized by the highly unbalanced dataset,
as it will be discussed in the Section 2.3.2, the SMOTE (Synthetic Minority Oversampling Technique)
algorithm [51] has been used. This approach is inspired by a technique developed in [52] characterized
by the implementation of extra training data by performing certain operations on real data. A dataset is
imbalanced if the classification categories are not approximately equally represented. Thus, the SMOTE
algorithm increases the number of samples of the minority class by creating new synthetic observations
generated through simple mathematical and geometric operations. In this case, the number of samples
of the minority class is set in order to equalize its size with the majority class.

2.3.1. Classification Algorithms

In this work, two widely used classification techniques are implemented and compared, namely:
Support Vector Machines (SVMs) and Multilayer Perceptron (MLP).
Support Vector Machines (SVMs): it is a non-probabilistic binary (two class) linear classifier
originally proposed by Vapnik and Cortes [53,54]. SVM separates data across a decision boundary,
named hyperplane f (x) = 0, by solving a constrained quadratic optimization problem based on the
structural risk minimization. The given data input xi (i = 1, 2, . . . , N ) consists of objects with different
labels corresponding to the two classes namely positive and negative class. In the case of linearly data,
the hyperplane that separates the given data is obtained by Equation (1):

N
X
y = f ( x ) = WT x + b = Wi xi +b (1)
i=1

where N is the number of the samples, W is a N-dimensional vector and b is a scalar. The vector W and
scalar b are used to define the position of separating hyperplane. The optimal separating hyperplane is
the separating hyperplane that creates the maximum distance between the plane and the nearest data,
i.e., the maximum margin. SVM can also be used in non-linear classification tasks with the application
of the kernel function. Indeed, working in the high-dimensional feature space generates problems
due to handling non-linearly separable features which can be solved by using the kernel function.
The selection of the proper kernel function is very important, since it defines the feature space in which
the training dataset will be classified. In this paper, the kernel Radial Basis Function (RBF) is used.
The RBF kernel hyper parameter γ and the SVM penalty parameter C are optimally selected to obtain
the best classification performance. In this work, the values of these two parameters are γ = 0 and
C = 55.
Multilayer Perceptron (MLP) has been used as an Artificial Neural Networks (ANN)
classification algorithm. Since MLP requires normalized data as input, z-score normalization has been
made prior to the algorithm training by means of a “Normalizer” node. Then, the same technique has
been applied to test data. MLPs are a powerful class of nonlinear statistical models which consist of
multiple layers of nodes in a directed graph, with each layer fully connected to the next one. There are
three different type of layers, i.e., input, hidden and output layer. Thus, except for the input nodes,
each node is a neuron (or processing element) with a nonlinear activation function [50].
Sustainability 2020, 12, 4776 8 of 15

Given data input xi (i = 1, 2, . . . , N ), the neural model output y can be obtained by Equation (2):
N 
  X 
T
y = f W x = f Wi xi +b (2)
i=1

where f is the activation function, N is the number of the neurons, W are the ANN model weights and
b is the bias vector.
A binary classification MLP’s output is a value included in the interval between 0 and 1,
which could be considered as the probability of the positive target class. A parameter optimization
loop has been used to tune the hyperparameters of the model so that precision and recall over class “1”
are maximized, as it will be discussed in the Section 2.3.2. The model selection has been performed by
optimizing the number of hidden layers and the number of hidden neurons per layer of MLP. In this
work, three hidden layers composed of ten neurons for each layer are adopted.
MLP was trained using the resilient propagation algorithm (RPROP). RPROP is a first-order
neural network optimization algorithm, developed to train shallow MLP in a fast and robust way.
Unlike other optimization algorithms, it has no hyperparameter to choose from [55], this leads to a
significant reduction in the required computational resources. The algorithm implementation has been
performed by means of the “RProp MLP” node.

