sensors

Article
Structural Health Monitoring: An IoT Sensor System
for Structural Damage Indicator Evaluation
Mirco Muttillo 1 , Vincenzo Stornelli 1, * , Rocco Alaggio 2 , Romina Paolucci 1 ,
Luca Di Battista 3 , Tullio de Rubeis 1 and Giuseppe Ferri 1
1 Department of Industrial and Information Engineering and Economics (DIIIE), University of L’Aquila,
Piazzale Pontieri 1, Monteluco di Roio, 67100 L’Aquila, Italy; [email protected] (M.M.);
[email protected] (R.P.); [email protected] (T.d.R.); [email protected] (G.F.)
2 Department of Civil, Construction-Architectural and Environmental Engineering (DICEAA),
University of L’Aquila, Via Giovanni Gronchi 18, Zona industrial di Pile, 67100 L’Aquila, Italy;
[email protected]
3 Department of Information Engineering, Computer Science and Mathematics (DISIM),
University of L’Aquila, Via Vetoio, Coppito, 67100 L’Aquila, Italy; [email protected]
* Correspondence: [email protected]; Tel.: +39-0862-2434469




Received: 28 July 2020; Accepted: 28 August 2020; Published: 31 August 2020


Abstract: In the last decades, the applications of structural monitoring are moving toward the field
of civil engineering and infrastructures. Nevertheless, if the structures have damages, it does not
mean that they have a complete loss of functionality, but rather that the system is no longer in an
optimal condition so that, if the damage increases, the structure can collapse. Structural Health
Monitoring (SHM), a process for the identification of damage, periodically collects data from suitable
sensors that allow to characterize the damage and establishes the health status of the structure.
Therefore, this monitoring will provide information on the structure condition, mostly about its
integrity, in a short time, and, for infrastructures and civil structures, it is necessary to assess
performance and health status. The aim of this work is to design an Internet of Things (IoT)
system for Structural Health Monitoring to find possible damages and to see how the structure
behaves over time. For this purpose, a customized datalogger and nodes have been designed.
The datalogger is able to acquire the data coming from the nodes through RS485 communication and
synchronize acquisitions. Furthermore, it has an internal memory to allow for the post-processing
of the collected data. The nodes are composed of a digital triaxial accelerometer, a general-purpose
microcontroller, and an external memory for storage measures. The microcontroller communicates
with an accelerometer, acquires values, and then saves them in the memory. The system has been
characterized and the damage indicator has been evaluated on a testing structure. Experimental
results show that the estimated damage indicator increases when the structure is perturbed. In the
present work, the damage indicator increased by a maximum value of 24.65 when the structure is
perturbed by a 2.5 mm engraving.

Keywords: structural health monitoring; IoT structural monitoring; damage indicator;

damage detection system

1. Introduction
The monitoring applications cover various disciplines, from aerospace to the diagnostics of
malfunctions of machines and mechanical systems, in the last years have been utilized in civil
engineering and infrastructure. This topic is also studied as scientific research, as evidenced by a large
number of articles in the related literature. The purpose of monitoring is to know the behavior of a

Sensors 2020, 20, 4908; doi:10.3390/s20174908 www.mdpi.com/journal/sensors

Sensors 2020, 20, 4908 2 of 15

building in a timely manner and from different points of view. For this reason, energy monitoring
systems [1–7] are often used combined with environmental monitoring systems and sensors [8–15].
This combination allows the building to remain in optimal conditions in terms of consumption and
durability over time and for its increasing connection to the Internet of Thing (IoT) world.
On the other hand, the development of an optimal monitoring system is still an open challenge [16]:
determining which of the existing ones is the most appropriate is absolutely not trivial. Indeed,
the structures themselves are incredibly heterogeneous, both in terms of construction technology
and in terms of age. Therefore, it is challenging to find a system that fits them all indiscriminately.
Furthermore, monitoring is an extremely multidisciplinary topic and it is extremely complex to take
into account all the variables involved.
Structural monitoring systems encounter two types of challenges: aging, with consequent and
gradual loss of operating conditions, and the occurrence of a sudden and unexpected event, as an
earthquake [17]. However, there is a certain heterogeneity in the methods of applying structural
monitoring. The objectives go towards a more precise detection capacity, easier management,
and storage of data (even when they are in large quantities), timeliness, and reliability of the
information provided [18].
The application of a system that allows structural monitoring has a specific name in the literature,
which is Structural Health Monitoring (SHM). Schubel et al. [19] presented a review of structural health
monitoring techniques for wind turbine blades. Indeed, the potential of the structural monitoring for
these specific application savings to manufacturing time and reduces the cost of the quality control
phases. Furthermore, the complete knowledge of the behavior of the structure through monitoring
permits better design and manufacturing. An essential other application of structural health monitoring
is for aircraft. Diamanti et al. [20] presented an SHM technique for composite structures of the aircraft.
The size of the critical damage has been determined by the defect of the composite structures that can
be found with a visual inspection, electromagnetic testing, ultrasonic inspection, and other methods.
Furthermore, fiber optic sensor technology is increasingly used for aircraft monitoring to reduce the
cost of maintenance and to identify damage in the structure [21].
The structural monitoring, in addition to ensuring that the structure is always in excellent health,
also exempts from the need to carry out visual inspections and substitute the use of more conventional
methods (for example ultrasound methods). The advantage of the SHM is characterized by being a very
promising alternative and efficient with respect to the conventional methods. Indeed, visual inspections
are not always possible and are, in any case, expensive in terms of time and money, while the use of
conventional methods cannot give up on the operator experience [22].
Therefore, SHM is generally characterized by a non-destructive approach allowing continuous
and autonomous monitoring thanks to the use of integrated sensors [23,24]. A typical structural
monitoring system, then, is made up of a sensor system, a data processing system, and a health
evaluation system [25].
There is a wide variety of monitoring solutions, in addition to the number and complexity of
sensors. Indeed, there are methods based on the study of natural frequency [26], which allows the
study of vibrations. Furthermore, there are methods widely used in the case of sudden structural
degradations [27]. Other methods are based on the study of modal forms [28], where a system is
capable of limiting false alarms. Then there are the so-called “hybrid” methods because they are based
on the study of approaches [29]. Methods based on the use of artificial neural networks, capable of
learning from past data and formulating predictions on future evolutions of the structure, are also an
object of study [18,30].
Concerning the used methodology, the heart of SHM is damage detection. The occurrence
of damage, in fact, can entail, and often does, changes in the characteristics of the structure (for
example with regard to stiffness) [31] which, properly detected, needs maintenance work to avoid the
aggravation of the situation since, in the long run, the structure itself collapses. It follows that the rapid
identification of the damage is a fundamental step in SHM. A fully developed system should be capable
Sensors 2020, 20, 4908 3 of 15

