sensors

Review
A Review of the Piezoelectric Electromechanical
Impedance Based Structural Health Monitoring
Technique for Engineering Structures
Wongi S. Na * and Jongdae Baek
Department of Infrastructure Safety Research, Korea Institute of Civil Engineering and Building Technology,
Gyeonggi-Do 10223, Korea; [email protected]
* Correspondence: [email protected]; Tel.: +82-31-910-0155




Received: 13 February 2018; Accepted: 19 April 2018; Published: 24 April 2018


Abstract: The birth of smart materials such as piezoelectric (PZT) transducers has aided in
revolutionizing the field of structural health monitoring (SHM) based on non-destructive testing
(NDT) methods. While a relatively new NDT method known as the electromechanical (EMI) technique
has been investigated for more than two decades, there are still various problems that must be solved
before it is applied to real structures. The technique, which has a significant potential to contribute to
the creation of one of the most effective SHM systems, involves the use of a single PZT for exciting
and sensing of the host structure. In this paper, studies applied for the past decade related to the EMI
technique have been reviewed to understand its trend. In addition, new concepts and ideas proposed
by various authors are also surveyed, and the paper concludes with a discussion of the potential
directions for future works.

Keywords: structural health monitoring; piezoelectric transducers; electromechanical impedance;

non-destructive testing; impedance-based health monitoring

1. Introduction
The birth of smart materials such as piezoelectric (PZT) transducers has aided in revolutionizing
the field of structural health monitoring (SHM) based on non-destructive testing (NDT) methods.
One of the relatively new NDT methods, known as the electromechanical impedance (EMI)
technique, has been a focus of international research for two decades; however, it has not been
fully commercialized due to many existing limitations. In general, SHM is usually applied on-site,
where the majority of inspection work is visual [1]. For this reason, inspectors require knowledge
of the possible damage locations prior to inspection. Furthermore, inspecting a large scale civil
infrastructure can be dangerous and time consuming, leading to an inefficient SHM system. In addition,
critical part inspections of structures usually require NDT methods that involve large and heavy
equipment, increasing the difficulty when monitoring structures such as cable-stayed and long-span
suspension bridges.
A well-designed SHM system can minimize the overall maintenance costs of a structure by
detecting damage at an early age, allowing action to be taken early to prevent further damage.
One possible approach is the EMI technique, as it uses a single PZT transducer that acts as both actuator
and sensor [2]. The advantage of the technique is that it has the ability to detect internal damage at a
relatively low cost. In addition, since it uses high frequency range excitation, vibrations caused from
the outside environment including vehicles and wind will not have a significant influence on the EMI
technique. However, most of the research has been applied under a well-controlled environment, or
inside a laboratory. In addition, issues such as durability and repeatability performance of the EMI
technique are being continuously investigated.

Sensors 2018, 18, 1307; doi:10.3390/s18051307 www.mdpi.com/journal/sensors

Sensors 2018, 18, 1307 2 of 18

Excellent reviews of the EMI technique can be found in previous studies, including Park et al. [3],
Yang et al. [4] and Annamdas and Soh [5]. For this reason, our review of the EMI technique in this
study is mainly focused on the investigations performed during the past decade, to give readers an
understanding of how the technique has been improved thus far. The first two parts of the study
focus on the theory and how to perform the EMI technique in order to give a quick overview of the
EMI technique. The next part reviews applications and studies related to frequency ranges of the
EMI technique, and how several authors proposed a method of creating a suitable frequency range if
there are none to select from. In addition, various damage identification studies using artificial neural
network algorithms are reviewed, and practical issues of the EMI technique are discussed. The finite
element modeling (FEM) simulation work applied by researchers is also discussed. The last part of this
study deals with future issues including EMI technique such as bond durability, PZT deterioration,
limited sensing range, reference signature issues and the possibility of incorporating the use of a drone
into the EMI technique for an efficient SHM system.

2. Theory behind the EMI Technique

The theoretical development of the EMI technique was first introduced by Liang et al. [2] and
has been studied by numerous researchers ever since. This technique utilizes a single PZT transducer,
usually with a size of 10 mm2 , to excite the host structure above 20 kHz, at which point the response
is acquired using the same PZT transducer. The PZT is typically attached to the surface of the host
structure using an adhesive, and a low voltage of 1V is used to deliver excitation to the structure.
The 1-D model introduced by Liang et al. [2] is shown in Equation (1), which confirms that the electrical
admittance Y (ω ) (inverse of impedance) of the PZT is directly related to the mechanical impedance
of the structure Zs (ω ). Thus, any changes in the properties of the structure can be identified by
monitoring the changes in the electrical impedance of the attached PZT transducer. The remaining
E
variables of ω, a, ε T33 , δ, d23x , Y xx represent the input frequency, geometric constant, dielectric constant,
loss tangent, coupling constant and Young’s modulus, respectively. Since the dielectric constant ε T33 is
temperature-sensitive, affecting the imaginary part of the impedance only, ideally the real part of the
impedance signature should be used for the EMI technique, as variations in the signature can cause
false alarms [3,6].  
T Zs (ω ) 2 E
Y (ω ) = iωa ε 33 (1 − iδ) − d Y (1)
Zs (ω ) + Za (ω ) 3x xx
Various authors including Sun et al. [7], Zhou et al. [8], Esteban [9] and Park et al. [10] have also
pioneered theoretical work regarding the EMI method during that time, where Xu and Liu [11] was the
first to consider the adhesive layer regarding the impedance model where it was represented as 1-D
spring-mass-damper system, placing in series with the structure. Regarding 2-D impedance models,
Zagrai and Giurgiutiu [12] derived a theoretical model for a circular 2-D structure and validated with
the results from experiments. The model considers axial and flexural vibrations of a target structure
taking into account for both structural and sensor dynamics. The proposed model (Equation (2))
showed promising results compared to the experimental data where ϕ a = ωr a /c. Here, r a is the radius
of a disk, k2p is the planar coupling, χ(ω ) is the dynamic stiffness ratio, ν is Poisson’s ratio, J0 and J1
are the Bessel functions of first kind, order zero and one, respectively.
#)−1
k2p
( "
  (1 + ν) J1 ( ϕ a )
Z (ω ) = iωC 1 − k2p × 1+ 2
(2)
1 − k p ϕ a J0 ( ϕ) − (1 − ν) J1 ( ϕ a ) − χ(ω )(1 + ν) J1 ( ϕ a )

Bhalla et al. [13] proposed a simplified 2-D impedance models considering shear lag effect caused
by the adhesive bond layer between the PZT transducer and the host structure (Figure 1). The derived
model shown in Equation (3) was compared with a previous work of Bhalla and Soh [14] showing the
shear lag phenomenon reasonably well. Here, l and h are the dimensions of the PZT, T is the complex
Sensors 2018, 18, 1307 3 of 18

Sensors 2018, 18, x FOR PEER REVIEW    3 of 18 
tangent ratio theoretically equal to tan(κl)/κl where κ is the wave number, and Za,e f f and Zs,e f f are
the effective impedance of the PZT and the host structure, respectively
  is the complex tangent ratio theoretically equal to tan( [14].
)/   where    is the wave number, and 
,   and  , are the effective impedance of the PZT and the host structure, respectively [14]. 
" ! #
l2 T 2d231 Y E 2d231 Y E Za,e f f
ε 33 − 2
(1 − ν) (21 − ν) Zs,e f f ,eq , + Za,e f f
Y = G + Bj = 4ωj + T (3)
4 h   (3) 
1 1 , , ,

 
Figure 1. Bond layer and piezoelectric (PZT) transducer deformation [14]. 
Figure 1. Bond layer and piezoelectric (PZT) transducer deformation [14].

Realizing the limitations (such as PZT shape and size) of the 1‐D and 2‐D models developed by 
Realizing the limitations (such as PZT shape and size) of the 1-D and 2-D models developed
various authors, Annamdas and Soh [15,16] presented a new 3‐D model utilizing 3D actuations and 
by various
PZT  authors,
transducer.  The  Annamdas
model  was  andvalidated 
Soh [15,16] with presented a newof 
experiments  3-D model utilizing
embedded  3D actuations
and  surface  bonded 
and
transducer  specimens.  However,  these  single  PZT‐structure  interaction  models  presented  bonded
PZT transducer. The model was validated with experiments of embedded and surface another 
transducer specimens. However, these single PZT-structure interaction models presented another
issue when modeling multiple PZT‐structure scenarios as previous models neglected the mass of the 
issue when modeling multiple PZT-structure scenarios as previous models neglected the mass of
PZT transducers. Thus, Annamdas and Soh [17] developed a multiple PZT‐structure (MPZT‐S) model 
theconsider 
to  PZT transducers.
the  mass Thus, Annamdas
influence  and Soh
of  multiple  PZT [17] developedThe 
transducers.  a multiple
proposed  PZT-structure (MPZT-S)
model  for  predicting 
model to consider the mass influence of multiple PZT transducers. The
impedance by the authors is shown in equation 4. Here, N is the number of PZT transducers with the  proposed model for predicting
impedance by
dimension  L,  the authors
W  and  H.  is shown1in Equation ⁄ 1 (4).1Here, 2 N, is the number
⁄ 1 of  PZT
and  transducers
,  ,   with the
are  the 
dimension L, W and H. Y R = Y ( 1 − ν ) / ( 1 + ν )( 1 − 2ν ) , R = ν/
response factors along directions X, Y and Z, respectively. Lastly, the equations for  ( 1 − ν ) and λ 1 , λ 2 , λ 3 are the response
,    and   
factors along directions X, Y and Z,
can be found in the aforementioned literature.  respectively. Lastly, the equations for A 0 , C 0 and E0 can be found
in the aforementioned literature.
  sin
A 2 jωLW [ ε + Y { d λ {[ N A sin kL − d ] + R [ NC sin kW − d ]+
Y =N cos 2H 2 33 R 31 1 0 sin 31 0 32
(4) 
R[ E0 kcos
cos k2H 2 − d 33 ]} + d
32 2 λ { R [ N A sin
0sin kL − d 31 ] + [ NC 0 sin kW
− d 32 ]+
(4)
R[ E0 k cos k2H − d33 ]} + d33 λ3 { Rcos [ N A20 sin kL − d31  ] + R[ NC0 sin kW − d32 ]+
[ E0 k cos k2H − d33 ]}}]
3. Applying the EMI Technique 
3. Applying the EMI Technique