2.3.2. Model Evaluation Metrics

As usual, a predictive maintenance-oriented binary classification problem based on real operating
machinery data suffers from the highly unbalanced distribution of the classes over the samples.
Failure is a very rare event in the dataset, so only less than 1% of original data belong to class “1” and
more than 99% to class “0”.
Machine learning algorithms do not work properly with this type of dataset, neither do standard
evaluation metrics. In this regard, it is important to emphasize the fact that overall accuracy is a useless
metric of model evaluation for this type of study, although it is the most commonly used evaluation
metrics for classification models.
True positives (TP) and true negatives (TN) are outcomes of the positive class and negative class,
respectively, correctly classified by the model, whereas false positives (FP) and false negatives (FN) are
incorrectly classified outcomes. Overall accuracy (OA) is defined by the Equation (3):

(TP + TN)
Overall Accuracy = (3)
(TP + FP + TN + FN)

Mediated between the two classes, so it is possible for the model to achieve a nearly perfect overall
accuracy by always predicting the majority class.
For highly unbalanced datasets, the best way to judge a model is to use per-class evaluation
metrics such as precision, recall, F1 score (weighted average of precision and recall), expressed as in
Equations (4)–(6), respectively:
TP
Precision = (4)
(TP + FP)
TP
Recall = (5)
(TP + FN)
2·(Presision·Recall)
F1 score = (6)
(Precision + Recall)
Additional significant parameters adopted are Cohen’s Kappa and Area under the curve (AUC)
of the ROC (receiver operating characteristic) Curve.
Since the goal of the study is to predict whether the machine will fail or not in a future interval,
recall and precision over the class “1” are the evaluation metric chosen to be maximized.
Sustainability 2020, 12, 4776 9 of 15

Furthermore, machine learning models training for predictive maintenance-oriented failure

prediction requires a clear understanding of the business requirements and the tolerance to false
positives and false negatives. Some businesses could afford having a high number of the former whilst
some others of the latter, thus the optimization loop aims at maximizing the metric associated with
the requirements.

3. Results
The “Scorer” node outputs confusion matrix and evaluation metrics statistics tables. The performance
of the two adopted algorithms is evaluated by a confusion matrix as illustrated in Tables 1 and 2 for
SVM and MLP algorithm, respectively, where the columns are the predicted class and the rows are the
actual class.

Table 1. Confusion matrix of the adopted dataset using Support Vector Machine algorithm.

Class 0 Class 1
Class 0 27,745 71
Class 1 456 175

Table 2. Confusion matrix of the adopted dataset using Multilayer Perceptron algorithm.

Class 0 Class 1
Class 0 27,684 132
Class 1 366 265

3.1. Results—Support Vector Machine Algorithm

Table 3 reports the evaluation metric for SVM algorithm. It can be noticed that a high value of
overall accuracy equal to 98.1% has been achieved. High values of precision over class “0” and class “1”
have been obtained which corresponds to the portion of positive and negative correct classifications,
respectively. On the other hand, the value of recall over class “1”, which is the percentage of actual
positives identified correctly, is quite low. This means that the algorithm has been penalized by the
high number of false negatives such that it affects the fault predictive performance of the ML model.

Table 3. Evaluation metric for Support Vector Machine algorithm.

Recall Precision F1 Score Accuracy Cohen’s K

Class 0 99.7% 98.4% 99.1% - -
Class 1 27.7% 71.1% 39.9% - -
Overall - - - 98.1% 0.392

Figure 3 reports the timeline of both the actual and predicted classes during the selected period
of collected data. The first row of the timeline represents the predicted values obtained by using the
SVM algorithm, while the second row of the timeline represents the data collected from the sensors
mounted on the centrifugal pump. The figure shows that the selected algorithm allows us to predict
only one of the four failures that occurred and did not produce a false positive.
Sustainability 2020, 12, 4776 10 of 15
Sustainability 2020, 12, x FOR PEER REVIEW 10 of 15

Sustainability 2020, 12, x FOR PEER REVIEW 10 of 15

Figure 3. 3.Timeline
Figure of actual
Timeline of actualandand predicted
predicted classesclasses during
during the theanalysis
selected selected analysis
period period using
using SVM
SVMalgorithm.
algorithm.