of detecting and evidencing in near real-time the occurrence of a structural anomaly, identifying its
location, and associating it with a type of structural damage and intensity [32]. There are two ways
to monitor a structure. When its global behavior is analyzed, and the structure is considered as a
single system, we speak of global damage identification. On the other hand, when we focus only on
certain elements considered critical or already weakly damaged we speak of local identification of the
damage [33].
In this work, an enhanced version with a different use of the proposal shown in [34] is presented.
The proposed monitoring system for structural health is based on a microcontroller and two triaxial
accelerometric sensors. The data returned, and subsequently suitably processed, allows to determine
the identification of the damage indicator on an engrave steel bar.

2. Literature Background of Structural Health Monitoring

Starting from this common concept, there were different ways in which SHM has declined over
the years, also in reference to different degrees of complexity. In [35], for example, an SHM system was
born practically by chance because the rather poor sensors installed at the Meazza Stadium in Milan,
were not originally intended to monitor the structure. Only later, it became clear that the information
returned was also interesting from that point of view. In [36], on the other hand, a much more complex
system is presented, which makes use of more than 600 sensors, testifying how wide and varied this
field of research is.
Wang et al. [37], proposed a wireless structural health monitoring system for real-time data
acquisition. This kind of system is limited in the number of sensors and the capability of the
synchronization of the samples. Indeed, to increase the sampling rate, the number of sensors connected
in the same network decreases. Therefore, with a sampling frequency of 100 Hz, the number of nodes
is equal to 12. Furthermore, to use a wireless monitoring system in a large structure, such as a bridge,
a peer-to-peer wireless sensor network must be designed and improved. Hu et al. [38] developed
a wireless monitoring system integrated into the Zhengdian Highway Bridge for structural health
monitoring. This system is able to acquire samples in continuous mode using a microcontroller and
ADC to acquire analog accelerometers. The main problem of analog accelerometers is the output drift
due to the temperature, and compensation circuits are needed. The work shows the limitation of the
proposed wireless system in terms of output data. Indeed, the results are limited due to the noise
interference of the analog circuits and data losses of the transmission.
The main component of the proposed monitoring system is the accelerometer ADXL355 [39].
This accelerometer is a digital sensor that is able to acquire the three-axis accelerations internally and
send them to an external microcontroller. A better explanation of the proposed monitoring system
is given in Section 3.1, and before presenting the proposed IoT sensor system for structural health
monitoring, a literature review of the related works that used the same accelerometers is proposed.
Multiple works in the literature [40–47] present a monitoring system for structural health
monitoring using the accelerometer ADXL355. These works are divided into the system based on a
wireless sensor network (WSN) and wired monitoring systems. Valenti et al. [40] proposed a low-cost
WSN for SHM and the system has been used for identification of modal frequency. The problem of
this system is the synchronization of the samples. Indeed, the only synchronization refers to the start
time and stop time that the master sends to the node. Other work that used a WSN was proposed by
Wondra et al. [45]. The WSN was also used for monitoring the wind turbine tower to wind excitation.
The limit of this system is the maximum sampling rates and the maximum nodes (31 Hz) and three
nodes, respectively. Furthermore, synchronization is also a critical problem for this WSN.
The wired sensor monitoring system is an alternative to a WSN. An application for a wired system
that used the ADXL355 is for earthquake detection [41,42,46]. Microseismic events is an important
research field, and this kind of system can send warning messages when an event occurs. The limit
of these systems is the small number of sensors that can be used. Indeed, the system is composed in
general of one node that sends data to a web server. Other nodes are disconnected from each other and
Sensors 2020, 20, 4908 4 of 15

positioned at distances of kilometers. Pierleoni et al. [43] proposed a wired monitoring system with 64
samples per second without synchronization from each node that communicates the data through an
ethernet connection. This system can appreciate the lowest modal frequency of the structures but not
the highest due to the low sampling frequency. Quqa et al. [44], instead, realized a single node wired
monitoring system for structural health. The system is able to identify the natural frequency and modal
parameters in real-time. The system is based on a single-board computer and accelerometer ADXL355
that limits the synchronization and the maximum number of nodes connected in the same network.
Navabian et al. [47] proposed an event monitoring system for structural health. This system acquires
data if the event exceeds the threshold, and the duration of the acquisition is about 70 s. Although,
like [40], the synchronization is also available from the start and stop acquire campaign.
Based on the literature review of the monitoring system that used the same accelerometer of the
system proposed in this work, a summary of the comparison between wired and wireless can be done.
The existing wired monitoring system has a very high cost, a typically low number of sensors that
can be connected in the same network, high bandwidth, high sensor data rate, and very high sensor
synchronicity. On the other side, the wireless monitoring system has a low cost, a high number of
sensors that can be connected in the same network, limited bandwidth, low sensor data rate, and critical
synchronization of nodes [48].
The main novelty of the wired proposed monitoring system, based on the previous analysis,
are the following:

1. The high number of nodes that can be connected in the same network, the only limitation is due
to the RS485 protocol;
2. High bandwidth;
3. High data rate;
4. High synchronization between nodes;
5. Low-cost system.