3.1. Impedance Measuring Hardware 
3.1. Impedance Measuring Hardware
Since the introduction of the EMI technique, it has generally been applied using an impedance 
Since the introduction of the EMI technique, it has generally been applied using an impedance
analyzer such as the Agilent 4194 A. However, the high cost of such devices created a demand for 
analyzer such as the Agilent 4194 A. However, the high cost of such devices created a demand for
lower‐cost ways of applying the EMI technique, as an impedance analyzer can cost up to US$40,000. 
lower-cost ways of applying the EMI technique, as an impedance analyzer can cost up to US$40,000.
Peairs et al. [18] introduced a low‐cost method for applying the EMI technique using an FFT analyzer 
Peairs et al. [18] introduced a low-cost method for applying the EMI technique using an FFT analyzer
with the simple circuit (voltage divider) shown in Figure 2a. Using this approach, the impedance (Z) 
with the simple circuit (voltage divider) shown in Figure 2a. Using this approach, the impedance (Z)
can be approximated by dividing the input voltage (V
can be approximated by dividing the input voltage (Vi ) iby ) by the current through the sensing resistor 
the current through the sensing resistor (I).
(I).  The 
The development 
development of  a  low‐cost 
of a low-cost EMI  technique 
EMI technique has  continued 
has continued to  be  investigated 
to be investigated by  various 
by various researchers.
researchers. Xu and Giurgiutiu [19] used a function generator HP33120A with a two‐channel DAQ 
Xu and Giurgiutiu [19] used a function generator HP33120A with a two-channel DAQ card, and more
card, and more work on the low‐cost system can be found in the work of Baptista and Vieira Filho. 
[20], Bhalla et al. [21] and Panigrahi et al. [22]. 

 
Sensors 2018, 18, 1307 4 of 18

work on the low-cost system can be found in the work of Baptista and Vieira Filho. [20], Bhalla et al. [21]
and Panigrahi et al. [22].
Sensors 2018, 18, x FOR PEER REVIEW    4 of 18 

   
(a)  (b) 

Figure 2.  Impedance

Figure 2. Impedance  measuring
measuring  device: (a) circuit 
device:  (a)  circuit for
for  approximating
approximating  impedance
impedance  [18];
[18];  (b)
(b) photo
photo of
of 
AD5933 evaluation board.
AD5933 evaluation board. 

The EMI technique is a local damage detection method that is very effective for assessing the 
The EMI technique is a local damage detection method that is very effective for assessing the
health of a nearby area. However, it is a method that can be difficult to manage when monitoring 
health of a nearby area. However, it is a method that can be difficult to manage when monitoring large
large  structures, 
structures, as it mayas  involve
it  may  involve 
hundreds hundreds  to  thousands 
to thousands of  piezoelectric 
of piezoelectric transducers. transducers.  As  the 
As the traditional
traditional wired systems with a large number of sensors can be very costly, this has highlighted the 
wired systems with a large number of sensors can be very costly, this has highlighted the significance of
significance of wireless sensor nodes. The benefits of the wireless systems are well summarized by 
wireless sensor nodes. The benefits of the wireless systems are well summarized by Spencer et al. [23]
Spencer 
and Lynch et and
al.  [23] 
Loh and 
[24]. Lynch  and 
Thus far, Loh  [24]. 
various Thus 
authors havefar, researched
various  authors 
wirelesshave  researched 
systems wireless 
[25,26] with the
systems [25,26] with the AD5933 impedance measurement chip manufactured by Analog Devices, 
AD5933 impedance measurement chip manufactured by Analog Devices, Inc. At the size of a small
Inc. At the size of a small coin, the chip itself is equipped with FFT functionality, an analog to digital 
coin, the chip itself is equipped with FFT functionality, an analog to digital converter and digital to
converter 
analog and  digital 
converter. to  analog 
In addition, the converter. 
company alsoIn  addition,  the  company 
commercialized a small also  commercialized 
impedance measuringa device
small 
impedance  measuring  device  using  the  chip  known  as  the  ‘AD5933  Evaluation 
using the chip known as the ‘AD5933 Evaluation Board’ shown in Figure 2b, which can measure the Board’  shown  in 
Figure 2b, which can measure the impedance up to 100 kHz. The board is currently sold for less than 
impedance up to 100 kHz. The board is currently sold for less than US$100, and many researchers
US$100, and many researchers have investigated the EMI technique using this device. 
have investigated the EMI technique using this device.
One  of
One of  the
the  advantages
advantages  of of  the
the  EMI
EMI  technique
technique  isis  that
that  aa  single
single  PZT
PZT  transducer
transducer  acts
acts  as
as  both
both  an
an 
actuator and sensor, simultaneously. In general, PZT ceramics have been used for majority of research 
actuator and sensor, simultaneously. In general, PZT ceramics have been used for majority of research
related  to 
related to the the 
EMIEMI  technique 
technique where 
where piezoelectric 
piezoelectric diaphragms 
diaphragms (buzzers) 
(buzzers) have 
have also also  to
proven proven  to  be 
be effective
effective  for  conducting  the  technique  [27–29].  However,  due  to  the 
for conducting the technique [27–29]. However, due to the ceramic nature of the conventional PZT ceramic  nature  of  the 
conventional PZT ceramics, the brittle property makes them vulnerable to breakage and the difficulty 
ceramics, the brittle property makes them vulnerable to breakage and the difficulty in attaching it onto
in attaching it onto curved surfaces cannot be ignored. For this reason, various authors [30–35] have 
curved surfaces cannot be ignored. For this reason, various authors [30–35] have investigated using
investigated using macro‐fiber composite (MFC) created from NASA Langley Research Center [36]. 
macro-fiber composite (MFC) created from NASA Langley Research Center [36]. MFC transducer
MFC transducer consists of multiple layers of adhesive, interdigitated electrodes and polyimide film 
consists of multiple layers of adhesive, interdigitated electrodes and polyimide film where piezoelectric
where piezoelectric materials are sandwiched between. This allows it to be both flexible and durable, 
materials are sandwiched between. This allows it to be both flexible and durable, making it suitable to
making it suitable to be attached onto curved surfaces such as pipe structures. Thus, one can expect 
be attached onto curved surfaces such as pipe structures. Thus, one can expect to see more research
to see more research using MFC transducers for EMI technique in the future. 
using MFC transducers for EMI technique in the future.

3.2. Statistical Metrics for Damage Quantification 
3.2. Statistical Metrics for Damage Quantification
When the properties of a structure change (e.g., due to damage, variations in temperature, etc.), 
When the properties of a structure change (e.g., due to damage, variations in temperature, etc.),
the impedance signature also changes, with greater variations subjected to larger damage. For this 
the impedance signature also changes, with greater variations subjected to larger damage. For this
reason, once the impedance signatures are acquired, it is necessary to quantify the severity of damage. 
reason, once the impedance signatures are acquired, it is necessary to quantify the severity of damage.
In  general,
In general,  there
there  are
are  44  different
different  statistical
statistical  equations
equations  used
used  for
for  the
the  quantitative
quantitative  assessment
assessment  of
of  the
the 
signatures [37]. These are root mean square deviation (RMSD), mean absolute percentage deviation 
signatures [37]. These are root mean square deviation (RMSD), mean absolute percentage deviation
(MAPD), covariance (Cov) and correlation coefficient (CC), and are represented in Equations (5)–(8). 
Here,  Re   represents  the  reference  impedance  signature  (real  part)  and  Re the 
corresponding signature (real part). N is the number of impedance signatures, with the symbols  ̅  
signifying mean values and    signifying standard deviation. 
Sensors 2018, 18, 1307 5 of 18

(MAPD), covariance (Cov) and correlation coefficient (CC), and are represented in Equations (5)–(8).
Here, Re( Zk )i represents the reference impedance signature (real part) and Re( Zk ) j the corresponding
signature (real part). N is the number of impedance signatures, with the symbols Z signifying mean
values and σZ signifying standard deviation.
 h i2 1/2
N N
RMSD = ∑ k =1
Re( Zk ) j − Re( Zk )i /∑ k =1
[ Re( Zk )i ]2 (5)

N
1 h i 
MAPD = ∑ Re ( Z ) − Re ( Z ) /Re ( Z ) (6)
 
k j k i k i
N

k =1
N
1 h
i 
∑


Cov = Re( Zk ) j − Re Z j
· Re( Zk )i − Re Z i (7)
N k =1
N
1 h
i 
∑


CC = Re( Zk ) j − Re Z j
· Re( Zk )i − Re Z i (8)
NσZj σZi k =1

Thus far, most of the researchers who have been investigating the EMI technique have used
one of these 4 statistical equations, with the RMSD metric having been used the most. Tseng and
Naidu [38] investigated the performance of the 4 different statistical metrics, and it was found through
experiments that RMSD and MAPD were more suitable for locating and characterizing the growth
of damage, whereas covariance and CC were more suitable for identifying the increase in damage
size at a fixed location. Tawie and Lee [39] used the EMI technique to monitor concrete curing and
strength gain for up to 28 days using the three statistical metrics (RMSD, MAPD and CC). The PZT
transducer was attached to the 150 mm3 sized concrete specimen where impedance was measured on
days 7, 14 and 28. From the study, the authors discovered that MAPD correlated better than both the
RMSD and CC metrics. Xu and Jiang [40] investigated the EMI technique by monitoring bolt loosening
of a concrete-steel composite girder using RMSD, MAPD and CCD. A number of PZT transducers
were surface-bonded to the upper flange of the steel girder and the concrete slab, where bolt loosening
was identified by monitoring impedance signatures. Hu et al. [41] focused on detecting damage on a
concrete slab (500 mm × 300 mm × 50 mm) using RMSD, CCD (CCD = 1 − CC) and a new damage
index proposed by the authors, Ry/Rx. The results showed that the frequency range below 100 kHz
performed better than ranges above this frequency range for damage identification, with the authors
confirming that the proposed new damage index showed better results. Wandowski et al. [42] used
Chessboard Distance (CB), a statistical metric other than the aforementioned four equations, to evaluate
the performance of the EMI technique on composite structures subjected to damage. Here, the authors
found that CB performed better than RMSD as CB values increased, with an increased level of damage
at different temperatures.
Other statistical metrics for analyzing the impedance signatures include average square deviation
(ASD), united mechanical impedance (UMI), ellipse damage index (EDI) and others, and new metrics
in the field of the EMI technique are being proposed by various researchers [43–45]. However, the
use of a single type of metric for EMI based damage detection has some limitations, as small damage
at close range to the PZT transducer can present the same results as larger damage further away.
One approach to overcoming this problem could be to combine several statistical metrics to increase
the potential of correctly identifying and locating structural damage.

