Available online at www.sciencedirect.

com
Available online at www.sciencedirect.com

ScienceDirect
ScienceDirect
Available online at www.sciencedirect.com

Procedia Manufacturing 00 (2019) 000–000

ScienceDirect
Procedia Manufacturing 00 (2019) 000–000 www.elsevier.com/locate/procedia
www.elsevier.com/locate/procedia

Procedia Manufacturing 42 (2020) 450–456

International
International Conference
Conference on
on Industry
Industry 4.0
4.0 and
and Smart
Smart Manufacturing
Manufacturing (ISM
(ISM 2019)
2019)

A
A Real-Time Condition Monitoring
Real-Time Condition Monitoring and
and Maintenance
Maintenance Management
Management System
System
for Low
for Low Voltage
Voltage Industrial
Industrial Motors
Motors Using
Using Internet-of-Things
Internet-of-Things
Konstantinos
Konstantinos Mykoniatis
Mykoniatis
Auburn University, Alabama, LA, 36832
Auburn University, Alabama, LA, 36832
,
,
* Corresponding author. E-mail address: [email protected]
* Corresponding author. E-mail address: [email protected]

Abstract
Abstract
Maintenance and sound operating industrial equipment are critical for any manufacturing company. Standardization of the
Maintenance and sound operating industrial equipment are critical for any manufacturing company. Standardization of the
manufacturing infrastructure and establishment of a systematic maintenance program is essential for this process. However,
manufacturing infrastructure and establishment of a systematic maintenance program is essential for this process. However,
condition monitoring must also be an integral part of a smart manufacturing program that seeks to improve and optimize the
condition monitoring must also be an integral part of a smart manufacturing program that seeks to improve and optimize the
operational efficiency of production systems. Absence of this type of data-driven observations and mindset in manufacturing
operational efficiency of production systems. Absence of this type of data-driven observations and mindset in manufacturing
decision making, may result in safety risks, missing critical signs, or occurrence of unexpected repairs that could bring the
decision making, may result in safety risks, missing critical signs, or occurrence of unexpected repairs that could bring the
equipment to a halt. Moreover, routine maintenance may short the actual useful life of some equipment and when scheduled
equipment to a halt. Moreover, routine maintenance may short the actual useful life of some equipment and when scheduled
maintenance shutdowns occur too often, they increase downtime and cost.
maintenance shutdowns occur too often, they increase downtime and cost.
This paper presents the design and development of a real-time condition monitoring system for managing industrial low voltage
This paper presents the design and development of a real-time condition monitoring system for managing industrial low voltage
motors using internet-of-things. The system can record and monitor vibration and temperature conditions of an industrial motor
motors using internet-of-things. The system can record and monitor vibration and temperature conditions of an industrial motor
and transmit the data through a wireless network to a data logging center. The current prototype was developed using open source
and transmit the data through a wireless network to a data logging center. The current prototype was developed using open source
software and hardware and can successfully identify abnormal motor conditions from sensor input values that exceed predefined
software and hardware and can successfully identify abnormal motor conditions from sensor input values that exceed predefined
setpoints. When a motor is approaching an abnormal condition, the system changes state and informs the user through a mobile
setpoints. When a motor is approaching an abnormal condition, the system changes state and informs the user through a mobile
alert. The motor management system requires the user to perform an RFID enforced inspection. The alarm system remains active
alert. The motor management system requires the user to perform an RFID enforced inspection. The alarm system remains active
until authorized personnel visits the equipment and scans the dedicated RFID batch. Furthermore, the motor management system
until authorized personnel visits the equipment and scans the dedicated RFID batch. Furthermore, the motor management system
can be accessed remotely to allow the user to visualize the current condition of the motor. Finally, we present a pilot experiment
can be accessed remotely to allow the user to visualize the current condition of the motor. Finally, we present a pilot experiment
that was conducted to test the condition monitoring capabilities of the prototype and we discuss future steps in order to further
that was conducted to test the condition monitoring capabilities of the prototype and we discuss future steps in order to further
develop the current system to a smart predictive maintenance system capable of detecting and predicting specific motor faults.
develop the current system to a smart predictive maintenance system capable of detecting and predicting specific motor faults.
© 2020
© 2020The TheAuthors.
Authors. Published
Published by Elsevier
by Elsevier B.V. B.V.
© 2020
This The Authors. Published by Elsevier B.V.BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
This isisan
anopen
open access
access article
article under
under the CCthe BY-NC-ND
CC license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
This is an open
Peer-reviewunder access article under
underresponsibility
responsibility the CC BY-NC-ND license of
(http://creativecommons.org/licenses/by-nc-nd/4.0/)
Peer-review ofof
thethe scientific
scientific committee
committee of the the International
International Conference
Conference on Industry
on Industry 4.0Smart
4.0 and and Smart Manufacturing.
Manufacturing.
Peer-review under responsibility of the scientific committee of the International Conference on Industry 4.0 and Smart Manufacturing.
Keywords: Industrial Motors; Condition Monitoring; Internet of Things; Temperature; Vibration; RFID, Notifications
Keywords: Industrial Motors; Condition Monitoring; Internet of Things; Temperature; Vibration; RFID, Notifications