3. Materials and Methods

In the present work, for the detection of the damage on the beam model, we proceeded “by
comparison”: first, some measurements were performed on the intact test structure, assuming this as
the reference state; subsequently, they were repeated on the same structure deliberately perturbed
through an incision of 2.5 mm.
The conducted test has been divided into three phases: in the first phase, the sampling frequencies
and the duration of the test were chosen; in the second one, we proceeded to start the system, acquire the
samples, save them in an SD card, stop the acquisition, and send the data to the PC; the third and final
phase consisted of the post-processing of data through Matlab. The test operating phases to derive the
damage indicator are shown in Figure 1 as a sort of flow-chart. Phase 2 consists of two tests: the first
is the test with a healthy structure and the second is with a damaged structure. At system startup,
N = 1. Therefore, the system acquires and saves to the SD card repeatedly for the test time. After that,
the acquisitions were stopped, and data was sent to the PC. Instead, the second test consists of N = 2
and with the damaged structure. The system started and acquired the samples, saved them on the SD
card, and then sent the data to the PC after the test time. The final phase is the post-processing of data
for damage indication.
Sensors 2020, 20, 4908 5 of 15
Sensors 2020, 20, x; 6 of 15

Figure 1.
Figure 1. Damage
Damage indicator
indicator test
test operating phases.
operating phases.

3.1. Damage Indicator

In addition, the microcontroller provides data storage on an external SD (Secure Digital) card,
whose presence
Damage detectionis necessary considering
is a problem that that the number
has been studiedofusing
samples acquired
various can quickly
methods [49–63].reach the
A fuzzy
order of millions. Therefore, at the end of the single acquisition, it is particularly
neural network for two-stage damage detection is presented by Jiang et al. [49]. A damage assessment useful to store the
data on
based onana fuzzy
SD card so that
neural it can also
network be sent
for the to thehas
first stage master
beenlater.
performed. Whereas in the second stage,
thanks to the using of the union of data fusion and fuzzythe
Another critical point of the system is related to fact that
models, theevaluation
a final code execution
has beentime on the
achieved.
microcontroller for data acquisition and saving is much longer than the time occurring
This approach can identify more patterns than the single-stage fuzzy model. Gui et al. [50] illustrated between one
a
sample
three and the next.
optimization This always
algorithm happens,kernel
for Gaussian even for sampling
function frequencies
parameters. equal
These to 1 kHz, and
optimization makes
algorithms
it impossible
are based on to acquire
vector and save
machines andallare
data sequentially.
allowed To avoid
to use them data loss,
for damage the integrated
detection. OtherDMA has
methods
been used
used on the Swarm
the Particle microcontroller,
Optimization which, through direct
algorithm access to memory,
[51], Operational allows thewith
Modal Analysis bypassing
dynamicof
the control unit of the microcontroller itself and to store the data directly
measurements [52], frequency response functions with artificial neural network-based for damage in the SD card memory.
To complete
detection [53], andthe description of Neural
1D Convolutional the system,Networksand for
in particular, of thedamage
vibration-based nodes, detection
it shouldand be
emphasizedin
localization that, as can[54].
real-time be seen in Figure 2, each of them is made up of two accelerometers, both
connected to the same microcontroller. This is made possible by the fact that communication, in this
One of the first works on the identification of the damage index addressed from the one-dimensional
case, is managed via the I2C protocol, which allows for connection of more than one device to the
point of view is that shown in [55], in which a method to evaluate the integrity of the structures
same bus, each with its own address, chosen via external hardware settings. In particular, the sensors
non-destructively is shown. In particular, it is described how the measurement of vibrations carried
used are integrated triaxial digital accelerometers. The fact that they are integrated makes it possible
to calm the price of the system, making it effectively competitive even from a purely economic point
of view. Specifically, the sensor model used is the Analog Device ADXL355, whose basic
Sensors 2020, 20, 4908 6 of 15

out in a single station in the structure can be used, in combination with a suitable theoretical model,
to indicate both the position and the extent of the damage.
The proposed experiment illustrates the application of the system for structural health monitoring
using a damage detection method based on Stochastic Subspace Identification concepts [56].
The method, being based on a non-parametric test, does not require to explicitly know system parameters
and is suitable for automatic data-driven damage detection monitoring of in-service structures.
Any damage diagnosis method requires the extraction of damage-sensitive features from the
measurement data of the monitored system. The feature vector is generally defined in a way that it is
approximately Gaussian distributed with zero mean in the reference (undamaged) state and non-zero
mean in the damaged state, hence the designation of the residual vector [57,58]. Many residuals have
been used in the literature [59,60]; in this paper, the subspace residual, representing the orthonormality
defect between subspaces characterizing the dynamic response in the current state of the structure
with respect to its reference, is adopted, specifically the robust subspace residual [61] less prone to
changes in excitation covariance.
Measures of the dynamic response of the structure in its reference state are acquired over
time to produce a statistical model of the residuals under changing environmental conditions [62].
If no structural damage occurs, the orthonormality assumption between the mentioned subspaces,
evaluated for different data sets, remains approximately valid according to small residues. However,
possible structural damage causes an increase in residues. This increase involves, with the choice of an
adequate metric, a significant rise in the scalar damage indicator. Therefore, if this value falls beyond
an appropriate threshold, it indicates the presence of damage [63].

3.2. System Description

The whole general scheme of the proposed monitoring system, with typical connection and node
architecture, is shown in Figure 2.
The system is composed of nodes, described in more detail below, which, via the RS485 protocol,
communicate with a master. The choice of this protocol is not casual: thanks to its characteristics,
in fact, the nodes can be positioned even at a distance of hundreds of meters, without compromising
their capability to communicate correctly with the master. This aspect is fundamental, as it allows the
master to synchronize the various nodes, to recover the data sent by them and to forward them to the
PC for post-processing via Matlab.
The single node, as seen in the previous figure, is made up, of a microcontroller, the SAM3X8E
ARM Cortex-M3 [64], equipped with an integrated Direct Memory Access (DMA). One of its tasks is to
manage communication with the master.
The microcontroller, of course, needs to be powered. However, since the total current consumption
of the node is only 100 mA, it is configured as a low power system. This allows it to be powered also
through photovoltaic panels with a battery and, then, the possibility of positioning the nodes even at
great distances and in environments with no electricity.
In addition, the microcontroller provides data storage on an external SD (Secure Digital) card,
whose presence is necessary considering that the number of samples acquired can quickly reach the
order of millions. Therefore, at the end of the single acquisition, it is particularly useful to store the
data on an SD card so that it can also be sent to the master later.
Another critical point of the system is related to the fact that the code execution time on the
microcontroller for data acquisition and saving is much longer than the time occurring between one
sample and the next. This always happens, even for sampling frequencies equal to 1 kHz, and makes
it impossible to acquire and save all data sequentially. To avoid data loss, the integrated DMA has
been used on the microcontroller, which, through direct access to memory, allows the bypassing of the
control unit of the microcontroller itself and to store the data directly in the SD card memory.
To complete the description of the system, and in particular, of the nodes, it should be emphasized
that, as can be seen in Figure 2, each of them is made up of two accelerometers, both connected to the
Sensors 2020, 20, 4908 7 of 15