4. Investigations on the EMI Technique

4.1. Applications
The real application of the EMI technique was adopted for the first time by Annamdas and Yang
(2012) in Singapore, Telok Blangah district [46]. The work presented monitoring results of the soil
excavation conducted for construction of new mass road transport station using the EMI technique.
Sensors 2018, 18, 1307 6 of 18

PZT transducers were installed on a temporary support structure (for preventing soil collapse during
construction) where impedance signatures were monitored up to a year. Although no significant
damage was reported in the structure, it was found that the EMI technique was effective detecting
changes in loading caused from the surrounding soil.
Although that there are still many issues to be dealt with regarding the EMI technique, one can
expect to see more applications of the EMI technique onto real structures in the future. Since the
EMI technique can detect any changes in structural property, the range of its application is very wide.
The research for this technique has reached beyond mechanical and civil engineering fields as various
authors have attempted the EMI technique to be used for biological applications. Tabrizi et al. [47]
investigated using the EMI technique for assessing the stability of dental implants after surgical
placement. The experiment involved chemically degrading the prepared specimens with nitric acid
where calcium loss amount was correlated with the impedance signature. Regardless of the promising
results obtained from the experiments, the authors stated that the approach should be tested for
repeatability to bring a step closer to be accepted in the field of dental implantology. Bhalla and
Suresh examined the possibility of using the EMI technique for monitoring conditions of human and
rabbit bones [48]. The study included detecting cracks and fracture as well as the healing process
after a fracture. In addition, changes in bone density was also monitored using the conductance of
the signatures.
Regarding mechanical and civil engineering field applications, the EMI technique has proven
to be effective in various areas including crack and load damage of various structures, debonding
of composite adhesive layer, corrosion of metallic structures, bolt loosening of bridge components,
prestressed concrete (PSC) related force loss and concrete strength predictions. Detecting a crack or
notch using the EMI technique of structures can be said to be one of the most investigated areas of the
EMI technique. This includes detecting damage subjected to different structure types such as metal,
composites and concrete structures [49–59]. A good correlation relationship can be seen between the
distance and severity of damage compared to a statistical metric. In general, closer and severer damage
will results in a larger statistical metric value where such results can be seen in some of the recent
studies. Corrosion and bolt loosening detection of metallic structures have also been continuously
investigated over the past several years. Regarding composite materials, since adhesive bonding is
a preferred choice over the mechanical fastening approach, adhesive layer monitoring has been an
important field in EMI technique. Some of the latest research includes composite-to-composite and
composite-to-concrete debonding detections [60–65]. Kim et al. [66] investigated on the possibility of
utilizing the EMI technique for predicting the prestress forces loss for PSC girder bridges. In this work,
impedance signatures were measured at 5 prestress levels where the frequency was measured up to
1000 kHz. From the results, signature variations were clearly visible subjected to different prestress
levels. Other studies related to concrete have also been active during the past several years as many
authors applied the EMI technique for predicting concrete strength. Lu et al. [67] proposed an idea of
using ‘Smart Probe’ to monitor the strength development of cementitious materials. Dynamic modulus
of elasticity of cementitious materials was predicted with the proposed analytical model where it
showed good correlation with the compressive strength of the mortar. Talakokula et al. [68] focused on
monitoring the hydration process of cementitious materials for the first 28 days. The study measured
both the real and imaginary part of admittance (inverse of impedance) where the results showed that
the magnitude of the signatures gradually decreased with time. The authors conclude by stating that
the approach proposed in the work has the potential to be applied to in situ hydration monitoring of
reinforced concrete structures. More research in the field of concrete strength prediction can be found
in recent studies [69–75].

4.2. Selecting or Creating a Suitable Frequency Range

Although that the frequency domain analysis of the EMI technique has been researched in vast
amounts, very little has been published related to time domain analysis methods where it uses the
 Sensors 2018, 18, 1307 7 of 18

excitation circuit proposed by Baptista and Vieira [20]. From various authors, it has been known that
time domain analysis methods are more sensitive for detecting damage compared to the conventional
frequency domain analysis of the EMI technique [76–78]. However, it was found that the efficiency lied
within detecting existence of damage and not on its exact location. For this reason, the remainder of
the study will continue to  focus of the frequency domain analysis of the EMI technique. The
Sensors 2018, 18, x FOR PEER REVIEW  frequency
7 of 18 
range of the EMI technique is selected by manually seeking multiple peaks and setting that specific
range for analysis. In previous studies, the frequency ranges used for the EMI technique have been up
specific range for analysis. In previous studies, the frequency ranges used for the EMI technique have 
to 1000 kHz, with higher frequency ranges resulting in a smaller sensing region and lower frequency
been up to 1000 kHz, with higher frequency ranges resulting in a smaller sensing region and lower 
ranges resulting in a larger sensing region. In general, a frequency range below 400 kHz is ideal,
frequency ranges resulting in a larger sensing region. In general, a frequency range below 400 kHz is 
with ranges below 80 kHz covering a relatively large area. Regarding frequency range selection,
ideal, with ranges below 80 kHz covering a relatively large area. Regarding frequency range selection, 
Yang and Divsholi [79] applied the EMI technique to a concrete mass in the frequency range from
Yang and Divsholi [79] applied the EMI technique to a concrete mass in the frequency range from 30 
30 kHz to 400 kHz. This range was divided into 6 sub-frequency groups (30–99.9 kHz, 100–149.9 kHz,
kHz to 400 kHz. This range was divided into 6 sub‐frequency groups (30–99.9 kHz, 100–149.9 kHz, 
150–199.9 kHz, 200–249.9 kHz, 250–299.9 kHz, and 300–400 kHz) to evaluate the performance of each
150–199.9 kHz, 200–249.9 kHz, 250–299.9 kHz, and 300–400 kHz) to evaluate the performance of each 
sub-frequency range against damage. The results of the experiment showed that damage close to the
sub‐frequency range against damage. The results of the experiment showed that damage close to the 
PZT changed the impedance signatures significantly in the higher frequency ranges, whereas damage far
PZT changed the impedance signatures significantly in the higher frequency ranges, whereas damage 
away from the PZT changed the signatures significantly in the lower frequency ranges. Min et al. [80]
far away from the PZT changed the signatures significantly in the lower frequency ranges. Min et al. 
investigated the autonomous selection of a suitable frequency range for the EMI technique using an
[80] investigated the autonomous selection of a suitable frequency range for the EMI technique using 
artificial neural network technique. The proposed artificial neural network (ANN) scheme allowed one
an artificial neural network technique. The proposed artificial neural network (ANN) scheme allowed 
one  to to automaticallyidentify 
automatically  identifya 
a damage‐sensitive 
damage-sensitive frequency
frequency region
region toto 
bebe 
used forfor 
used  the the 
experiment applied in
experiment 
the study.
applied in the study.   
If no peaks can be found when sweeping the frequency ranges to search a region with multiple
If no peaks can be found when sweeping the frequency ranges to search a region with multiple 
peaks, peaks, the technique 
the  EMI  EMI techniquemay mayfail fail to detect
to  detect  anystructural 
any  structuraldamage. 
damage. Without 
Without any
any resonance,
resonance,  the changes
the 
in the impedance signature can be quite small. To solve this problem, Na and Lee proposed a method
changes in the impedance signature can be quite small. To solve this problem, Na and Lee proposed 
of creating a resonance frequency range when applying the EMI technique (Na and Lee [81]. The idea
a method of creating a resonance frequency range when applying the EMI technique (Na and Lee 
was to sandwich a metal disc between the PZT and the structure, where the resonance frequency
[81]. The idea was to sandwich a metal disc between the PZT and the structure, where the resonance 
of the metal disc is maintained regardless of the properties of the structure (Figure 3). Thus, the
frequency of the metal disc is maintained regardless of the properties of the structure (Figure 3). Thus, 
resonance frequency range could be changed simply by changing the size of the metal disc. This idea
the resonance frequency range could be changed simply by changing the size of the metal disc. This 
was extended by using a steel wire to apply the EMI technique [82], which allowed the EMI technique
idea was extended by using a steel wire to apply the EMI technique [82], which allowed the EMI 
to beto applied
technique  to any
be  applied  kindkind 
to  any  of surface, as the
of  surface,  as tip
the of the
tip  of wire could
the  wire  be attached
could  virtually
be  attached  anywhere.
virtually 
anywhere.  The  resonance 
The resonance frequency 
frequency ranges becould 
ranges could changed be  by
changed  by  using 
using different different 
sized sized  PZT for the
PZT transducers
transducers  for  EMI
steel wire the  steel  wire  EMI 
technique. Thus,technique. 
when the EMI Thus,  when  the 
technique EMI  technique 
is applied is  applied 
and no resonance and  no  with
frequencies
resonance frequencies with peaks are found, an alternative solution is to create a suitable frequency 
peaks are found, an alternative solution is to create a suitable frequency range with multiple peaks for
range with multiple peaks for successful damage detection. 
successful damage detection.  

 
Figure 3. Resonance frequency creating concept [81]. 
Figure 3. Resonance frequency creating concept [81].