3.2. Results—Multilayer Perceptron

3.2. Results—Multilayer Perceptron Algorithm
Algorithm
Figure 3. Timeline
By analyzing of actual
Table andbepredicted classes during the accuracy
selected analysis period using SVM
By analyzing Table 4,4,ititcan
can be noticed
noticed that an
that overall
an overall of 98.2%
accuracy of has
98.2%beenhas
achieved.
been achieved.
algorithm.
Although this value is slightly higher than that obtained with SVM algorithm, avalue
Although this value is slightly higher than that obtained with SVM algorithm, a lower lowerof value of
precision over class “1” is achieved. On the other hand, the use of MLP algorithm provides a high
precision over class “1” isPerceptron
3.2. Results—Multilayer achieved. On the other hand, the use of MLP algorithm provides a high
Algorithm
value of recall over class “1” equal to 42.2%, which corresponds to the percentage of actual positives
valueidentified
of recall over class
correctly.
By analyzing “1”
These
Table equal
4, metrics to 42.2%,
it can beresults
noticed which
arethat corresponds
computed using ato
by accuracy
an overall ofthe
standard percentage
98.2% has been of
threshold actual
upon the positives
achieved.
identified correctly.
MLP output
Although These
thisprobability metrics results
equal tohigher
value is slightly are
0.5. Finally,
than thatcomputed
by comparing by
obtained with using
the two a
SVMcases,standard threshold
it can abelower
algorithm, noticed thatofa the MLP
upon
value
output probability
higher
precisionvalue equal
overofclass
the “1” to
Cohen’s 0.5. Finally,
Kappa
is achieved. is On by other
achieved
the comparing
by using
hand,MLP the
the two cases,
algorithm.
use of MLP In it can
this case,be
algorithm noticed
Cohen’s
provides that a higher
Kappa
a high
isof
0.507
valuevalue ofand
the it represents
recall
Cohen’sover class
Kappa theis
“1” extent
equal toto42.2%,
achieved whichby the
whichprediction
using MLPvalues
corresponds are percentage
to the
algorithm. correct
In representations
thisofcase, of the Kappa is
actualCohen’s
positives
real observed data used.
0.507identified correctly.
and it represents These
the metrics
extent results are
to which the computed
prediction byvalues
using aare standard
correctthreshold upon the of the real
representations
MLP output probability equal to 0.5. Finally, by comparing the two cases, it can be noticed that a
observed data used. Table 4. Evaluation metric for Multilayer Perceptron algorithm.
higher value of the Cohen’s Kappa is achieved by using MLP algorithm. In this case, Cohen’s Kappa
is 0.507 and it represents the extent
RECALL to which the prediction
PRECISION SCOREvalues
F1Multilayer are correct representations
ACCURACY COHEN’S K of the
Table 4. Evaluation metric for Perceptron algorithm.
real observed data
CLASS 0 used.99.5 % 98.7 % 99.1 % ‐ ‐
CLASS 1 42.2 %Recall 66.8 % Precision 51.6 F1% Score Accuracy
‐ Cohen’s
‐ K
Table 4. Evaluation metric for Multilayer Perceptron algorithm.
OVERALL Class 0 ‐ 99.5% ‐ 98.7% ‐ 99.1% 98.2 % - 0.507
-
Class 1RECALL 42.2%PRECISION 66.8% F1 SCORE 51.6% ACCURACY - COHEN’S - K
In this case Overall
too, the timeline- of both the actual
- and -
predicted 98.2%
classes during the 0.507
selected period
CLASS 0 99.5 % 98.7 % 99.1 % ‐ ‐
of collected data 1is reported
CLASS 42.2 % in Figure66.84.
% The MLP 51.6algorithm
% allows
‐ predicting two ‐ of the four
failures
In that occurred
OVERALL
this case too, in the
‐ selected
the timeline of both period providing
‐ the actual ‐ better
and predicting
98.2 classes
predicted % performance
0.507thethan
during SVM period of
selected
algorithm. On the other hand, false positive predicted values are detected. This means that the
collected data is reported in Figure 4. The MLP algorithm allows predicting two of the four failures
algorithm
In thishas
casealso identified
too, a fault
the timeline during
of both thethe normal
actual andoperating
predictedconditions of the
classes during machinery.
the selected period
that occurred
of collectedindata
the is
selected
reportedperiod providing
in Figure 4. The MLP better predicting
algorithm allowsperformance
predicting twothan SVM
of the fouralgorithm.
On the
failures that occurred in the selected period providing better predicting performance than SVM has also
other hand, false positive predicted values are detected. This means that the algorithm
identified a fault
algorithm. Onduring
the otherthehand,
normal
falseoperating conditions
positive predicted of the
values are machinery.
detected. This means that the
algorithm has also identified a fault during the normal operating conditions of the machinery.