same microcontroller. This is made possible by the fact that communication, in this case, is managed
via the I2C protocol, which allows for connection of more than one device to the same bus, each with
its own address, chosen via external hardware settings. In particular, the sensors used are integrated
triaxial digital accelerometers. The fact that they are integrated makes it possible to calm the price of
Sensors 2020, 20, x; 7 of 15
the system, making it effectively competitive even from a purely economic point of view. Specifically,
the sensor model
characteristics areused is the
voltage Analog
supply Device
range ADXL355,
equal whose
to 2.25–3.6 basic characteristics
V, settable are
range ±2, 4, 8 g forvoltage supply
each axis, for
range equal to 2.25–3.6 V, settable range ±2, 4, 8 g for each axis, for ±2 g the sensitivity
±2 g the sensitivity is 3.9 µg/LSB, low power device with 200 µA consumption in measurement mode is 3.9 µg/LSB,
low power
and 20-bit device
internalwithanalog-to-digital
200 µA consumption in measurement
converter (ADC). Themode and 20-bit of
sensitivity internal analog-to-digital
this sensor changes,
±0.01%/ ◦ C,
according to the temperature, of ±0.01%/°C, with respect to the ambient value ofof25
converter (ADC). The sensitivity of this sensor changes, according to the temperature, °C. The
with respect to the ◦
accelerometer has ambient value temperature
an internal of 25 C. The accelerometer
sensor that the has microcontroller
an internal temperature
can read sensor
for that
the
the microcontroller
data compensation. can read for the data compensation.

Figure 2. The
The scheme
scheme of
of the
the proposed
proposed monitoring
monitoring system. In
In this
this scheme,
scheme, the
the architecture
architecture of the node
is presented.

4. Experimental
4. Experimental Set-Up
Set-Up
The proposed
The proposed monitoring
monitoring system
system has
has been
been tested
tested utilizing
utilizing the
the experimental setup shown
experimental setup shown inin
Figure 3,
Figure 3, where
where the
the identification
identification of damage indicator
of damage indicator procedure
procedure has
has been
been applied. The environment
applied. The environment
temperature test was equal to 25 ◦ C, and under these test conditions, the sensitivity of the accelerometers
temperature test was equal to 25 °C, and under these test conditions, the sensitivity of the
does not change.does
accelerometers The not
cantilever
change. structure (aluminum
The cantilever bar) has
structure been anchored
(aluminum withbeen
bar) has a bench vice. The
anchored two
with a
accelerometers of the acquisition node have been put on the aluminum bar. The first
bench vice. The two accelerometers of the acquisition node have been put on the aluminum bar. The accelerometer
first accelerometer has been mounted at the end of the bar and the second is positioned at 16.6 cm
distance from the blocking point.
For this test, one master and one node that communicate through the RS485 bus were used. The
node acquires the data from two three-axis accelerometers, saves them on an SD card, and at the end
of the test, transmits to the master device. Moreover, an external power supply for the node and
Sensors 2020, 20, 4908 8 of 15

has been mounted at the end of the bar and the second is positioned at 16.6 cm distance from the
blocking point.
For this test, one master and one node that communicate through the RS485 bus were used.
The node acquires the data from two three-axis accelerometers, saves them on an SD card, and at the
end of the test, transmits to the master device. Moreover, an external power supply for the node and
master is required. A picture of the complete testing system is shown in Figure 4.
Sensors 2020, 20, x; 8 of 15
Sensors 2020, 20, x; 8 of 15

Figure 3. Testing the structure for the identification of the damage indicator.
Figure3.3.Testing
Figure Testingthe
the structure
structure for
for the
the identification
identificationofofthe
thedamage
damageindicator.
indicator.

Figure
Figure 4.4.Experimental
Experimentalset-up
set-up of
of testing
testing structure
structure for
forthe
theidentification
identificationofofthe
thedamage
damageindicator.
indicator.
Figure 4. Experimental set-up of testing structure for the identification of the damage indicator.
Havingadapted
Having adaptedthethe sampling
sampling frequency
frequencyofof250 Hz,
250 Hz,two tests
two have
tests beenbeen
have carried out. The
carried out.first
The test
first
concerned
test concerned the acquisition campaign
the acquisition with
campaign the healthy
with the structure,
healthy and
structure, after that,
andbeen the aluminum
aftercarried
that, the bar
aluminum was
Having adapted the sampling frequency of 250 Hz, two tests have out. The first bar
test
wasdamaged
damaged forfor
thethe
second
secondexperiment.
experiment.ForFor
damage
damage detection,
detection,a perturbation
a perturbation to to
thethe
structure
structure was
was
concerned the acquisition campaign with the healthy structure, and after that, the aluminum bar was
induced. Indeed, on the testing structure (Figure 5), a 2.5 mm engrave was realized. For both the tests,
damaged for the second experiment. For damage detection, a perturbation to the structure was
the bar was stressed with only ambient noise. The approach of damage detection is based on an
induced. Indeed, on the testing structure (Figure 5), a 2.5 mm engrave was realized. For both the tests,
algorithm that processes the output data of the acquisition system when the structure is subjected to
theexternal
bar was stressed These
excitations. with output
only ambient noise. two
data represent The measurements
approach of damage
lasting 15detection
min of theishealthy
based andon an
algorithm that processes the output data of the acquisition system when the structure
damaged structure. The algorithm allows the evaluation of the damage indicator of a structure. is subjected to
external excitations. These output data represent two measurements lasting 15 min of the healthy and
Sensors 2020, 20, 4908 9 of 15

induced. Indeed, on the testing structure (Figure 5), a 2.5 mm engrave was realized. For both the
tests, the bar was stressed with only ambient noise. The approach of damage detection is based on an
algorithm that processes the output data of the acquisition system when the structure is subjected to
external excitations. These output data represent two measurements lasting 15 min of the healthy and
damaged structure. The algorithm allows the evaluation of the damage indicator of a structure.
Sensors 2020,
Sensors 20,20,
2020, x; x; 9 of
9 of 1515