4.3. Artificial Neural Networks and EMI Technique 
4.3. Artificial Neural Networks and EMI Technique
The  ultimate  goal  of  SHM  technique  is  to  identify  the  location  of  damage  and  its  size,  and 
The ultimate goal of SHM technique is to identify the location of damage and its size, and
estimate the remaining service life of the target structure. Using the EMI technique, many authors 
estimate the remaining service life of the target structure. Using the EMI technique, many authors
have studied how the impedance signatures change when subjected to different sizes and locations 
of damage. One well known technique of locating damage is by using an artificial neural network 
(ANN), and various authors have already evaluated how an ANN can perform when incorporated 
into the EMI technique. When an ANN is applied to the EMI technique, a decent amount of training 
is usually required for the algorithm to successfully identify damage. Some common types of ANN 
techniques  can  be  found  in  Giurgiutiu  and  Kropas‐Hughes  [83],  in  which  the  researchers  have 
Sensors 2018, 18, 1307 8 of 18

have studied how the impedance signatures change when subjected to different sizes and locations
of damage. One well known technique of locating damage is by using an artificial neural network
(ANN), and various authors have already evaluated how an ANN can perform when incorporated
into the EMI technique. When an ANN is applied to the EMI technique, a decent amount of training
is usually required for the algorithm to successfully identify damage. Some common types of ANN
techniques can be found in Giurgiutiu
Sensors 2018, 18, x FOR PEER REVIEW    and Kropas-Hughes [83], in which the researchers have applied 8 of 18 
the EMI technique to aluminum alloy based circular plates of 100 mm diameter and 1.6 mm thickness.
Thealgorithm for the aforementioned study. Min et al. applied a multi‐layer feed‐forward NN with a 
probabilistic neural network (PNN) algorithm was applied with a featured extraction algorithm for
theback‐propagation algorithm in conjunction with the EMI technique for damage identification [84]. A 
aforementioned study. Min et al. applied a multi-layer feed-forward NN with a back-propagation
series of experiments was carried out to evaluate the proposed technique subjected to loose bolts and 
algorithm in conjunction with the EMI technique for damage identification [84]. A series of experiments
crack damages on both lab scale pipe structure and on a box girder bridge. Na and Lee [85] used a 
was carried out to evaluate the proposed technique subjected to loose bolts and crack damages on both
lab glass‐epoxy composite plate with a size of 200 mm × 200 m × 3.5 mm with 3 PZT transducers attached 
scale pipe structure and on a box girder bridge. Na and Lee [85] used a glass-epoxy composite plate
withto each of the 3 corners of the plate, with the purpose of the study being to predict the location of 
a size of 200 mm × 200 m × 3.5 mm with 3 PZT transducers attached to each of the 3 corners of the
damage 
plate, with the using  the  developed 
purpose of the studyPNN  beingalgorithm. 
to predict In 
thethis  study, 
location 6  different 
of damage using areas  were  classified 
the developed PNN for 
testing, and 24 damaged holes out of 30 were correctly predicted, an 80% accuracy rate. The study 
algorithm. In this study, 6 different areas were classified for testing, and 24 damaged holes out of 30
concludes with the statement that increasing the number of training data could improve the outcome 
were correctly predicted, an 80% accuracy rate. The study concludes with the statement that increasing
theof  the  results. 
number Palomino 
of training et  al.  improve
data could [86]  focused  on  the  classification 
the outcome of the results.of Palomino
two  different  types 
et al. [86] of  damages 
focused on
the(cracks and rivet losses) on an aluminum panel using PNN and fuzzy cluster analysis methods. The 
classification of two different types of damages (cracks and rivet losses) on an aluminum panel
authors 
using PNN and discovered  that  analysis
fuzzy cluster the  location  of  the 
methods. ThePZT  transducer 
authors discovered is  that
vital the
for  successful 
location of thedamage 
PZT
identification. Oh et al. [87] combined ANN with the EMI technique for predicting the strength of 
transducer is vital for successful damage identification. Oh et al. [87] combined ANN with the EMI
concrete by monitoring the impedance signatures for 28 days during the curing of the concrete. For 
technique for predicting the strength of concrete by monitoring the impedance signatures for 28 days
training of the ANN algorithm, water‐and‐cement ratio, curing time, curing temperature, maturity 
during the curing of the concrete. For training of the ANN algorithm, water-and-cement ratio, curing
and CC value calculated from the impedance signatures were used (Figure 4). The proposed ANN 
time, curing temperature, maturity and CC value calculated from the impedance signatures were
model estimated the strength of prepared concrete specimens with negligible error. de Oliveira et al. 
used (Figure 4). The proposed ANN model estimated the strength of prepared concrete specimens
with[88] proposed an enhanced strategy for damage detection via Savitzky–Golay (SG) filter, PNN and 
negligible error. de Oliveira et al. [88] proposed an enhanced strategy for damage detection via
simplified  fuzzy ARTMAP 
Savitzky–Golay (SG) filter, PNN network,  with  these 
and simplified fuzzythree  methods applied 
ARTMAP network, with to an  experiment using 
these three methods an 
aluminum plate containing three PZT transducers. The results showed a promising outcome in terms 
applied to an experiment using an aluminum plate containing three PZT transducers. The results
of successfully detecting damage. 
showed a promising outcome in terms of successfully detecting damage.

 
Figure 4. Concept of concrete strength prediction with artificial neural network (ANN) [87]. 
Figure 4. Concept of concrete strength prediction with artificial neural network (ANN) [87].

The trend of combining the ANN algorithm with the EMI technique can be said to be increasing, 
The trend of combining the ANN algorithm with the EMI technique can be said to be increasing,
as  the  statistical  metrics  such  as  RMSD,  MAPD,  covariance  and  CC  are  not  sufficient  to  quantify, 
as the statistical metrics such as RMSD, MAPD, covariance and CC are not sufficient to quantify, locate
locate and identify what kind of damage the structure has received. 
and identify what kind of damage the structure has received.
4.4. Practical Issues 
Since the introduction of the EMI technique, countless studies have been carried out in this area, 
with  many  reporting  promising  results.  However,  most  of  the  studies  performed  were  under 
laboratory conditions or mainly theoretically based, which raises questions regarding their potential 
in real‐life applications, particularly under harsh environments. Although structure damage changes 
the impedance signature, other factors such as variations in temperature and the durability of PZT 
transducers can also cause the signatures to change as well. Yang et al. [4] investigated the signature 
repeatability of PZT transducers both indoor and outdoor up to 15 months, and found that silicone 
 Sensors 2018, 18, 1307 9 of 18

4.4. Practical Issues

Since the introduction of the EMI technique, countless studies have been carried out in this
area, with many reporting promising results. However, most of the studies performed were under
laboratory conditions or mainly theoretically based, which raises questions regarding their potential in
real-life applications, particularly under harsh environments. Although structure damage changes
the impedance signature, other factors such as variations in temperature and the durability of PZT
transducers can also cause the signatures to change as well. Yang et al. [4] investigated the signature
repeatability of PZT transducers both indoor and outdoor up to 15 months, and found that silicone
rubber protection over the PZT resulted in RMSD values that were lower than unprotected PZTs
(Figure 5). In addition, the authors
Sensors 2018, 18, x FOR PEER REVIEW    experimentally evaluated how the bonding layers of 9 of 18  the PZT
transducers and variations in the temperature influenced the impedance signatures. Sun et al. [6]
applied 
appliedthe the
EMI  EMItechnique 
techniqueunder  a  temperature 
under a temperature varying 
varyingcondition  and and
condition discovered 
discovered that  a  rise 
that in  in
a rise
temperature  seemed  to  soften  the  overall  stiffness  of  the  host  structure,  causing  the 
temperature seemed to soften the overall stiffness of the host structure, causing the resonance frequency resonance 
frequency range to shift and the peak amplitudes to change. The authors also used cross‐correlation 
range to shift and the peak amplitudes to change. The authors also used cross-correlation to minimize
to minimize the signature variation due to temperature change by horizontally shifting the signature. 
the signature variation due to temperature change by horizontally shifting the signature. Park et al. [89]
Park et al. [89] discovered that the real part of the impedance signatures should be preferred to the 
discovered that the real part of the impedance signatures should be preferred to the imaginary part,
imaginary part, as they found out that the real part signature of the free PZT patch changed only 
as they found out that the real part signature of the free PZT patch changed only negligibly with
negligibly  with  temperature 
temperature variations.  Furthermore, 
variations. Furthermore, the  authors 
the authors suggested suggested 
another another 
technique technique  to the
to compensate
compensate the signature variations due to temperature changes; many other researchers have also 
signature variations due to temperature changes; many other researchers have also studied this
studied 
area. this 
Grissoarea. 
andGrisso 
Inmanand 
[90]Inman  [90] aproposed 
proposed techniquea fortechnique 
separating for the
separating 
temperature the  temperature 
variation effect
variation 
from sensor defects, in which a frame structure was used to conduct experimentsexperiments 
effect  from  sensor  defects,  in  which  a  frame  structure  was  used  to  conduct  under various
under various temperatures. The results of their experiments reveal that the measured susceptance 
temperatures. The results of their experiments reveal that the measured susceptance slope showed a
slope  showed 
linear a  linear 
relationship torelationship 
temperatureto  temperature 
variations. variations. 
Sepehry Sepehry 
et al. [91] combinedet  al. 
ANN[91] based
combined  ANN 
on radial basis
based on radial basis function (RBF) with the EMI technique to compensate the temperature effect. A 
function (RBF) with the EMI technique to compensate the temperature effect. A steel plate and gas
steel plate and gas pipe were used for the study, and damage was artificially created by drilling a 
pipe were used for the study, and damage was artificially created by drilling a hole and loosening
hole and loosening bolt, respectively. Wandowski et al. [92] used carbon fibre reinforced polymer 
bolt, respectively. Wandowski et al. [92] used carbon fibre reinforced polymer samples to evaluate the
samples to evaluate the proposed temperature compensation algorithm. The principle of the 2‐step 
proposed temperature compensation algorithm. The principle of the 2-step algorithm is to first shift
algorithm is to first shift the signature in the horizontal direction using cross‐correlation and then in 
the signature in the horizontal direction using cross-correlation and then in the vertical direction using
the vertical direction using signal normalization with its root mean square values. 
signal normalization with its root mean square values.  

 
Figure 5. Test specimen setup testing PZT durability (c.f. Yang et al., 2008 [4]).
Figure 5. Test specimen setup testing PZT durability (c.f. Yang et al., 2008 [4]). 

Although multiple techniques have been proposed to compensate the temperature effect of the 
Although multiple techniques have been proposed to compensate the temperature effect of the
EMI technique, it is extremely difficult to compensate this effect completely. Sometimes the variations 
EMI technique, it is extremely difficult to compensate this effect completely. Sometimes the variations
in  impedance  signatures 
in impedance signaturescan can
be  be
unpredictable,  as  as
unpredictable, some  peaks 
some peaksmay 
mayshift  or or
shift increase/decrease  in  in
increase/decrease
amplitude more than others [93]. 
amplitude more than others [93].