Figure 4. Timeline of actual and predicted classes during the selected analysis period using
MLP algorithm.
 Sustainability 2020, 12, x FOR PEER REVIEW 11 of 15

Sustainability 2020, 12, 4776 11 of 15

Figure 4. Timeline of actual and predicted classes during the selected analysis period using MLP
algorithm.
An ROC Curve is another commonly used method of visualizing the performance of a binary
An ROC Curve is another commonly used method of visualizing the performance of a binary
classifier. In this work, the ROC curve of the MLP, our best classifier, is reported. In this case, it results
classifier. In this work, the ROC curve of the MLP, our best classifier, is reported. In this case, it
from varying the decision threshold and plotting the true positive rate against the false positive rate
results from varying the decision threshold and plotting the true positive rate against the false
(see Figure 5).
positive rate (see Figure 5).

1.0 P (included Failure = 1)

0.9 Random

0.8

0.7
TRUE POSITIVE RATE

0.6

0.5

0.4

0.3

0.2

0.1
P (included Failure =1) (0.573)

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
FALSE POSITIVE RATE
Figure 5. ROC
Figure Curve
5. ROC for for
Curve thethe
Multilayer Perceptron
Multilayer algorithm.
Perceptron algorithm.

Analyzing the ROC Curve, as illustrated in Figure 3, the AUC (Area Under Curve) results
Analyzing
equal 0.5726; whilethe it isROC Curve,
possible as illustrated
to notice that by in Figure 3,
changing thethreshold
the AUC (Area Under
value, Curve)
the FP results
number equal
could
0.5726; while
be reduced withoutit isreducing
possiblethe
to notice that by changing the threshold value, the FP number could be
TP number.
reduced without reducing the TP number.
All these obtained metrics values are considered satisfactory when taking various factors
All these obtained metrics values are considered satisfactory when taking various factors into
into consideration:
consideration:
• The failure events register results incomplete and un-standardized.
 The failure events register results incomplete and un‐standardized.
• The dataset includes just four non-ambiguous fault events, which is a very small number.
 The dataset includes just four non‐ambiguous fault events, which is a very small number.
Consequently, the dataset is highly unbalanced between classes and the algorithm training
Consequently, the dataset is highly unbalanced between classes and the algorithm training set
set contains very few examples of class “1”. By using the SMOTE algorithm, classes are
contains very few examples of class “1”. By using the SMOTE algorithm, classes are
balanced—but still the information quality provided by a “synthetic” sample is not of as high a
balanced—but still the information quality provided by a “synthetic” sample is not of as high a
quality as is provided by real data.
quality as is provided by real data.
•  Inadequate
Inadequate sensory
sensoryequipment,
equipment, since thethe
since machine
machine parts areare
parts notnot
equipped
equippedwith thethe
with mostmost
appropriate sensors.
appropriate sensors.
•  Sampling
Samplingfrequency
frequencyhashas
been set set
been to one sample
to one perper
sample hour, providing
hour, a smaller
providing amount
a smaller of data
amount to to
of data
be stored and analyzed.
be stored and analyzed.
While limitations
While inducted
limitations by the
inducted byfirst
thethree
first factors are difficult
three factors to overcome
are difficult in a shortinterm,
to overcome a higher
a short term, a
sampling
higher frequency is easily achievable
sampling frequency and would
is easily achievable andresult in aresult
would largerinamount
a largerofamount
data, which could
of data, which
Sustainability 2020, 12, 4776 12 of 15

improve ML algorithm performances. Moreover, data cleaning and feature extraction and selection
processes will be improved.