1. Introduction lack of justification for upgrading, or because they cannot

1. Introduction lack of justification for upgrading, or because they cannot
afford to have their processes shut down for upgrades [2]. One
afford to have their processes shut down for upgrades [2]. One
As technology continues to evolve into the 21st century, important area that is impacted by this is industrial motors.
As technology continues to evolve into the 21st century, important area that is impacted by this is industrial motors.
industrial equipment and machines that were installed ten to Low voltage motors are an essential part of manufacturing
industrial equipment and machines that were installed ten to Low voltage motors are an essential part of manufacturing
twenty years ago are becoming outdated. While larger plants systems that are used in a wide range of industrial machines and
twenty years ago are becoming outdated. While larger plants systems that are used in a wide range of industrial machines and
are more likely to employ advanced manufacturing technology applications such as oil, paper, food, pharmaceuticals, water,
are more likely to employ advanced manufacturing technology applications such as oil, paper, food, pharmaceuticals, water,
than smaller ones [1], there are still industries that use outdated and packaging. Every motor is designed to operate under
than smaller ones [1], there are still industries that use outdated and packaging. Every motor is designed to operate under
machines due to sunk costs and uncertain market conditions, specific conditions. Industrial low voltage motors use a specific
machines due to sunk costs and uncertain market conditions, specific conditions. Industrial low voltage motors use a specific
2351-9789 © 2020 The Authors. Published by Elsevier B.V.
2351-9789© 2020 The
This is an open Authors.
access Published
article underby Elsevier
the B.V.
CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
2351-9789© 2020 The Authors. Published by Elsevier B.V.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)Peer-review under responsibility of the
Peer-review
This is an open under responsibility
access article of the
under the CC scientificlicense
BY-NC-ND committee of the International Conference on Industry 4.0 and
(http://creativecommons.org/licenses/by-nc-nd/4.0/)Peer-review underSmart Manufacturing.
responsibility of the
scientific committee of the International Conference on Industry 4.0 and Smart Manufacturing.
10.1016/j.promfg.2020.02.050
scientific committee of the International Conference on Industry 4.0 and Smart Manufacturing.
 Konstantinos Mykoniatis et al. / Procedia Manufacturing 42 (2020) 450–456 451
2 Author name /Procedia Manufacturing 00 (2019) 000–000