Figure
Figure 5. Engraved aluminumbar
baranchored
anchoredin
in the bench
bench vice for the damage detection test.
Figure 5. 5.Engraved
Engraved aluminum
aluminum bar anchored in the
the benchvice
vicefor
forthe
thedamage
damagedetection
detectiontest.
test.
5. Results
5. Results andand Discussion
Discussion
5. Results and Discussion
Measurements have been performed on six axes, three for each triaxial sensor. In Figures 6 and
Measurements
Measurements have
have beenperformed
been performedon onsix
sixaxes,
axes,three
threefor
foreach
eachtriaxial
triaxial sensor.
sensor. In
In Figures
Figures 66 and
and 7,
7, the acceleration measurements relative to all axes located on the structure for the two tests
the7,acceleration
the acceleration
are shown.
measurements
measurements relative
relative to all to alllocated
axes axes located
on the on the structure
structure for the for tests
two the two
are tests
shown.
are shown.

FigureFigure
6. Six6.axes
Six acquired
axes acquired data through
data through the proposed
the proposed monitoring
monitoring systemsystem
of the of the undamaged
undamaged structure.
structure. The first three measurements are related to sensor 1 (X , Y , Z ), and the
The first three measurements are related to sensor 1 (X1 , Y1 , Z1 ), and the others are from sensor
1 1 1 others are from
2 (X2 ,
Figure
sensor 6.2 Six
(X 2, axes
Y acquired data through the proposed monitoring system of the undamaged
2, Z2).
Y2 , Z2 ).
structure. The first three measurements are related to sensor 1 (X1, Y1, Z1), and the others are from
sensor 2 (X2, Y2, Z2).
Sensors 2020, 20, 4908 10 of 15
Sensors 2020, 20, x; 10 of 15

Figure 7.
Figure 7. Six
Six axes
axes acquired
acquired data data through
through the
the proposed
proposed monitoring
monitoring system
system of of the
the damaged
damaged structure
structure
(engravingof
(engraving of2.5
2.5mm).
mm).The Thefirst
firstthree
threemeasurements
measurementsarearerelated
relatedtotosensor 1 ,1,YY
sensor1 1(X(X 11,,ZZ11),), and
and the
the others
others
are from
are from sensor
sensor22(X(X22,, Y
Y22, Z22).).

The
Theacquired
acquiredsamples
samples were 224,400
were for the
224,400 forhealthy structure
the healthy and 213,602
structure for the damaged
and 213,602 structure.
for the damaged
Indeed,
structure. Indeed, with sample time equal to 5 ms, the whole experimental time is about 15 healthy
with sample time equal to 5 ms, the whole experimental time is about 15 min for the min for
structure
the healthyand approximately
structure 14.27 min for14.27
and approximately the damaged structure.
min for the damaged In structure.
order to estimate
In ordertheto damage
estimate
indicator,
the damage theindicator,
first measurements have been divided
the first measurements into four
have been series
divided with
into about
four 50,000
series withsamples each.
about 50,000
These series have been called UD1, UD2, UD3, and UD0. The latter series UD0
samples each. These series have been called UD1, UD2, UD3, and UD0. The latter series UD0 has been has been used to
calculate the damage
used to calculate the indicator as the reference
damage indicator subspace.subspace.
as the reference Similarly,Similarly,
the measurements of the damaged
the measurements of the
structure have been divided into four series called D1, D2, D3, and D4 with the
damaged structure have been divided into four series called D1, D2, D3, and D4 with the same same number of samples.
Therefore,
number with the reference subspace UD0, the algorithm returns three damage indicator values
of samples.
for the healthy structure
Therefore, with theand four forsubspace
reference the damaged
UD0,structure. Thesereturns
the algorithm values are smaller
three damage in structural
indicator
health condition than the structural damage condition. The values of the damage
values for the healthy structure and four for the damaged structure. These values are smaller indicator are shown in
in Figure 8.health condition than the structural damage condition. The values of the damage indicator
structural
Finally,ininFigure
are shown Table 1,
8.the calculated damage indicators have been reported. The values of the damage
indicator have an increase of ten times, with only a 2.5 mm engrave. The results show that the proposed
system, with synchronous samples between the two sensors, is able to detect damages in a monitored
structure. However, the proposed monitoring system with a damage indicator approach will detect
structural defects or damage after events such as earthquakes or landslides.
Sensors 2020, 20, 4908 11 of 15
Sensors 2020, 20, x; 11 of 15

Figure 8. Values of the damage indicator. The green bars represent the structure in a healthy condition
Figure 8. Values of the damage indicator. The green bars represent the structure in a healthy condition
and the red bars show the damage indicator with the engraved structure.
and the red bars show the damage indicator with the engraved structure.