4.5. Experiment vs. Simulation 
4.5. Experiment vs. Simulation
To To
gain  a  abetter 
gain betterunderstanding  of an
understanding of an engineering
engineering  system, 
system, modeling 
modeling is essential
is an an  essential  field 
field of of 
research,
research, and a large number of studies are published in this area each year. Considering the EMI 
and a large number of studies are published in this area each year. Considering the EMI technique,
technique, PZT‐structure interaction models have been intensively developed. However, there are 
PZT-structure interaction models have been intensively developed. However, there are limitations
limitations to these solutions when they are applied to real structures, as they can be only applied to 
simple structures such as beams, shells and plates with simple boundary conditions. As such, finite 
element  method  (FEM)  became  known  as  an  alternative  solution.  With  powerful  FEM  tools,  the 
efficiency of FE modeling is increasing every day, particularly with the advancement of computer 
technology. Figure 6 shows two impedance signatures for a 20 mm × 10 mm × 0.5 mm PZT transducer 
Sensors 2018, 18, 1307 10 of 18

to these solutions when they are applied to real structures, as they can be only applied to simple
structures such as beams, shells and plates with simple boundary conditions. As such, finite element
method (FEM) became known as an alternative solution. With powerful FEM tools, the efficiency of FE
modeling is increasing every day, particularly with the advancement of computer technology. Figure 6
shows two impedance signatures for a 20 mm × 10 mm × 0.5 mm PZT transducer where one was
Sensors 2018, 18, x FOR PEER REVIEW 
measured using AD5933 evaluation   10 of 18 
board and the other using a FEM simulation tool. The resonance
frequency ranges for both impedance signatures are within 65–80 kHz range.

 
Figure 6. Simulation vs. experiment impedance signatures on a 20 mm × 10 mm PZT. 
Figure 6. Simulation vs. experiment impedance signatures on a 20 mm × 10 mm PZT.

A well summarized work on FE‐based simulation of PZT‐structure interaction can be found in 
A well summarized work on FE-based simulation of PZT-structure interaction can be found in
the PhD dissertation by Lalande [94], in which the author used ANSYS software to show promising 
the PhD dissertation by Lalande [94], in which the author used ANSYS software to show promising
outcomes  on  ring  and  shell  structures.  Tseng  and  Wang  [95]  applied  both  experimental  and 
outcomes on ring and shell structures. Tseng and Wang [95] applied both experimental and numerical
numerical  studies  on  2  concrete  beam  structures,  which  led  to  the  result  of  the  RMSD  values 
studies on 2 concrete beam structures, which led to the result of the RMSD values increasing when crack
increasing  when  crack  damage  increased  and  when  crack  damage  occurred  closer  to  the  PZT 
damage increased and when crack damage occurred closer to the PZT transducers. Lim and Soh [49]
transducers. Lim and Soh [49] investigated the possibility of using the EMI technique for estimating 
investigated the possibility of using the EMI technique for estimating the remaining fatigue life of a
the remaining fatigue life of a 1‐D aluminum structure using ANSYS 12.1, incorporating the linear 
1-D aluminum structure using ANSYS 12.1, incorporating the linear elastic fracture mechanics theory.
elastic fracture mechanics theory. In their work, the authors stated that the CCDM (cross correlation 
In their work, the authors stated that the CCDM (cross correlation deviation mean) results showed
deviation mean) results showed better outcomes compared to the RMSD results when characterizing 
better outcomes compared to the RMSD results when characterizing crack length. Yang et al. [96] used
crack length. Yang et al. [96] used ANSYS to apply the EMI technique on various models including 
ANSYS to apply the EMI technique on various models including bonding layers on an aluminum
bonding layers on an aluminum beam, and the results showed excellent agreement compared to the 
beam, and the results showed excellent agreement compared to the experimental data up to 1000 kHz.
experimental data up to 1000 kHz. In addition, the authors simulated the EMI technique from 30 °C 
In addition, the authors simulated the EMI technique from 30 ◦ C to 60◦ C in the frequency range up
to 60°C in the frequency range up to 400 kHz, also showing good agreement with the experimental 
to 400 kHz, also showing good agreement with the experimental results. Hamzeloo et al. [97] used
results.  Hamzeloo  et  al.  [97]  used  the  EMI  technique  to  perform  both  experiment  and  simulation 
the EMI technique to perform both experiment and simulation (using ABAQUS) on aluminum and
(using ABAQUS) on aluminum and steel specimens. The test specimens were hollow cylinders with 
steel specimens. The test specimens were hollow cylinders with different thicknesses, and the test was
different thicknesses, and the test was performed in the frequency range from 10 to 40 kHz. Using 
performed in the frequency range from 10 to 40 kHz. Using the RMSD metric, the results showed that
the  RMSD  metric,  the  results  showed  that  accumulation  of  damage  does  not  always  increase  the 
accumulation of damage does not always increase the RMSD values. Rugina et al. [98] reported on
RMSD  values.  Rugina  et al.  [98] reported  on  the use of  the  EMI  technique  on  thin  circular  plates, 
the use of the EMI technique on thin circular plates, comparing the results of the analytical method,
comparing the results of the analytical method, FE simulation and experiments. At undamaged state, 
FE simulation and experiments. At undamaged state, all the impedance signatures showed good
all  the  impedance  signatures  showed  good  agreement  when  compared  to  one  another  in  the 
agreement when compared to one another in the frequency range up to 40 kHz. Djemana et al. [99]
frequency range up to 40 kHz. Djemana et al. [99] focused on detecting and locating damage on an 
focused on detecting and locating damage on an aluminum beam structure with the application
aluminum  beam  structure  with  the  application  of  an  extreme  learning  machine‐based  algorithm 
of an extreme learning machine-based algorithm developed for the study. An overall accuracy of
developed  for  the  study.  An  overall  accuracy  of  84.5%  was  achieved  in  terms  of  predicting  the 
84.5% was achieved in terms of predicting the location of a single damage in structure with the aid of
location of a single damage in structure with the aid of FEM simulations. 
FEM simulations.
Thus far, the majority of the FEM simulation studies related to the EMI technique have used 
either ANSYS or ABAQUS. Since the EMI technique is a non‐modeled method that is heavily reliant 
on acquired impedance signatures, incorporating FEM may enable the achievement of the ultimate 
goal of locating, identifying and quantifying the size of damage. 

5. Future Work 
Sensors 2018, 18, 1307 11 of 18

Thus far, the majority of the FEM simulation studies related to the EMI technique have used
either ANSYS or ABAQUS. Since the EMI technique is a non-modeled method that is heavily reliant
on acquired impedance signatures, incorporating FEM may enable the achievement of the ultimate
goal of locating, identifying and quantifying the size of damage.

5. Future Work

5.1. Bond Durability, PZT Deterioration, Sensing Range and Reference Signature
For the EMI technique to be applied to SHM system, the durability of the PZT transducer and the
bonding layer must be high. While many PZT manufacturers claim that the PZT deterioration rate is
less than 5% per decade, environmental factors must be taken into account. Environmental factors
such as changes in temperature over a long period of time can lead to thermal fatigue, while rain or
sea water exposure could result in acidic and alkaline attacks. In addition, the PZT transducer should
endure both the mechanical and electrical cyclic loading. Although there are ways to protect the PZT
transducers, such as embedding them into the structure or covering them up with silicon or epoxy,
the PZTs cannot be replaced if they are damaged. In addition, changes in the properties of the cover
(silicon or epoxy) can cause the impedance signatures to change, causing a fault alarm. For this reason,
the influence of these factors needs to be eliminated through compensation algorithms, or perhaps by
using a control or multiple PZT transducers.
Regarding sensing range, the EMI technique is a local method that uses high frequency excitation.
For this reason, the sensing range is limited to several meters, depending on the properties of the PZT
and the host structure. As such, cost issues arise when dealing with large structures, as a vast number
of PZT transducers need to be installed. This shows that using the EMI technique alone to develop
SHM system is very difficult, and a global method (e.g., accelerometer on bridge cables) should be
incorporated to create an efficient SHM system. A well-known combination of the EMI technique with
the guided wave-based technique has been investigated in Giurgiutiu et al. [100] where it was found
that the EMI technique was found to be suitable for damage detection in the near field while the guided
wave based technique was more suitable for far-field damage detection [101–105]. Lastly, one of the
first steps for conducting the EMI technique is to acquirea reference signature in a ‘no damage’ state.
This reference signature is then compared to a damaged state at a later stage to identify any damage
to the structure. As it is virtually impossible to acquire a ‘no damage’ state impedance signature for
existing structures, future studies could be focused on using the EMI technique without the use of a
reference signature, either by using the method alone or in combination with other NDT techniques.

5.2. Drone EMI Technique

As the Fourth Industrial Revolution approaches, various areas of technology have become key
items of interest around the globe. Technologies such as artificial intelligence (AI) and drones can be
said to be closely related to the EMI technique. Although the amount of ANN work in conjunction
to the EMI technique is increasing every day, the combination of drones with the EMI technique
(or any other non-visual based NDT methods) is in the very beginning stage of research. Although
visual inspection-based drone monitoring technology is advancing, this technology has limitations
when searching for internal damage or even small cracks, which can be difficult to identify using a
camera. Some of the work on visual based SHM using drones can be found in Ellenberg et al. [106],
Ham et al. [107], Flammini et al. [108] and Reagan et al. [109]. On the other hand, the EMI technique is
a contact-based method which has the potential of detecting internal damage of structures.
The first work on combining the EMI technique with drone was performed by Na and Baek [110],
using a re-attachable PZT transducer using a magnet (Figure 7). In this research, the authors discovered
that the impedance signature did not change during the hovering state, regardless of the drone
vibrations. However, re-attaching the PZT transducer caused the impedance signature to change
where additional future work is required to overcome to problem. Regardless of the signature change
Sensors 2018, 18, 1307 12 of 18

due to the re-attachment, the authors found that the proposed drone-EMI based technique showed
Sensors 2018, 18, x FOR PEER REVIEW 
potential for detecting thickness  reduction in steel structures. 12 of 18 

 
Figure 7. Drone electromechanical (EMI) technique concept [110].
Figure 7. Drone electromechanical (EMI) technique concept [110]. 