amount of energy, have a certain operating temperature based The remainder of the paper is organized as follows: Section
on their insulation system, vibrate at certain frequencies and are 2 describes the IoT prototype including the software and
capable of handling load according to the peak torque they can hardware architecture as well as the MMS implemented
provide. Any change in these factors may indicate a pending algorithms and control logic of it. Section 3 provides research
failure or the need for inspection and maintenance. methodology initial experimental results of the developed
Mechanical failures are ranked as the most frequently prototype. Finally, in Section 4 we discuss future work and
occurred cause for stoppages, breakdowns and catastrophic conclusions.
accidents among 100 petrochemical plants world-wide [3], [4].
Previous work on several areas of application has shown that 2. Prototype Design Description And Implementation
monitoring and analyzing key performance indicators and
physical parameters associated with operating machines results This section describes the design process of the MMS,
in increased safety and usage of life limited parts, reduction in including the implemented algorithms, the hardware
unscheduled maintenance and manual inspection as well as architecture and the prototype implementation. The MMS can
reduction in fault not found [5], [6], [7], [8], [9], [10]. track and record temperature and standard vibration measures
In addition, the associated costs of problematic components of industrial motors. If a motor reaches the predefined set point
can be reduced when they can be detected prior to a failure and of temperature or detects abnormal vibration frequencies, then
when on-time maintenance is performed with minimum the alarm system is activated, and the system notifies the
disruption to productivity. Several surveys that analyzed the operator via an email for the current motor condition and its
impact of condition monitoring maintenance management on corresponding state. In addition, the alarm will not be
the economics of manufacturing industry in America, Canada, dismissed unless an operator arrives at the motor location for
Europe, and Australia reveal its significant role in ensuring inspection and scans his/her dedicated RFID batch on the RFID
continuing prosperity [11]. More specifically, through these reader which is attached to the IoT device.
studies it was found out that condition-based maintenance
caused: 50-80 % reductions in repair costs, 30% increase in
revenue, 50-80 % reduction in maintenance costs, 30% 2.1. Software
reduction in spares inventories, and 20-60 % increase in overall
profitability of plants [11]. Several algorithms were developed and integrated, such as
Vital utilities that are accomplished by the Internet-of- for capturing the most dominant vibration frequency using a
Things (IoT) applications, in both domestic [12], [13], and Fast Fourier Transformation (FFT) algorithm, for notifying the
industrial settings, are monitoring, and hence prompt decision operator via an email about the motor’s condition, and for
making for efficient management [14]. The McKinsey Global enforcing motor inspection through an RFID batch to dismiss
Institute reports that predictive maintenance using IoT the alarm.
technology was estimated to be capable of saving Due to memory and sampling rate limitations of the Arduino
approximately $240-$627 billion a year by 2025 [15]. This is a microcontroller, an FFT algorithm [16] was integrated in the
cost savings estimate of 10-40% per year. Considering more system to be able to capture the most dominant frequency, and
sophisticated and expensive machines that are being used a signal generator was used to adjust the frequency, waveforms,
nowadays makes it easier for companies that use outdated offset, and amplitude of standard vibration signals.
technology to justify condition monitoring expenditure. Prior to the completion of the MMS activation the user must
This paper presents the design and development of a low- define the maximum allowable temperature and vibration range
cost real-time condition monitoring system for managing depending on the low voltage motor’s specifications and the
industrial low voltage motors using IoT. The current IoT desired safety factors. The temperature boundaries are specified
prototype can record and monitor vibration and temperature in 3 different states: stable, warning and alarm. Once the user
conditions of industrial motors and transmit data through a has manually input the temperature and vibration thresholds the
wireless network to a data logging center. The Motor vibration and temperature sensors scan for input data, send
Management System (MMS) IoT device can identify abnormal them to a data logging center (a computer), and using a python
motor conditions from sensor input values that exceed script program are stored in a csv file for further analysis.
predefined setpoints. The temperature and vibration frequency Contingent upon the vibration frequency and temperature
are prime indicators of a motor’s operating performance. When values, the color of the RGB LED, which is placed on the MMS,
a motor is approaching an abnormal condition, the system changes. The green LED indicates a stable state, the yellow
changes state and informs the user through a mobile alert to LED color indicates a warning state and the red LED color
perform an RFID enforced inspection. The alarm system indicates an alarm state. When a motor operates at a stable
remains activated until an inspector visits the equipment. The vibration frequency and regular temperature level, within the
MMS can be accessed remotely to allow the user to examine predefined thresholds, the green light of the LED remains on.
the current condition and behavior of the motor. Furthermore, a If a motor’s vibration frequency deviates from the most
pilot experiment was conducted to test the condition monitoring dominant frequency, due to drastic changes or spikes that may
capabilities of the prototype. Finally, the author discusses about occur, or if the motor’s temperature exceeds the stable
future plans in order to further develop the current system to a temperature boundaries and falls within the warning or alarm
smart predictive motor management and maintenance system. state boundaries, the MMS alarm mechanism is activated and
452 Konstantinos Mykoniatis et al. / Procedia Manufacturing 42 (2020) 450–456
Author name / Procedia Manufacturing 00 (2019) 000–000 3