4. Conclusions
This paper proposes the preliminary development of a machine learning-supervised algorithm
for the fault diagnosis of rotating machinery in the oil and gas industry. The basic idea is to develop a
simple and easy to implement ML model aiming at performing agile and informed decision-making.
The data come from a real operating centrifugal pump that works within the production line of
SARLUX refinery (Sarroch, Italy). Eight different sensors have been used to build the features: one of
flow rate, two of bearings vibration, two of axial displacement, and three of motor coil temperature.
The adoption of a real dataset strongly influences further decisions of techniques to use for feature
engineering, data labeling and machine learning techniques. Two different algorithms are used and
compared—the Support Vector Machine (SVM), and the Multilayer Perceptron (MLP).
The main purpose of this work is not based on finding a highly and extremely accurate ML model,
but rather to show how, with a simple and intuitive ML algorithm, it is possible to have good
forecast results. Indeed, the results show that the proposed algorithms present good overall classification
performances, achieving a good capability to identify the health status of the monitored machine.
The SVM algorithm shows higher precision than the MLP but lower recall over positive class,
while MLP shows better classification performance, predicting two of the four failures that occurred
in the selected period. Thus, based on these promising algorithm assessments, a larger scale
experiment series is planned, aiming at reducing maintenance costs by the optimal scheduling
of sustainable maintenance actions and increasing the life expectancy of the centrifugal pump.
To this end, availability-based maintenance planning procedures and the associated costs might be
another worthwhile step. Future work will be aimed at improving the overall performance metric by
both having more robust data-set and considering a different set of features to improve and make the
results generally applicable—or at least reduce possible bias. Finally, additional work is needed to
validate the proposed method by considering different industrial components.

Author Contributions: Conceptualization, P.F.O. and R.C.; methodology, S.A., A.Z. and C.M.; validation, A.Z.,
C.M. and L.S.; investigation, S.A. and A.Z.; data curation, L.S.; writing—original draft preparation, S.A., A.Z.,
and C.M.; writing—review and editing, All Authors; supervision, P.F.O., L.S. and R.C. All authors have read and
agreed to the published version of the manuscript.
Funding: This research received no external funding.
Acknowledgments: Zoccheddu A. gratefully acknowledges Saras Ricerche e Tecnologie Srl for the internship
opportunity. The authors wish to thank SARLUX Refinery (Sarroch, Italy) for kindly supplying the data used in
the paper.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Susto, G.A.; Schirru, A.; Pampuri, S.; McLoone, S.; Beghi, A. Machine Learning for Predictive Maintenance:
A Multiple Classifier Approach. IEEE Trans. Ind. Inform. 2015, 11, 812–820. [CrossRef]
2. Carvalho, T.P.; Soares, F.A.A.M.N.; Vita, R.; Francisco, R.d.P.; Basto, J.P.; Alcalá, S.G.S. A systematic literature
review of machine learning methods applied to predictive maintenance. Comput. Ind. Eng. 2019, 137.
[CrossRef]
3. Li, K.; Xue, Y.; Cui, S.; Niu, Q. Intelligent Computing in Smart Grid and Electrical Vehicles; Springer: Cham,
Switzerland, 2014; Volume 463, ISBN 9783662452851.
4. Kostrzewski, M.; Varjan, P.; Gnap, J. Solutions dedicated to internal logistics 4.0. In Sustainable Logistics and
Production in Industry 4.0; Springer: Cham, Switzerland, 2020; pp. 243–262. ISBN 978-3-030-33368-3.
5. Alpaydin, E. Introduction to Machine Learning, 2nd ed.; The MIT Press: Cambridge, MA, USA, 2010;
ISBN 9780262012430.
Sustainability 2020, 12, 4776 13 of 15