5.3. Multi‐functional Sensing Possibilities 
5.3. Multi-functional Sensing Possibilities
By 2050, it has been predicted that up to 70% of the world’s population will live in urban areas 
By 2050, it has been predicted that up to 70% of the world’s population will live in urban areas
and with this fact, the importance of developing a smart‐city is becoming more and more important. 
and with this fact, the importance of developing a smart-city is becoming more and more important.
One of the vital aspects for achieving an efficient smart‐city lies within the field of sensors. With a 
One of the vital aspects for achieving an efficient smart-city lies within the field of sensors. With a
vast number of sensors installed around the city, management of the overall system may depend on 
vast number of sensors installed around the city, management of the overall system may depend on
factors such as the number of sensors. For this reason, the importance of multi‐functional sensors will 
factors such as the number of sensors. For this reason, the importance of multi-functional sensors
become more important as this may reduce the total number of sensors required for an effective smart 
will become more important as this may reduce the total number of sensors required for an effective
city. As mentioned in a previous section, the impedance signatures during the EMI technique has 
smart city. As mentioned in a previous section, the impedance signatures during the EMI technique
been experimentally proven to vary with change in temperatures. This gives the possibility of the 
has been experimentally proven to vary with change in temperatures. This gives the possibility of
EMI 
the EMItechnique  not not
technique only 
onlyto to
detect 
detectdamage 
damage but 
butalso 
alsopredict 
predictthe 
thetemperature 
temperatureof  of the 
the host  structure, 
host structure,
simultaneously. With this idea, it may be possible to reduce two sensors into a single sensor to reduce 
simultaneously. With this idea, it may be possible to reduce two sensors into a single sensor to reduce
the overall number of sensors required for a structural health monitoring system. 
the overall number of sensors required for a structural health monitoring system.
With sensing range limitations of the EMI technique, one may suggest that the real application 
With sensing range limitations of the EMI technique, one may suggest that the real application of
of such technique can be very limited. However, a future technology such as the ‘Hyperloop’ system 
such technique can be very limited. However, a future technology such as the ‘Hyperloop’ system can
can demand such technique as the EMI technique is effective for detecting damage at a very early 
demand such technique as the EMI technique is effective for detecting damage at a very early stage.
stage. 
The The  Hyperloop 
Hyperloop system issystem 
a new is 
typea  new  type  transportation
of ground of  ground  transportation  technology 
technology which which  can 
can theoretically
theoretically travel up to 1200 km per hour, moving at a speed faster than an airplane. The concept 
travel up to 1200 km per hour, moving at a speed faster than an airplane. The concept of the technology
of the technology can be seen in Figure 8 where the basic idea is that a pod travels inside a near‐
can be seen in Figure 8 where the basic idea is that a pod travels inside a near-vacuum stated tube
vacuum  to
structure stated  tube 
reduce air structure 
resistance.to 
Thereduce  air  resistance. 
two well-known USAThe  two  well‐known 
companies, Hyperloop USA  companies, 
Transportation
Hyperloop Transportation Technologies (HTT) and Virgin Hyperloop One are currently researching 
Technologies (HTT) and Virgin Hyperloop One are currently researching the technology to make it
the technology to make it into a reality where details can be found in [111]. Such advanced technology 
into a reality where details can be found in [111]. Such advanced technology will require sensors to
will  require 
detect damagesensors  to early
at a very detect  damage 
age at  a  very 
for an efficient early  age  for 
maintenance an  efficient 
system. maintenance 
Furthermore, the tubesystem. 
length
Furthermore, the tube length could be built up to several hundreds to thousands of kilometers, where 
could be built up to several hundreds to thousands of kilometers, where a large number of sensors
a large number of sensors will be possibly required. For this reason, multi‐functional sensors could 
will be possibly required. For this reason, multi-functional sensors could be an important factor to
be an important factor to achieve such technology. 
achieve such technology.  

 
Sensors 2018, 18, 1307 13 of 18
Sensors 2018, 18, x FOR PEER REVIEW    13 of 18 

 
Figure  8.  Hyperloop  concept  from  Korea  Institute  of  Civil  Engineering  and  Building  Technology 
Figure 8. Hyperloop concept from Korea Institute of Civil Engineering and Building Technology (KICT).
(KICT). 

6. Conclusions
6. Conclusions 
In this work, we have reviewed studies that have been conducted on the application of EMI
In  this  work,  we  have  reviewed  studies  that  have  been  conducted  on  the  application  of  EMI 
technique, focusing on the latest research in the field. While the EMI technique has been researched
technique, focusing on the latest research in the field. While the EMI technique has been researched 
for over two decades, there are still various problems to be solved in its application. These problems
for over two decades, there are still various problems to be solved in its application. These problems 
include the limited sensing range, the need to select a suitable frequency range, the need to compensate
include  the  limited  sensing  range,  the  need  to  select  a  suitable  frequency  range,  the  need  to 
for environmental factors such as temperature variations, efficient statistical metrics etc., which need
compensate for environmental factors such as temperature variations, efficient statistical metrics etc., 
to be solved to achieve one of the ultimate goals of locating and identifying damage type. For this
which need to be solved to achieve one of the ultimate goals of locating and identifying damage type. 
reason, one of the trends is that the studies using ANN has increased everyday where it has proven
For this reason, one of the trends is that the studies using ANN has increased everyday where it has 
to give promising outcome for the EMI technique. As well, ANN has been investigated not only for
proven to give promising outcome for the EMI technique. As well, ANN has been investigated not 
measuring structural damage but for a variety of applications including selecting suitable frequency
only for measuring structural damage but for a variety of applications including selecting suitable 
range, temperature compensation, predicting concrete strength, bolt loosening etc.
frequency range, temperature compensation, predicting concrete strength, bolt loosening etc.   
The EMI technique has great potential, with numerous studies being conducted to solve various
The EMI technique has great potential, with numerous studies being conducted to solve various 
problems. However, since most of the studies have been conducted under lab conditions, future
problems.  However,  since  most  of  the  studies  have  been  conducted  under  lab  conditions,  future 
research should be more focused on real structures in order to identify critical problems using the EMI
research should be more focused on real structures in order to identify critical problems using the 
technique. Nevertheless, continuous research by experts and advancement in computer technologies
EMI  technique.  Nevertheless,  continuous  research  by  experts  and  advancement  in  computer 
will bring new solutions and ideas, with FEM simulations becoming more effective for bringing the
technologies will bring new solutions and ideas, with FEM simulations becoming more effective for 
EMI technique one step closer to commercialization.
bringing the EMI technique one step closer to commercialization.   
Funding: This research was funded by Korea Institute of Civil engineering and building Technology (KICT),
Funding: This research was funded by Korea Institute of Civil engineering and building Technology (KICT), 
Korea, grant number 20180034-001.
Korea,  grant number 
Conflicts 20180034‐001. 
of Interest: The authors declare no conflict of interest.
Conflicts of Interest: The authors declare no conflict of interest. 
References
References 
1. Chang, P.C.; Flatau, A.; Liu, S.C. Health monitoring of civil infrastructure. Struct. Health Monit. 2003, 2,
1. 257–267. [CrossRef]
Chang, P.C.; Flatau, A.; Liu, S.C. Health monitoring of civil infrastructure. Struct. Health Monit. 2003, 2, 257–
2. Liang, C. Coupled Electromechanical Analysis of Adaptive Material System-Determination of Actuator
267. 
2. Power Consumption
Liang,  and System Energy
C.  Coupled  Electromechanical  Transfer.
Analysis  of J.Adaptive 
Intell. Mater. Syst. Struct.
Material  1994, 5, 12–20. [CrossRef]
System‐Determination  of  Actuator 
3. Park, G.; Sohn, H.; Farrar, C.R.; Inman, D.J. Overview of piezoelectric impedance-based health monitoring
Power Consumption and System Energy Transfer. J. Intell. Mater. Syst. Struct. 1994, 5, 21–20. 
3. and path forward. Shock Vib. Dig. 2003, 35, 451–464. [CrossRef]
Park, G.; Sohn, H.; Farrar, C.R.; Inman, D.J. Overview of piezoelectric impedance‐based health monitoring 
4. Yang, Y.; Lim, Y.Y.; Soh, C.K. Practical issues related to the application of the electromechanical impedance
and path forward. Shock Vib. Dig. 2003, 35, 451–464. 
4. technique in the structural health monitoring of civil structures: I. Experiment. Smart Mater. Struct. 2008, 17,
Yang, Y.; Lim, Y.Y.; Soh, C.K. Practical issues related to the application of the electromechanical impedance 
035008. [CrossRef]
technique in the structural health monitoring of civil structures: I. Experiment. Smart Mater. Struct. 2008, 
5. Annamdas, V.G.M.; Soh, C.K. Application of electromechanical impedance technique for engineering
17, 035008. 
5. structures: Review
Annamdas,  and
V.G.M.;  future
Soh,  issues.
C.K.  J. Intell. of 
Application  Mater. Syst. Struct. 2010,
electromechanical  21, 41–59.technique 
impedance  [CrossRef]for  engineering 
structures: Review and future issues. J. Intell. Mater. Syst. Struct. 2010, 21, 41–59. 