an email notification about the motor’s condition and state is

sent to the operator.
Another algorithm waits for input in order to allow the alert
dismissal. The system for the alert dismissal consists of an
RFID badge and a reader. Once the alarm state is activated, the
red LED and buzzer remains on until an RFID badge is swiped
on the RFID reader. The RFID reader is used for security
reasons, meaning only authorized operators would have access
to the badges. The RFID enforced inspection is an additional
security measure to prevent the alarm state from being ignored
or purposely overridden by unauthorized personnel. This
process ensures that the machinery will not be running at a high
risk or in a damaged state without knowledge. In addition, the
RFID enforced inspection aims to reduce regular maintenance
and is anticipated to increase manager’s awareness. The overall
behavior of the MMS is described using a process flow diagram Fig. 2. UML Diagram for the MMS
(Fig. 1). Fig.2 presents a UML use case diagram that illustrates
the events that occur, and functionalities offered by the MMS
in order for the user to be informed about the motors’ condition. 2.2. Hardware Architecture

This subsection describes the configuration of the physical

components of the industrial MMS and their interrelationships.
The configuration of the hardware consists of an XBee
shield mounted on top of an Arduino microcontroller. A non-
contact infrared temperature sensor (MLX90614) and a
thermistor are connected to the Xbee shield as input sensor
devices to provide information about the temperature
conditions. An accelerometer-gyroscope sensor (GY521) is
used as an input for monitoring the motor vibration frequency
level. Two XBees S2C radio frequency communication
modules were used to send and receive signals for all the mobile
notifications as well as to allow wireless connectivity between
the MMS and a computer logging station.
The XBee modules were tested, paired and configured
through XCTU, an open-access multi-platform application that
allow developers to set-up, configure and test the XBee
modules [17]. The one XBee module was configured as an
output device to communicate and transmit real-time data of the
input sensors to another XBee receiver module. The XBee
receiver was setup as an input connected into a stationary
computer workstation that was used as a central point for data
logging and analysis.
In addition, the MMS is equipped with an RFID reader, an
RGB LED and a passive buzzer to further increase situation
awareness of the operator on the shop floor. The LED and the
sound alarm are used as outputs for extra precaution to indicate
the current state of the motor and provide visual and audio
feedback when critical or abnormal conditions are detected. An
authorized individual can manually turn off the alarm by
swiping the RFID card onto the RFID reader. The architecture
diagram (Fig. 3) illustrates an internal view of the system’s
architecture in order to understand the different components of
Fig. 1. High Level Process flow diagram of the MMS the MMS components and how they interact.
 Konstantinos Mykoniatis et al. / Procedia Manufacturing 42 (2020) 450–456 453
4 Author name /Procedia Manufacturing 00 (2019) 000–000

3. Methodology and Pilot Testing Results

This section describes the experimental methodology that

was followed as well as the pilot testing results of the MMS
prototype’s functionality under real-time operating conditions.
The prototype was tested on two industrial low voltage motors.
The characteristics of the two motors are illustrated in Table 1
and Table 2, respectively.

Table 1. Characteristics of Motor I.

Motor I
Manufacturer Marathon Electric
Model AVH 364TTDC4036AAS
Type TDC
Fig. 3. Motor Management System Architecture
Class B
Horsepower (H.P.) 60
2.3. Prototype Description Volts (V) 230/460
F.L. (Amps) 142/7.1
The MMS prototype has been developed to improve overall
Frequency (Hz) 60
equipment effectiveness, support decision making of
maintenance management, as well as to allow real-time Service Factor (SF) 1.15
condition monitoring of industrial machines equipped with low Speed (RPM) 1775
voltage motors. The MMS exhibits the following NEMA Nominal Efficiency (%) 93.6
functionalities: Nominal Power Factor (%) 84.5

• Ability to monitor and store vibration and temperature data

Table 2. Characteristics of Motor II.
• Program to set temperature and vibration limits
• Ability to detect abnormal motor conditions Motor II

• Alarm system through light and audible indicator Manufacturer Emerson Motor Co.
• Mobile alert through push notifications. Model HJ40S2BV-P
• RFID security and enforced inspection control Type CTF1
• Wireless connectivity Class F
• Remote access to view motor’s condition through a GUI
Horsepower (H.P.) 40
The system components of the MMS IoT device are Volts (V) 230/460
illustrated in Fig. 4. F.L. (Amps) 46-92
Frequency (Hz) 60
Service Factor (SF) 1.15
Speed (RPM) 1775
NEMA Nominal Efficiency (%) 92.4
Nominal Power Factor (%) 87.4