Finally, inDamage
Table 1. Table 1, the calculated
indicator values fordamage indicators
all measurements have been and
in undamaged reported.
damagedThe values of the
structures.
damage indicator have an increase of ten times, with only a 2.5 mm engrave. The results show that
Condition Series Damage Indicator
the proposed system, with synchronous samples between the two sensors, is able to detect damages
in a monitored structure. However, the proposed UD1 monitoring system 1.68 with a damage indicator
Healthy Structure UD2 2.60
approach will detect structural defects or damage after events such as earthquakes or landslides.
UD3 2.09
D1 24.50
Table 1. Damage indicator values for all measurements
D2 in undamaged and damaged structures.
24.65
Damaged Structure
D3 22.01
Condition Series
D4 23.28Damage Indicator

UD1 1.68
6. Conclusions
Healthy Structure UD2 2.60
In this work, an IoT monitoring system for structural health is presented. The IoT system,
with application in Smart Buildings, allows for the measurement of the main parameters
UD3 2.09 for evaluating
the damage indicator. The system is based on the microcontroller Sam3X8E ARM cortex-M3 and
high-resolution digital accelerometers ADXL355. D1Furthermore, thanks to the use 24.50
of an SD card and
DMA, the system allows for the acquisition of a high D2 number of samples and communicates
24.65 through
the RS485 to the master
Damaged device.
Structure
The reliable results have been ensured with the D3high synchronization between 22.01 the sensors and their
high resolution. Instead, the problems of the traditional analog sensors used in the typical monitoring
D4 23.28
systems have been eliminated with the use of the digital accelerometers. Therefore, the proposed
monitoring system is cheaper than an analog solution.
6. Conclusions
The system has been used to evaluate damage in an aluminum bar locked in a bench vice. The test
gaveInthe possibility
this work, an to
IoTassess, with the
monitoring developed
system system,
for structural its capability
health of damage
is presented. The IoTidentification.
system, with
An engraving
application was realized
in Smart Buildings,in allows
the structure for comparing
for the measurement of the
the evaluated damage
main parameters forindicator in
evaluating
bothdamage
the conditions.
indicator. The system is based on the microcontroller Sam3X8E ARM cortex-M3 and
Therefore, to
high-resolution implement
digital a structural
accelerometers health monitoring
ADXL355. Furthermore,system,
thanksthetodetection
the use ofofananySDdamage
card and is
important.
DMA, Futureallows
the system development of the proposed
for the acquisition system
of a high will concern
number of samplesthe and
installation of the device
communicates on
through
realRS485
the structures,
to thesuch as buildings.
master device.
The reliable results have been ensured with the high synchronization between the sensors and
their high resolution. Instead, the problems of the traditional analog sensors used in the typical
Sensors 2020, 20, 4908 12 of 15

Author Contributions: Conceptualization, M.M., V.S., R.A., R.P., and L.d.B.; investigation, R.P.; data curation,
L.d.B. and R.A.; writing—original draft preparation, R.P., M.M., L.d.B., and T.d.R.; supervision, V.S., R.A., and G.F.
All authors have read and agreed to the published version of the manuscript.
Funding: This research received no external funding.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Ghosh, A.; Raha, A.; Mukherjee, A. Energy-Efficient IoT-Health Monitoring System using Approximate
Computing. Internet Things 2020, 9, 100166. [CrossRef]
2. Luan, H.; Leng, J. Design of energy monitoring system based on IOT. In Proceedings of the 28th Chinese
Control and Decision Conference (CCDC), Yinchuan, China, 28–30 May 2016.
3. Kychkin, A. Synthesizing a System for Remote Energy Monitoring in Manufacturing. Metallurgist 2016, 59,
752–760. [CrossRef]
4. De Rubeis, T.; Nardi, I.; Muttillo, M.; Ranieri, S.; Ambrosini, D. Room and window geometry influence for
daylight harvesting maximization—Effects on energy savings in an academic classroom. Energy Procedia
2018, 148, 1090–1097. [CrossRef]
5. De Rubeis, T.; Nardi, I.; Muttillo, M. Development of a low-cost temperature data monitoring. An upgrade
for hot box apparatus. J. Phys. Conf. Ser. 2017, 923, 012039. [CrossRef]
6. De Rubeis, T.; Muttillo, M.; Nardi, I.; Pantoli, L.; Stornelli, V.; Ambrosini, D. Integrated Measuring and
Control System for Thermal Analysis of Buildings Components in Hot Box Experiments. Energies 2019,
12, 2053. [CrossRef]
7. Pantoli, L.; Muttillo, M.; Stornelli, V.; Ferri, G.; Gabriele, T. A low cost flexible power line communication
system. In Lecture Notes in Electrical Engineering; Springer: Cham, Switzerland, 2017; pp. 413–420.
8. Deshmukh, A.; Shinde, U. A low cost environment monitoring system using raspberry Pi and arduino with
Zigbee. In Proceedings of the 2016 International Conference on Inventive Computation Technologies (ICICT),
Tamilnadu, India, 26–27 August 2016.
9. Novas, N.; Gázquez, J.; MacLennan, J.; García, R.; Fernández-Ros, M.; Manzano-Agugliaro, F. A real-time
underground environment monitoring system for sustainable tourism of caves. J. Clean. Prod. 2017, 142,
2707–2721. [CrossRef]
10. Kim, S.; Jeong, J.; Hwang, M.; Kang, C. Development of an IoT-based atmospheric environment monitoring
system. In Proceedings of the 2017 International Conference on Information and Communication Technology
Convergence (ICTC), Jeju Island, Korea, 18–20 October 2017.
11. Liu, L.; Zhang, Y. Design of greenhouse environment monitoring system based on Wireless Sensor Network.
In Proceedings of the 3rd International Conference on Control, Automation and Robotics (ICCAR), Nagoya,
Japan, 22–24 April 2017.
12. Barile, G.; Leoni, A.; Pantoli, L.; Stornelli, V. Real-Time Autonomous System for Structural and Environmental
Monitoring of Dynamic Events. Electronics 2018, 7, 420. [CrossRef]
13. Fusacchia, P.; Muttillo, M.; Leoni, A.; Pantoli, L.; Parente, F.; Stornelli, V.; Ferri, G. A Low Cost Fully Integrable
in a Standard CMOS Technology Portable System for the Assessment of Wind Conditions. Procedia Eng.
2016, 168, 1024–1027. [CrossRef]
14. Pantoli, L.; Paolucci, R.; Muttillo, M.; Fusacchia, P.; Leoni, A. A multisensorial thermal anemometer system.
In Lecture Notes in Electrical Engineering; Springer: Cham, Switzerland, 2017; pp. 330–337.
15. Pantoli, L.; Muttillo, M.; Ferri, G.; Stornelli, V.; Alaggio, R.; Vettori, D.; Chinzari, L.; Chinzari, F. Electronic
system for structural and environmental building monitoring. In Lecture Notes in Electrical Engineering;
Springer: Cham, Switzerland, 2019; Volume 539, pp. 481–488.
16. Ooijevaar, T.H. Vibration Based Structural Health Monitoring of Composite Skin-Stiffener Structures. Ph.D.
Thesis, Universiteit Twente, Enschede, The Netherlands, March 2014. [CrossRef]
17. Li, H.; Ren, L.; Jia, Z.; Yi, T.; Li, D. State-of-the-art in structural health monitoring of large and complex civil
infrastructures. J. Civ. Struct. Health Monit. 2015, 6, 3–16. [CrossRef]
18. Ko, J.; Ni, Y. Technology developments in structural health monitoring of large-scale bridges. Eng. Struct.
2005, 27, 1715–1725. [CrossRef]
Sensors 2020, 20, 4908 13 of 15