 
Sensors 2018, 18, 1307 14 of 18

6. Sun, F.P.; Chaudhry, Z.; Liang, C.; Rogers, C.A. Truss structure integrity identification using PZT
sensor-actuator. J. Intell. Mater. Syst. Struct. 1995, 6, 134–139. [CrossRef]
7. Sun, F.P.; Liang, C.; Rogers, C.A. Structural modal analysis using collocated piezoelectric actuator/sensors:
An electromechanical approach. Smart Structures and Materials 1994: Smart Structures and Intelligent
Systems. Int. Soc. Opt. Photonic 1994, 2190, 238–250.
8. Zhou, S.W.; Liang, C.; Rogers, C.A. An impedance-based system modeling approach for induced strain
actuator-driven structures. J. Vib. Acoust. 1996, 118, 323–331. [CrossRef]
9. Esteban, J. Analysis of the Sensing Region of a PZT Actuator-Sensor. Ph.D. Thesis, Virginia Polytechnic
Institute and State University, Blacksburg, VA, USA, 1996.
10. Park, G.; Cudney, H.H.; Inman, D.J. An integrated health monitoring technique using structural impedance
sensors. J. Intell. Mater. Syst. Struct. 2000, 11, 448–455. [CrossRef]
11. Xu, Y.G.; Liu, G.R. A modified electro-mechanical impedance model of piezoelectric actuator-sensors for
debonding detection of composite patches. J. Intell. Mater. Syst. Struct. 2002, 13, 389–396. [CrossRef]
12. Zagrai, A.N.; Giurgiutiu, V. Health monitoring of aging aerospace structures using the electromechanical
impedance method. Smart Nondestructive Evaluation for Health Monitoring of Structural and Biological
Systems. Int. Soc. Opt. Photonics 2002, 4702, 289–301.
13. Bhalla, S.; Kumar, P.; Gupta, A.; Datta, T.K. Simplified impedance model for adhesively bonded
piezo-impedance transducers. J. Aerosp. Eng. 2009, 22, 373–382. [CrossRef]
14. Bhalla, S.; Soh, C.K. Electromechanical impedance modeling for adhesively bonded piezo-transducers.
J. Intell. Mater. Syst. Struct. 2004, 15, 955–972. [CrossRef]
15. Annamdas, V.G.M.; Soh, C.K. Three-dimensional electromechanical impedance model. I: Formulation of
directional sum impedance. J. Aerosp. Eng. 2007, 20, 53–62. [CrossRef]
16. Annamdas, V.G.M.; Soh, C.K. Three-dimensional electromechanical impedance model. II: Damage analysis
and PZT characterization. J. Aerosp. Eng. 2007, 20, 63–71. [CrossRef]
17. Annamdas, V.G.M.; Soh, C.K. Three-dimensional electromechanical impedance model for multiple
piezoceramic transducers—Structure interaction. J. Aerosp. Eng. 2008, 21, 35–44. [CrossRef]
18. Peairs, D.M.; Park, G.; Inman, D.J. Improving accessibility of the impedance-based structural health
monitoring method. J. Intell. Mater. Syst. Struct. 2004, 15, 129–139. [CrossRef]
19. Xu, B.; Giurgiutiu, V. A Low-Cost and Field Portable Electromechanical (E/M) Impedance Analyzer for Active
Structural Health Monitoring; South Carolina University: Columbia, CA, USA, 2005.
20. Baptista, F.G.; Vieira Filho, J. A new impedance measurement system for PZT-based structural health
monitoring. IEEE Trans. Instrum. Meas. 2009, 58, 3602–3608. [CrossRef]
21. Bhalla, S.; Gupta, A.; Bansal, S.; Garg, T. Ultra low-cost adaptations of electro-mechanical impedance
technique for structural health monitoring. J. Intell. Mater. Syst. Struct. 2009, 20, 991–999. [CrossRef]
22. Panigrahi, R.; Bhalla, S.; Gupta, A. A low-cost variant of electro-mechanical impedance (EMI) technique for
structural health monitoring. Exp. Tech. 2010, 34, 25–29. [CrossRef]
23. Spencer, B.F.; Ruiz-Sandoval, M.E.; Kurata, N. Smart sensing technology: Opportunities and challenges.
Struct. Control Health Monit. 2004, 11, 349–368. [CrossRef]
24. Lynch, J.P.; Loh, K.J. A summary review of wireless sensors and sensor networks for structural health
monitoring. Shock Vib. Dig. 2006, 38, 91–130. [CrossRef]
25. Mascarenas, D.L.; Todd, M.D.; Park, G.; Farrar, C.R. Development of an impedance-based wireless sensor
node for structural health monitoring. Smart Mater. Struct. 2007, 16, 2137. [CrossRef]
26. Perera, R.; Pérez, A.; García-Diéguez, M.; Zapico-Valle, J.L. Active Wireless System for Struct. Health Monit.
Applications. Sensors 2017, 17, 2880. [CrossRef] [PubMed]
27. Freitas, E.S.; Baptista, F.G. Experimental analysis of the feasibility of low-cost piezoelectric diaphragms in
impedance-based SHM applications. Sens. Actuators A Phys. 2016, 238, 220–228. [CrossRef]
28. Annamdas, V.G.M.; Soh, C.K. Load monitoring using a calibrated piezo diaphragm based impedance strain
sensor and wireless sensor network in real time. Smart Mater. Struct. 2017, 26, 045036. [CrossRef]
29. Budoya, D.; Castro, B.D.; Campeiro, L.; Silveira, R.D.; Freitas, E.D.; Baptista, F. Analysis of Piezoelectric
Diaphragms in Impedance-Based Damage Detection in Large Structures. Proceedings 2018, 2, 131. [CrossRef]
30. Park, S.; Yun, C.B.; Inman, D.J. Structural health monitoring using electro-mechanical impedance sensors.
Fatigue Fract. Eng. Mater. Struct. 2008, 31, 714–724. [CrossRef]
Sensors 2018, 18, 1307 15 of 18

31. Min, J.; Park, S.; Yun, C.B.; Song, B. Development of a low-cost multifunctional wireless impedance sensor
node. Smart Struct. Syst. 2010, 6, 689–709. [CrossRef]
32. Ursu, I.; Giurgiutiu, V.; Toader, A. Towards spacecraft applications of structural health monitoring.
INCAS Bull. 2012, 4, 111.
33. Rabelo, D.S.; Tsuruta, K.M.; de Oliveira, D.D.; Cavalini, A.A.; Neto, R.F.; Steffen, V. Fault Detection of
a Rotating Shaft by Using the Electromechanical Impedance Method and a Temperature Compensation
Approach. J. Nondestruct. Eval. 2017, 36, 25. [CrossRef]
34. Faria, C.T.; Owen, R.B.; Inman, D.J. Macro-fiber composites performance under thermal cycling for
impedance-based SHM applications. Act. Passive Smart Struct. Integr. Syst. 2014, 9057, 905726.
35. Tian, Z. Guided Wave Based Integrated Integrated Structural Health Monitoring and Nondestructive
Evaluation. Ph.D. Thesis, University of South Carolina, Columbia, SC, USA, 2015.
36. Wilkie, W.K.; Bryant, R.G.; High, J.W.; Fox, R.L.; Hellbaum, R.F.; Jalink, A.; Little, B.D.; Mirick, P.H. Low-Cost
piezocomposite actuator for structural control applications. Smart structures and materials 2000: Industrial
and commercial applications of smart structures technologies. Int. Soc. Opt. Photonics 2000, 3991, 323–335.
37. Giurgiutiu, V. Structural Health Monitoring: With Piezoelectric Wafer Active Sensors; Elsevier: New York, NY,
USA, 2007.
38. Tseng, K.K.; Naidu, A.S.K. Non-parametric damage detection and characterization using smart piezoceramic
material. Smart Mater. Struct. 2002, 11, 317. [CrossRef]
39. Tawie, R.; Lee, H.K. Monitoring the strength development in concrete by EMI sensing technique.
Constr. Build. Mater. 2010, 24, 1746–1753. [CrossRef]
40. Xu, B.; Jiang, F. Concrete-steel composite girder bolt loosening monitoring using electromechanical
impedance measurements. In Earth and Space 2012: Engineering, Science, Construction, and Operations in
Challenging Environments; American Society of Civil Engineers: Reston, VA, USA, 2012; pp. 629–634.
41. Hu, X.; Zhu, H.; Wang, D. A study of concrete slab damage detection based on the electromechanical
impedance method. Sensors 2014, 14, 19897–19909. [CrossRef] [PubMed]
42. Wandowski, T.; Opoka, S.; Malinowski, P.; Ostachowicz, W. The performance of three Electromechanical
Impedance damage indicators on structural element with bolted joints. In Proceedings of the International
Symposium on NDT in Aerospace, Bremen, Germany, 16–18 November 2015.
43. Raju, V. Implementing Impedance-Based Health Monitoring. Ph.D. Thesis, Virginia Tech., Blacksburg, VA,
USA, 1997.
44. Ai, D.; Zhu, H.; Luo, H.; Yang, J. An effective electromechanical impedance technique for steel structural
health monitoring. Constr. Build. Mater. 2014, 73, 97–104. [CrossRef]
45. Tinoco, H.A.; Marulanda, D.J. A new index for damage identification in active beams with electromechanical
impedance technique approach to SHM. In Proceedings of the 2nd International Conference on Advanced
Mechatronics, Design, and Manufacturing Technology, Bogota, Colombia, 22–24 October 2014; pp. 1–6.
46. Annamdas, V.G.M.; Yang, Y. Practical implementation of piezo-impedance sensors in monitoring of
excavation support structures. Struct. Control Health Monit. 2012, 19, 231–245. [CrossRef]
47. Tabrizi, A.; Rizzo, P.; Ochs, M.W. Electromechanical impedance method to assess dental implant stability.
Smart Mater. Struct. 2012, 21, 115022. [CrossRef]
48. Bhalla, S.; Suresh, R. Condition monitoring of bones using piezo-transducers. Meccanica 2013, 48, 2233–2244.
[CrossRef]
49. Lim, Y.Y.; Soh, C.K. Fatigue life estimation of a 1D aluminum beam under mode-I loading using the
electromechanical impedance technique. Smart Mater. Struct. 2011, 20, 125001. [CrossRef]
50. Karayannis, C.G.; Voutetaki, M.E.; Chalioris, C.E.; Providakis, C.P.; Angeli, G.M. Detection of flexural damage
stages for RC beams using piezoelectric sensors (PZT). Smart Struct. Syst. 2015, 15, 997–1018. [CrossRef]
51. Zuo, C.; Feng, X.; Zhang, Y.; Lu, L.; Zhou, J. Crack detection in pipelines using multiple electromechanical
impedance sensors. Smart Mater. Struct. 2017, 26, 104004. [CrossRef]
52. Liu, P.; Wang, W.; Chen, Y.; Feng, X.; Miao, L. Concrete damage diagnosis using electromechanical impedance
technique. Constr. Build. Mater. 2017, 136, 450–455. [CrossRef]
53. Soh, C.K.; Lim, Y.Y. Fatigue damage diagnosis and prognosis using electromechanical impedance technique.
Struct. Health Monit. Aerosp. Struct. 2016, 429–446. [CrossRef]
Sensors 2018, 18, 1307 16 of 18