6. Pham, D.T.; Afify, A.A. Machine-learning techniques and their applications in manufacturing. Proc. Inst.
Mech. Eng. Part B J. Eng. Manuf. 2005, 219, 395–412. [CrossRef]
7. Lei, Y.; Yang, B.; Jiang, X.; Jia, F.; Li, N.; Nandi, A.K. Applications of machine learning to machine fault
diagnosis: A review and roadmap. Mech. Syst. Signal Process. 2020, 138, 106587. [CrossRef]
8. Diez-Olivan, A.; Del Ser, J.; Galar, D.; Sierra, B. Data fusion and machine learning for industrial prognosis:
Trends and perspectives towards Industry 4.0. Inf. Fus. 2019, 50, 92–111. [CrossRef]
9. Costello, J.J.A.; West, G.M.; McArthur, S.D.J. Machine learning model for event-based prognostics in gas
circulator condition monitoring. IEEE Trans. Reliab. 2017, 66, 1048–1057. [CrossRef]
10. Han, Y.; Zhao, D.; Hou, H. Oil-immersed Transformer Internal Thermoelectric Potential Fault Diagnosis
Based on Decision-tree of KNIME Platform. Proced. Comput. Sci. 2016, 83, 1321–1326. [CrossRef]
11. Czwajda, L.; Kosacka-Olejnik, M.; Kudelska, I.; Kostrzewski, M.; Sethanan, K.; Pitakaso, R. Application of
Prediction Markets Phenomenon as Decision Support Instrument in Vehicle Recycling Sector. LogForum
2019, 15, 265–278. [CrossRef]
12. Li, H.; Parikh, D.; He, Q.; Qian, B.; Li, Z.; Fang, D.; Hampapur, A. Improving rail network velocity: A machine
learning approach to predictive maintenance. Transp. Res. Part C Emerg. Technol. 2014, 45, 17–26. [CrossRef]
13. Allah Bukhsh, Z.; Saeed, A.; Stipanovic, I.; Doree, A.G. Predictive maintenance using tree-based classification
techniques: A case of railway switches. Transp. Res. Part C Emerg. Technol. 2019, 101, 35–54. [CrossRef]
14. Sahal, R.; Breslin, J.G.; Ali, M.I. Big data and stream processing platforms for Industry 4.0 requirements
mapping for a predictive maintenance use case. J. Manuf. Syst. 2020, 54, 138–151. [CrossRef]
15. Hajizadeh, Y. Machine learning in oil and gas; a SWOT analysis approach. J. Pet. Sci. Eng. 2019, 176, 661–663.
[CrossRef]
16. Hanga, K.; Kovalchuk, Y. Machine Learning and Multi-Agent Systems in Oil and Gas Industry Applications:
A Survey. Comput. Sci. Rev. 2019, 34, 100191. [CrossRef]
17. Trout, J.N.; Kolodziej, J.R. Reciprocating compressor valve condition monitoring using image-based pattern
recognition. In Proceedings of the Annual Conference of the Prognostics and Health Management Society,
Denver, CO, USA, 6 October 2016.
18. Deutsch, J.; He, D. Using Deep Learning-Based Approach to Predict Remaining Useful Life of Rotating
Components. IEEE Trans. Syst. Man Cybern. Syst. 2017, 48, 11–20. [CrossRef]
19. Kwak, D.S.; Kim, K.J. A data mining approach considering missing values for the optimization of
semiconductor-manufacturing processes. Expert Syst. Appl. 2012, 39, 2590–2596. [CrossRef]
20. Qian, W.; Li, S.; Yi, P.; Zhang, K. A novel transfer learning method for robust fault diagnosis of rotating
machines under variable working conditions. Meas. J. Int. Meas. Confed. 2019, 138, 514–525. [CrossRef]
21. Zhang, W.; Li, X.; Ding, Q. Deep residual learning-based fault diagnosis method for rotating machinery.
ISA Trans. 2018. [CrossRef]
22. Su, L.; Li, K.; Xiong, M.; Li, F.; Wu, J. A novel fault diagnosis algorithm for rotating machinery based on a
sparsity and neighborhood preserving deep extreme learning machine. Neurocomputing 2019, 350, 261–270.
23. Qian, W.; Li, S.; Wang, J.; Wu, Q. A novel supervised sparse feature extraction method and its application on
rotating machine fault diagnosis. Neurocomputing 2018, 320, 129–140. [CrossRef]
24. Zhang, Z.; Li, S.; Wang, J.; Xin, Y.; An, Z. General normalized sparse filtering: A novel unsupervised learning
method for rotating machinery fault diagnosis. Mech. Syst. Signal Process. 2019, 124, 596–612. [CrossRef]
25. Yang, B.; Lei, Y.; Jia, F.; Xing, S. An intelligent fault diagnosis approach based on transfer learning from
laboratory bearings to locomotive bearings. Mech. Syst. Signal Process. 2019, 122, 692–706. [CrossRef]
26. Bilski, P. Application of Support Vector Machines to the induction motor parameters identification. Meas. J.
Int. Meas. Confed. 2014, 51, 377–386. [CrossRef]
27. Jirdehi, M.A.; Rezaei, A. Parameters estimation of squirrel-cage induction motors using ANN and ANFIS.
Alex. Eng. J. 2016, 55, 357–368. [CrossRef]
28. Romeo, L.; Loncarski, J.; Paolanti, M.; Bocchini, G.; Mancini, A.; Frontoni, E. Machine learning-based design
support system for the prediction of heterogeneous machine parameters in industry 4.0. Expert Syst. Appl.
2020, 140. [CrossRef]
29. Giantomassi, A.; Ferracuti, F.; Iarlori, S.; Ippoliti, G.; Longhi, S. Electric Motor Fault Detection and Diagnosis
by Kernel Density Estimation and Kullback-Leibler Divergence based on Stator Current Measurements.
IEEE Trans. Ind. Electron. 2015, 62, 1770–1780. [CrossRef]
Sustainability 2020, 12, 4776 14 of 15