3.1. Temperature Testing

A non-contact infrared sensor and a thermistor were used for

measuring the temperature on the surface of each motor. Prior
to mounting the thermistor on the motor, the electrode and its
contacts were inspected to ensure they are free of scratches,
Fig. 4. MMS Components & Prototype Representation grease or oil. The sensors were mounted on “hot-spot” motor
locations, which tend to warm up faster than the rest of the
Apart from the visual components, the prototype includes a surface [18], in order to monitor potential issues in the stator
python script program for sending notifications and storing core of the motor. Masking tape was used to attach the sensors
information in a database for analysis of potential fail indicators on the motor and prevent movement due to mechanical shock
and a Graphical User Interface that allows the user to visualize and vibration. The threshold of the temperature sensors was
the motor condition. programmed to signal a warning based on the typical surface
temperature which results from the insulation hot-spot
classification of each motor. The hot-spot insulation class rating
454 Konstantinos Mykoniatis et al. / Procedia Manufacturing 42 (2020) 450–456
Author name / Procedia Manufacturing 00 (2019) 000–000 5

for each type of motor is defined by the maximum ambient 3.2. Vibration Testing
temperature (40° C) combined with the maximum temperature
rise that occurs when a motor is normally operated at full load Some of the well-known causes of vibration anomalies occur
to yield an average insulation life of 20,000 hours [19]. Table 3 due to electromagnetic or mechanical imbalance, loose
illustrates the insulation hot-spot class rating for 20,000 hours components, parts rubbing, motor resonance, or bearing failure
of average insulation life and the corresponding typical [20]. Since one of the most common problems for motor failure
temperature on the surface of the motor. is due to bearing related problems [21], [22], an accelerometer
was placed directly on the motor’s bearing surface. Fig 6.
Table 3. Class rating for 20,000 hours of average insulation life [19] . Illustrates the locations of the three main sensors attached on
Insulation Hot- Spot Insulation Hot-Spot Motor Surface the motors.
Class Temp Temp
A 105° C 80° C
B 130° C 105° C
F 155° C 130° C
H 180° C 155° C

Choosing a motor with an insulation class higher than

necessary can improve and prolong the motor life and become
more tolerant of other problems that typically shorten motor life
[19]. Since class B and F motors were used in this experiment Fig. 6. The MMS Prototype mounted on Marathon motor
105° C and 130° C must be the selected thresholds at which the
temperature sensors must signal a warning if exceeded. In order to test the functionality of the vibration alarm program
In order to verify the functionality of the temperature alarm and configure the alarm thresholds, vibration magnitude data
program, the class B and the class F motors were tested, and associated with the two motors were collected three times
512 and 240 temperature data were collected, respectively. Fig. respectively under stable operating conditions. The sampling
5. graphically depicts the box-and-whisker plots of the frequency was set up to 1000 hertz with rate of 128 samples per
collected temperature data for motor class B and class F. 2.14 seconds for each run. To combine the datasets from the
individual runs and assess the practical and statistical
significance, one-way analysis of variance was performed using
Tukey and Fisher multiple comparison procedures with 95%
confidence interval for the means. In the results tables of Tukey
and Fisher, means that do not share a letter will be significantly
different. Tables 4-7 summarize the results, which show that the
differences between the means were not statistically significant.
Therefore, the datasets were combined into single datasets for
(a) each motor.

Table 4. Motor B Grouping Information Using the Tukey Method and 95%
Confidence
Factor N Mean Grouping
Data Run1 Motor (B) 10625 2437.16 A
Data Run3 Motor (B) 10625 2433.44 A
Data Run2 Motor (B) 10625 2425.23 A

(b) Table 5. Motor B Grouping Information Using the Fisher LSD Method and
95% Confidence
Fig. 5. Box-and-Whisker plots of temperature data for (a) motor class B; (b) Factor N Mean Grouping
motor class F
Data Run1 Motor (B) 10625 2437.16 A
Data Run3 Motor (B) 10625 2433.44 A
Based on the collected data the stable state was set below the
75 percentile, the warning state range was defined between the
th Data Run2 Motor (B) 10625 2425.23 A
75th and 95th percentiles, and the alarm state above the 95th
percentile. The alarm program produced the desired output for
each state successfully.
 Konstantinos Mykoniatis et al. / Procedia Manufacturing 42 (2020) 450–456 455
6 Author name /Procedia Manufacturing 00 (2019) 000–000