19. Schubel, P.; Crossley, R.; Boateng, E.; Hutchinson, J. Review of structural health and cure monitoring
techniques for large wind turbine blades. Renew. Energy 2013, 51, 113–123. [CrossRef]
20. Diamanti, K.; Soutis, C. Structural health monitoring techniques for aircraft composite structures.
Prog. Aerosp. Sci. 2010, 46, 342–352. [CrossRef]
21. Guo, H.; Xiao, G.; Mrad, N.; Yao, J. Fiber Optic Sensors for Structural Health Monitoring of Air Platforms.
Sensors 2011, 11, 3687–3705. [CrossRef] [PubMed]
22. Gomes, G.F.; Mendéz, Y.A.D.; Alexandrino, P.D.S.L.; da Cunha, S.S., Jr.; Ancelotti, A.C., Jr. The use of intelligent
computational tools for damage detection and identification with an emphasis on composites—A review.
Compos. Struct. 2018, 196, 44–54. [CrossRef]
23. Heslehurst, R. Defects and Damage in Composite Materials and Structures; CRC Press, Taylor & Francis Group:
Boca Raton, FL, USA, 2017.
24. Gopalakrishnan, S.; Ruzzene, M.; Hanagud, S. Computational Techniques for Structural Health Monitoring;
Springer: Berlin/Heidelberg, Germany, 2011.
25. Mufti, A. Structural Health Monitoring of Innovative Canadian Civil Engineering Structures. Struct. Health
Monit. Int. J. 2002, 1, 89–103. [CrossRef]
26. Lifshitz, J.; Rotem, A. Determination of Reinforcement Unbonding of Composites by a Vibration Technique.
J. Compos. Mater. 1969, 3, 412–423. [CrossRef]
27. Velmurugan, R.; Balaganesan, G. Modal analysis of pre and post impacted nano composite laminates. Lat. Am.
J. Solids Struct. 2011, 8, 9–26. [CrossRef]
28. Fu, Y.; Liu, J.; Wei, Z.; Lu, Z. A two-step approach for damage Identification in plates. J. Vib. Control 2014, 22,
3018–3031. [CrossRef]
29. Kim, J.; Ryu, Y.; Cho, H.; Stubbs, N. Damage identification in beam-type structures: Frequency-based method
vs mode-shape-based method. Eng. Struct. 2003, 25, 57–67. [CrossRef]
30. Gomes, G.; Ancelotti, A.; da Cunha, S. Residual stress prediction in porous cfrp using artificial neural
networks. Compos. Mech. Comput. Appl. Int. J. 2018, 9, 27–40. [CrossRef]
31. Cawley, P.; Adams, R. The location of defects in structures from measurements of natural frequencies. J. Strain
Anal. Eng. Des. 1979, 14, 49–57. [CrossRef]
32. Gomes, G.; Diniz, C.; da Cunha, S.; Ancelotti, A. Design Optimization of Composite Prosthetic Tubes Using
GA-ANN Algorithm Considering Tsai-Wu Failure Criteria. J. Fail. Anal. Prev. 2017, 17, 740–749. [CrossRef]
33. Dixit, S.; Sharma, K. A Review of Studies in Structural Health Monitoring (SHM). In Proceedings of the
Creative Construction Conference, Budapest, Hungary, 29 June–2 July 2019.
34. Muttillo, M.; Battista, L.; de Rubeis, T.; Nardi, I. Structural health continuous monitoring of
buildings—A modal parameters identification system. In Proceedings of the 4th International Conference on
Smart and Sustainable Technologies (SpliTech), Bol/Split, Croatia, 18–21 June 2019.
35. Cigada, A.; Moschioni, G.; Vanali, M.; Caprioli, A. the measurement network of the san siro meazza stadium
in milan: Origin and implementation of a new data acquisition strategy for structural health monitoring.
Exp. Tech. 2010, 34, 70–81. [CrossRef]
36. Ni, Y.; Xia, Y.; Liao, W.; Ko, J. Technology innovation in developing the structural health monitoring system
for Guangzhou New TV Tower. Struct. Control Health Monit. 2009, 16, 73–98. [CrossRef]
37. Wang, Y.; Lynch, J.; Law, K. A wireless structural health monitoring system with multithreaded sensing
devices: Design and validation. Struct. Infrastruct. Eng. 2007, 3, 103–120. [CrossRef]
38. Hu, X.; Wang, B.; Ji, H. A Wireless Sensor Network-Based Structural Health Monitoring System for Highway
Bridges. Comput. Aided Civ. Infrastruct. Eng. 2012, 28, 193–209. [CrossRef]
39. ADXL355 Datasheet and Product Info. Analog Devices. Available online: https://www.analog.com/en/
products/adxl355.html (accessed on 4 June 2020).
40. Valenti, S.; Conti, M.; Pierleoni, P.; Zappelli, L.; Belli, A.; Gara, F.; Carbonari, S.; Regni, M. A low cost wireless
sensor node for building monitoring. In Proceedings of the IEEE Workshop on Environmental, Energy, and
Structural Monitoring Systems (EESMS), Salerno, Italy, 21–22 June 2018.
41. González, S.; Jiménez, J.C.; Guevara, R.; Palacios, I. IoT-based microseismic monitoring system for the
evaluation of structural health in Smart cities. In Proceedings of the Ibero-American Congress Smart Cities
(ICSC-CITIES), Soria, Spain, 26–27 September 2018; pp. 1–13.
42. Lee, J.; Khan, I.; Choi, S.; Kwon, Y. A Smart IoT Device for Detecting and Responding to Earthquakes.
Electronics 2019, 8, 1546. [CrossRef]
Sensors 2020, 20, 4908 14 of 15