30. Paolanti, M.; Romeo, L.; Felicetti, A.; Mancini, A.; Frontoni, E.; Loncarski, J. Machine learning approach for
predictive maintenance in industry 4.0. In Proceedings of the 14th IEEE/ASME International Conference on
Mechatronic and Embedded Systems and Applications (MESA), Oulu, Finland, 2–4 July 2018; pp. 1–6.
31. Yang, Y.; Zheng, H.; Li, Y.; Xu, M.; Chen, Y. A fault diagnosis scheme for rotating machinery using hierarchical
symbolic analysis and convolutional neural network. ISA Trans. 2019, 91, 235–252. [CrossRef]
32. Pang, S.; Yang, X.; Zhang, X.; Lin, X. Fault diagnosis of rotating machinery with ensemble kernel extreme
learning machine based on fused multi-domain features. ISA Trans. 2019. [CrossRef]
33. Chen, Z.; Gryllias, K.; Li, W. Mechanical fault diagnosis using Convolutional Neural Networks and Extreme
Learning Machine. Mech. Syst. Signal Process. 2019, 133, 106272. [CrossRef]
34. Zhang, X.L.; Chen, W.; Wang, B.J.; Chen, X.F. Intelligent fault diagnosis of rotating machinery using support
vector machine with ant colony algorithm for synchronous feature selection and parameter optimization.
Neurocomputing 2015, 167, 260–279. [CrossRef]
35. Wang, X.B.; Zhang, X.; Li, Z.; Wu, J. Ensemble extreme learning machines for compound-fault diagnosis of
rotating machinery. Knowl.-Based Syst. 2019. [CrossRef]
36. Panda, A.K.; Rapur, J.S.; Tiwari, R. Prediction of flow blockages and impending cavitation in centrifugal
pumps using Support Vector Machine (SVM) algorithms based on vibration measurements. Meas. J. Int.
Meas. Confed. 2018, 130, 44–56. [CrossRef]
37. Wang, Z.Y.; Lu, C.; Zhou, B. Fault diagnosis for rotary machinery with selective ensemble neural networks.
Mech. Syst. Signal Process. 2018, 113, 112–130. [CrossRef]
38. Liu, R.; Yang, B.; Zio, E.; Chen, X. Artificial intelligence for fault diagnosis of rotating machinery: A review.
Mech. Syst. Signal Process. 2018, 108, 33–47. [CrossRef]
39. Tang, S.; Shen, C.; Wang, D.; Li, S.; Huang, W.; Zhu, Z. Adaptive deep feature learning network with
Nesterov momentum and its application to rotating machinery fault diagnosis. Neurocomputing 2018, 305,
1–14. [CrossRef]
40. Yu, S.; Zhao, D.; Chen, W.; Hou, H. Oil-immersed Power Transformer Internal Fault Diagnosis Research
Based on Probabilistic Neural Network. Proced. Comput. Sci. 2016, 83, 1327–1331. [CrossRef]
41. Zenisek, J.; Holzinger, F.; Affenzeller, M. Machine learning based concept drift detection for predictive
maintenance. Comput. Ind. Eng. 2019, 137. [CrossRef]