Table 6. Motor F Grouping Information Using the Tukey Method and 95%
Confidence
Factor N Mean Grouping
Data Run 2 Motor (F) 10160 16933.9 A
Data Run 3 Motor (F) 10160 16926.5 A
Data Run 1 Motor (F) 10160 16903.9 A

Table 7. Motor F Grouping Information Using the Fisher LSD Method and
95% Confidence
Factor N Mean Grouping
Data Run 2 Motor (F) 10160 16933.9 A
Data Run 3 Motor (F) 10160 16926.5 A
Data Run 1 Motor (F) 10160 16903.9 A (a)

The combined datasets for each motor were examined and

analyzed for multiple change points using a nonparametric
approach proposed by Matteson and James [23]. Table 8
summarizes the inputs to the algorithm, with the numbers in the
brackets to be the ones considered in performing the algorithm.

Table 8. Inputs to nonparametric algorithm for multiple change point analysis

(b)
of multivariate data
Input Description Fig. 7. Box-and-Whisker plots of vibration frequency data for (a)
motor class B; (b) motor class F
X A T x d matrix containing the length T time series
with d-dimensional observations.
Based on the collected data for the two datasets the warning
sig.lvl The level at which to sequentially test if a proposed
change point is statistically significant. The level of
state for motor B and F were defined to be triggered above the
[0.05, 0.9] 95th percentile. The vibration alarm program produced the
significance was tested for values between 0.05 and
0.9, with no change in the results. desired output for each state successfully.
R The maximum number of random permutations to
[499]
use in each iteration of the permutation test. The 4. Conclusions and Future Work
permutation test p-value is calculated using the
method outlined by Gandy [24] . This paper presented the design and development of a real-
min.size Minimum number of observations between change time condition monitoring and maintenance management
points system for low voltage industrial motors using IoT. The device
alpha The moment index used for determining the monitors the temperature and vibration frequency level of
distance between and within segments. industrial motors, wirelessly transmits the data to a logging
station and sends a push notification in case of a spike, or once
Table 9 summarizes the outputs for vibration of motor B and the temperature or vibrations exceed predefined parameters.
F, where k.hat is the number of clusters within the data created The functionality and accuracy of the device were tested on two
by the change points, p.values are the approximate p-values different motors and pilot testing results were analyzed and
estimated for each permutation test, and permutations are the presented.
number of permutations performed by each of the sequential The current condition monitoring system will be further
permutation test. The results indicate that there are no change developed by integrating machine learning and data mining
points in the distribution of the data. approaches to capture early indicators of potential motor faults
and alert the user about specific faults. The additional plan
Table 9. Output of nonparametric algorithm for multiple change point includes, the development of a digital twin of the system to
analysis of multivariate data
simulate anomalies and collect simulation data that will be used
Vibration Motor Class B Vibration Motor Class F as an input to predictive maintenance machine learning
k.hat =1 k.hat =1 algorithms. Simulation will be used as a testing environment to
p.values =1 p.values =1 experiment with various policies and optimize the machine
permutations =499 permutations =499
learning algorithms in order to transform maintenance
scheduling from a painstaking manual task into a strategic
advantage.
Fig. 7. graphically depicts the box-and-whisker plots of the
collected temperature data for motors class B and class F,
respectively.
456 Konstantinos Mykoniatis et al. / Procedia Manufacturing 42 (2020) 450–456
Author name / Procedia Manufacturing 00 (2019) 000–000 7

References [13] A. Angelopoulou, K. Mykoniatis, K. Carlson and S. J.