43. Pierleoni, P.; Conti, M.; Belli, A.; Palma, L.; Incipini, L.; Sabbatini, L.; Valenti, S.; Mercuri, M.; Concetti, R.
IoT Solution based on MQTT Protocol for Real-Time Building Monitoring. In Proceedings of the IEEE 23rd
International Symposium on Consumer Technologies (ISCT), Ancona, Italy, 19–21 June 2019.
44. Quqa, S.; Landi, L.; Diotallevi, P. Real time damage detection through single low-cost smart sensor.
In Proceedings of the 7th International Conference on Computational Methods in Structural Dynamics and
Earthquake Engineering, Crete, Greece, 24–26 June 2019.
45. Wondra, B.; Malek, S.; Botz, M.; Glaser, S.; Grosse, C. Wireless High-Resolution Acceleration Measurements
for Structural Health Monitoring of Wind Turbine Towers. Data-Enabled Discov. Appl. 2019, 3, 4. [CrossRef]
46. Khan, I.; Choi, S.; Kwon, Y. Earthquake Detection in a Static and Dynamic Environment Using Supervised
Machine Learning and a Novel Feature Extraction Method. Sensors 2020, 20, 800. [CrossRef]
47. Navabian, N.; Beskhyroun, S. An Automated Wireless-Based System for Real-Time Health Monitoring of
Civil Infrastructures. In Proceedings of the 2020 New Zealand Society for Earthquake Engineering Annual
Technical Conference, Wellington, New Zealand, 22–24 April 2020.
48. Noel, A.; Abdaoui, A.; Elfouly, T.; Ahmed, M.; Badawy, A.; Shehata, M. Structural Health Monitoring Using
Wireless Sensor Networks: A Comprehensive Survey. IEEE Commun. Surv. Tutor. 2017, 19, 1403–1423.
[CrossRef]
49. Jiang, S.; Zhang, C.; Zhang, S. Two-stage structural damage detection using fuzzy neural networks and data
fusion techniques. Expert Syst. Appl. 2011, 38, 511–519. [CrossRef]
50. Gui, G.; Pan, H.; Lin, Z.; Li, Y.; Yuan, Z. Data-driven support vector machine with optimization techniques
for structural health monitoring and damage detection. KSCE J. Civ. Eng. 2017, 21, 523–534. [CrossRef]
51. Gökdağ, H.; Yildiz, A. Structural Damage Detection Using Modal Parameters and Particle Swarm
Optimization. Mater. Test. 2012, 54, 416–420. [CrossRef]
52. Magalhães, F.; Cunha, A.; Caetano, E. Vibration based structural health monitoring of an arch bridge: From
automated OMA to damage detection. Mech. Syst. Signal Process. 2012, 28, 212–228. [CrossRef]
53. Bandara, R.; Chan, T.; Thambiratnam, D. Structural damage detection method using frequency response
functions. Struct. Health Monit. Int. J. 2014, 13, 418–429. [CrossRef]
54. Abdeljaber, O.; Avci, O.; Kiranyaz, S.; Gabbouj, M.; Inman, D. Real-time vibration-based structural damage
detection using one-dimensional convolutional neural networks. J. Sound Vib. 2017, 388, 154–170. [CrossRef]
55. Adams, R.; Cawley, P.; Pye, C.; Stone, B. A Vibration Technique for Non-Destructively Assessing the Integrity
of Structures. J. Mech. Eng. Sci. 1978, 20, 93–100. [CrossRef]
56. Van Overschee, P.; De Moor, P.; De Moor, B. Subspace identification problem. Automatica 1993, 29, 649–660.
[CrossRef]
57. Allahdadian, S.; Döhler, M.; Ventura, C.; Mevel, L. Towards robust statistical damage localization via
model-based sensitivity clustering. Mech. Syst. Signal Process. 2019, 134, 106341. [CrossRef]
58. Döhler, M.; Mevel, L.; Qinghua, Z. Fault detection, isolation and quantification from Gaussian residuals with
application to structural damage diagnosis. Annu. Rev. Control 2016, 42, 244–256. [CrossRef]
59. Jaishi, B.; Ren, W. Damage detection by finite element model updating using modal flexibility residual.
J. Sound Vib. 2006, 290, 369–387. [CrossRef]
60. Titurus, B.; Friswell, M. Damage detection using successive parameter subset selections and multiple modal
residuals. Mech. Syst. Signal Process. 2014, 45, 193–206. [CrossRef]
61. Döhler, M.; Mevel, L.; Hille, F. Subspace-based damage detection under changes in the ambient excitation
statistics. Mech. Syst. Signal Process. 2014, 45, 207–224. [CrossRef]
62. Aloisio, A.; Di Battista, L.; Alaggio, R.; Antonacci, E.; Fragiacomo, M. Assessment of structural interventions
using Bayesian updating and subspace-based fault detection methods: The case study of S. Maria di
Collemaggio basilica, L’Aquila, Italy. Struct. Infrastruct. Eng. 2020, 1–15. [CrossRef]
Sensors 2020, 20, 4908 15 of 15

63. Aloisio, A.; Di Battista, L.; Alaggio, R.; Fragiacomo, M. Sensitivity analysis of subspace-based damage
indicators under changes in ambient excitation covariance, severity and location of damage. Eng. Struct.
2020, 208, 110235. [CrossRef]
64. ATSAM3X8E—32-bit SAM Microcontrollers. Available online: https://www.microchip.com/wwwproducts/
en/ATsam3x8e (accessed on 4 June 2020).

© 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/).