42. Lee, W.J.; Wu, H.; Yun, H.; Kim, H.; Jun, M.B.G.; Sutherland, J.W. Predictive maintenance of machine tool
systems using artificial intelligence techniques applied to machine condition data. Proced. CIRP 2019, 80,
506–511. [CrossRef]
43. Guedes, A.S.; Silva, S.M.; Cardoso Filho, B.d.J.; Conceição, C.A. Evaluation of electrical insulation in
three-phase induction motors and classification of failures using neural networks. Electr. Power Syst. Res.
2016, 140, 263–273. [CrossRef]
44. Wuest, T.; Weimer, D.; Irgens, C.; Thoben, K.D. Machine learning in manufacturing: Advantages, challenges,
and applications. Prod. Manuf. Res. 2016, 4, 23–45. [CrossRef]
45. Ruiz-Sarmiento, J.R.; Monroy, J.; Moreno, F.A.; Galindo, C.; Bonelo, J.M.; Gonzalez-Jimenez, J. A predictive
model for the maintenance of industrial machinery in the context of industry 4.0. Eng. Appl. Artif. Intell.
2020, 87, 103289. [CrossRef]
46. Ruiz-Sarmiento, J.R.; Galindo, C.; Gonzalez-Jimenez, J. Building multiversal semantic maps for mobile robot
operation. Knowl.-Based Syst. 2017, 119, 257–272. [CrossRef]
47. Burkov, A. The Hundred-Page Machine Learning Book, 1st ed.; Publishing Kindle Direct: Seattle, WA, USA, 2019.
48. Jahnke, P. Machine Learning Approaches for Failure Type Detection and Predictive Maintenance. Master’s
Thesis, Technische Universität Darmstadt, Darmstadt, Germany, 19 June 2015.
49. Lepot, M.; Aubin, J.B.; Clemens, F.H.L.R. Interpolation in time series: An introductive overview of existing
methods, their performance criteria and uncertainty assessment. Water 2017, 9, 796. [CrossRef]
50. Hastie, T.; Tibshirani, R.; Friedman, J. The Elements of Statistical Learning; Springer: Berlin, Germany, 2017.
51. Chawla, N.V.; Bowyer, K.W.; Hall, L.O.; Kegelmeyer, W.P. SMOTE: Synthetic Minority Over-Sampling
Technique. Journal of Artificial Intelligence Research. J. Artif. Intell. Res. 2002, 16, 321–357. [CrossRef]
52. Ha, T.M.; Bunke, H. Off-line, handwritten numeral recognition by perturbation method. IEEE Trans. Pattern
Anal. Mach. Intell. 1997, 19, 535–539. [CrossRef]
53. Cortes, C.; Vapnik, V.N. Support-Vector networks. Mach. Learn. 1995, 20, 273–297. [CrossRef]
Sustainability 2020, 12, 4776 15 of 15

54. Vapnik, V.N. Statistical Learning Theory; John Wiley & Sons Inc.: New York, NY, USA, 1998.
55. Riedmiller, M.; Braun, H. A direct adaptive method for faster backpropagation learning: The RPROP
algorithm. In Proceedings of the IEEE International Conference on Neural Networks, San Francisco, CA,
USA, 28 March–1 April 1993; pp. 586–591.

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).