Kim, "UCF Smart Mailbox: Reinforcing
Communications in the Neighborhoods. In," in
[1] T. Dunne, "Plant age and technology use in US International Conference on Human-Computer
manufacturing industries," The RAND Journal of Interaction, 2014.
Economics, pp. 488-499, 1994. [14] P. Asghari, A. M. Rahmani and H. H. S. Javadi,
[2] V. E. Lambson, "Industry evolution with sunk costs and "Internet of Things applications: A systematic review,"
uncertain market conditions," International Journal of Computer Networks, vol. 148, pp. 241-261, 2019.
Industrial Organization, vol. 9, no. 2, pp. 171-196, [15] J. Manyika, M. Chui, P. Bisson, J. Woetzel, J. R. Dobbs
1991. and B. a. D. Aharon, "Unlocking the potential of the
[3] W. G. Garrison, "One hundred largest losses-A Thirty- Internet of Things," McKinsey Global Institute, New
year review of property damage losses in the York, 2011.
hydrocarbon-chemical industries," M&M Protection [16] E. Condes, "Fast Fourier Transform for Arduino,"
Consultant, Chicago, 1988. [Online]. Available:
[4] P. A. Carson and C. J. Mumford, "An analysis of https://github.com/kosme/arduinoFFT. [Accessed
incidents involving major hazards in the chemical 2019].
industry.," Journal of Hazardous materials, vol. 3, no. [17] "Digi.com," 13 September 2019. [Online]. Available:
2, pp. 149-165, 1979. https://www.digi.com/products/embedded-systems/digi-
[5] S. A. McInerny and Y. Dai, "Basic vibration signal xbee-tools/xctu.
processing for bearing fault detection," IEEE [18] T. J. Barnish, M. R. Muller and D. J. Kasten, "Motor
Transactions on education, pp. 149-156, 2003. Maintenance: A Survey of Techniques and Results," in
[6] S. Nandi, H. A. Toliyat and X. Li, "Condition Proceedings of the 1997 ACEEE Summer Study on
monitoring and fault diagnosis of electrical motors—A Energy Efficiency in Industry. American Council for an
review," IEEE transactions on energy conversion, pp. Energy-Efficient Economy, Washington, DC, 1997.
719-729, 2005. [19] E. Cowern, "Motor Temperature Ratings," Baldor
[7] L. M. Bergasa, J. Nuevo, M. A. Sotelo, R. Barea and M. Basics: Motors, no. 317 , pp. 42-47, March 2017.
E. Lopez, "Real-time system for monitoring driver [20] R. J. Lawerie, "Common causes of vibration," EC & M:
vigilance," IEEE Transactions on Intelligent Electrical Construction and Maintenance, vol. 3, no.
Transportation Systems, pp. 63-77, 2006. 92, 1994.
[8] W. Yang, P. J. Tavner, C. J. Crabtree and M. [21] Y. R. Patni and M. Ansari, "Review paper on Stator and
Wilkinson, "Cost-effective condition monitoring for Rotor Fault Diagnosis Of 3-Phase Induction Motors,"
wind turbines," IEEE Transactions on industrial International Research Journal of Engineering and
electronics, pp. 263-271, 2009. Technology, vol. 5, no. 4, pp. 180-185, 2018.
[9] V. J. Hodge, S. O' Keefe, M. Weeks and A. Moulds, [22] PK, I. Diah, B. Y. Dewantara and W. M. Utomo,
"Wireless sensor networks for condition monitoring in "Healthy Monitoring and Fault Detection Outer Race
the railway industry: A survey," IEEE Transactions on Bearing in Induction Motor Using Stator Current,"
Intelligent Transportation Systems, pp. 1088-1106, International Journal of Integrated Engineering, vol.
2014. 11, no. 3, 2019.
[10] D. Goyal, A. Chaudhary, R. K. Dang, B. S. Pabla and S. [23] D. S. Matteson and N. A. James, "A nonparametric
S. Dhami, "Condition Monitoring of Rotating approach for multiple change point analysis of
Machines: A review," Wold Scientific News, pp. 98-108, multivariate data," Journal of the American Statistical
2018. Association , vol. 109, no. 505, pp. 334-335, 2014.
[11] B. K. N. Rao, Handbook of Condition monitoring, [24] A. Gandy, "Sequential implementation of Monte Carlo
Elsevier, 1996. tests with uniformly bounded resampling risk," Journal
[12] S. D. T. Kelly, N. K. Suryadevara and S. C. of the American Statistical Association, vol. 104, no.
Mukhopadhyay, "Towards the implementation of IoT 488, pp. 1504-1511, 2009.
for environmental condition monitoring in homes,"
IEEE sensors journal, vol. 13, no. 10, pp. 3846-3853,
2013.