54. Annamdas, V.G.M.; Pang, J.H.L.; Chew, Y.X.; Hoh, H.J.; Zhou, K.; Song, B. Fatigue Monitoring of double
surface defects using PZT based Electromechanical Impedance and Digital image correlation methods.
Adv. Mater. Res. 2014, 891, 551–556. [CrossRef]
55. Dixit, A.; Bhalla, S. Prognosis of fatigue and impact induced damage in concrete using embedded
piezo-transducers. Sens. Actuators A Phys. 2018, 274, 116–131. [CrossRef]
56. Ai, D.; Luo, H.; Wang, C.; Zhu, H. Monitoring of the load-induced RC beam structural tension/compression
stress and damage using piezoelectric transducers. Eng. Struct. 2018, 154, 38–51. [CrossRef]
57. Na, W.S. Distinguishing crack damage from debonding damage of glass fiber reinforced polymer plate
using a piezoelectric transducer based nondestructive testing method. Compos. Struct. 2017, 159, 517–527.
[CrossRef]
58. Boukabache, H.; Escriba, C.; Zedek, S.; Medale, D.; Rolet, S.; Fourniols, J.Y. System-on-Chip integration of a
new electromechanical impedance calculation method for aircraft structure health monitoring. Sensors 2012,
12, 13617–13635. [CrossRef] [PubMed]
59. Wandowski, T.; Malinowski, P.; Radzienski, M.; Opoka, S.; Ostachowicz, W. Methods for Assessment of
Composite Aerospace Structures. In Smart Structures and Materials; Springer: Cham, Switzerland, 2017;
pp. 227–244.
60. Na, W.S. Low cost technique for detecting adhesive debonding damage of glass epoxy composite plate using
an impedance based non-destructive testing method. Compos. Struct. 2018, 189, 99–106. [CrossRef]
61. Sun, R.; Sevillano, E.; Perera, R. Identification of intermediate debonding damage in FRP-strengthened
RC beams based on a multi-objective updating approach and PZT sensors. Compos. Part B Eng. 2017, 109,
248–258. [CrossRef]
62. Wandowski, T.; Malinowski, P.H.; Ostachowicz, W.M. Delamination detection in CFRP panels using EMI
method with temperature compensation. Compos. Struct. 2016, 151, 99–107. [CrossRef]
63. Roth, W.; Giurgiutiu, V. Adhesive disbond detection using piezoelectric wafer active sensors. Structural
Health Monitoring and Inspection of Advanced Materials, Aerospace, and Civil. Int. Soc. Opt. Photonics
2015, 9437, 94370S.
64. Zhuang, Y.; Kopsaftopoulos, F.; Dugnani, R.; Chang, F.K. Integrity monitoring of adhesively bonded joints
via an electromechanical impedance-based approach. Struct. Health Monit. 2017. [CrossRef]
65. Fiborek, P.; Malinowski, P.H.; Kudela, P.; Wandowski, T.; Ostachowicz, W.M. Time-domain spectral element
method for modelling of the electromechanical impedance of disbonded composites. J. Intell. Mater.
Syst. Struct. 2018. [CrossRef]
66. Kim, J.T.; Park, J.H.; Hong, D.S.; Cho, H.M.; Na, W.B.; Yi, J.H. Vibration and impedance monitoring for
prestress-loss prediction in PSC girder bridges. Smart Struct. Syst. 2009, 5, 81–94. [CrossRef]
67. Lu, X.; Lim, Y.Y.; Soh, C.K. A novel electromechanical impedance–based model for strength development
monitoring of cementitious materials. Struct. Health Monit. 2017. [CrossRef]
68. Talakokula, V.; Bhalla, S.; Gupta, A. Monitoring early hydration of reinforced concrete structures using
structural parameters identified by piezo sensors via electromechanical impedance technique. Mech. Syst.
Signal Process. 2018, 99, 129–141. [CrossRef]
69. Jothi Saravanan, T.; Balamonica, K.; Bharathi Priya, C.; Gopalakrishnan, N.; Murthy, S.G.N. Piezoelectric
EMI–Based Monitoring of Early Strength Gain in Concrete and Damage Detection in Structural Components.
J. Infrastruct. Syst. 2017, 23, 04017029. [CrossRef]
70. Kim, J.; Lee, C.; Park, S. Artificial neural network-based early-age concrete strength monitoring using
dynamic response signals. Sensors 2017, 17, 1319. [CrossRef] [PubMed]
71. Kim, W.C.; Jo, H.J.; Park, G. Concrete cure monitoring using piezoelectric admittance measurements.
In Proceedings of the 2015 World Congress on Advances in Structural Engineering and Mechanic, Incheon,
Korea, 25–29 August 2015.
72. Lim, Y.Y.; Kwong, K.Z.; Liew, W.Y.H.; Soh, C.K. Non-destructive concrete strength evaluation using smart
piezoelectric transducer—A comparative study. Smart Mater. Struct. 2016, 25, 085021. [CrossRef]
73. Lim, Y.Y.; Soh, C.K. Electro-mechanical impedance (EMI)-based incipient crack monitoring and critical crack
identification of beam structures. Res. Nondestruct. Eval. 2014, 25, 82–98. [CrossRef]
74. Negi, P.; Chakraborty, T.; Kaur, N.; Bhalla, S. Investigations on effectiveness of embedded PZT patches at
varying orientations for monitoring concrete hydration using EMI technique. Constr. Build. Mater. 2018, 169,
489–498. [CrossRef]
Sensors 2018, 18, 1307 17 of 18

75. Choi, S.K.; Tareen, N.; Kim, J.; Park, S.; Park, I. Real-Time Strength Monitoring for Concrete Structures Using
EMI Technique Incorporating with Fuzzy Logic. Appl. Sci. 2018, 8, 75. [CrossRef]
76. Oliveira, M.A.; Vieira Filho, J.; Lopes Jr, V. A novel time-domain technique for damage detection applied to
SHM using Savitzky-Golay filter. In Proceedings of the 9th International Workshop on Structural Health
Monitoring (IWSHM), Stanford, CA, USA, 10–12 September 2013; pp. 996–1003.
77. Conceição Junior, P.O.; Ferreira, F.I.; Aguiar, P.R.; Batista, F.G.; Bianchi, E.C.; Daddona, D.M. Time-domain
Analysis Based on the Electromechanical Impedance Method for Monitoring of the Dressing Operation.
Procedia CIRP 2018, 67, 319–324. [CrossRef]
78. Vieira Filho, J.; Baptista, F.G.; Inman, D.J. Time-domain analysis of piezoelectric impedance-based structural
health monitoring using multilevel wavelet decomposition. Mech. Syst. Signal Process. 2011, 25, 1550–1558.
[CrossRef]
79. Yang, Y.; Divsholi, B.S. Sub-frequency interval approach in electromechanical impedance technique for
concrete structure health monitoring. Sensors 2010, 10, 11644–11661. [CrossRef] [PubMed]
80. Min, J.; Park, S.; Yun, C.B. Impedance-based structural health monitoring using neural networks for
autonomous frequency range selection. Smart Mater. Struct. 2010, 19, 125011. [CrossRef]
81. Na, S.; Lee, H.K. Resonant frequency range utilized electro-mechanical impedance method for damage
detection performance enhancement on composite structures. Compos. Struct. 2012, 94, 2383–2389. [CrossRef]
82. Na, S.; Lee, H.K. Steel wire electromechanical impedance method using a piezoelectric material for composite
structures with complex surfaces. Compos. Struct. 2013, 98, 79–84. [CrossRef]
83. Giurgiutiu, V.; Kropas-Hughes, C.V. Comparative study of neural network damage detection from a statistical
set of electro-mechanical impedance spectra. Smart Nondestruct. Eval. Health Monit. Struct. Biol. Syst. II Int.
Soc. Opt. Photonics 2003, 5047, 108–120.
84. Min, J.; Park, S.; Yun, C.B.; Lee, C.G.; Lee, C. Impedance-based structural health monitoring incorporating
neural network technique for identification of damage type and severity. Eng. Struct. 2012, 39, 210–220.
[CrossRef]
85. Na, S.; Lee, H.K. Neural network approach for damaged area location prediction of a composite plate using
electromechanical impedance technique. Compos. Sci. Technol. 2013, 88, 62–68. [CrossRef]
86. Palomino, L.V.; Steffen, V.; Finzi Neto, R.M. Probabilistic neural network and fuzzy cluster analysis methods
applied to impedance-based SHM for damage classification. Shock Vib. 2014. [CrossRef]
87. Oh, T.K.; Kim, J.; Lee, C.; Park, S. Nondestructive concrete strength estimation based on electro-mechanical
impedance with artificial neural network. J. Adv. Concr. Technol. 2017, 15, 94–102. [CrossRef]
88. De Oliveira, M.A.; Araujo, N.V.; da Silva, R.N.; da Silva, T.I.; Epaarachchi, J. Use of Savitzky–Golay Filter
for Performances Improvement of SHM Systems Based on Neural Networks and Distributed PZT Sensors.
Sensors 2018, 18, 152. [CrossRef] [PubMed]
89. Park, G.; Kabeya, K.; Cudney, H.H.; Inman, D.J. Impedance-based structural health monitoring for
temperature varying applications. JSME Int. J. Ser. A Solid Mech. Mater. Eng. 1999, 42, 249–258. [CrossRef]
90. Grisso, B.L.; Inman, D.J. Temperature corrected sensor diagnostics for impedance-based SHM. J. Sound Vib.
2010, 329, 2323–2336. [CrossRef]
91. Sepehry, N.; Shamshirsaz, M.; Abdollahi, F. Temperature variation effect compensation in impedance-based
structural health monitoring using neural networks. J. Intell. Mater. Syst. Struct. 2011, 22, 1975–1982.
[CrossRef]
92. Wandowski, T.; Malinowski, P.H.; Ostachowicz, W.M. Temperature and damage influence on
electromechanical impedance method used for carbon fibre–reinforced polymer panels. J. Intell. Mater.
Syst. Struct. 2017, 28, 782–798. [CrossRef]
93. Na, W.S.; Lee, H. Experimental investigation for an isolation technique on conducting the electromechanical
impedance method in high-temperature pipeline facilities. J. Sound Vib. 2016, 383, 210–220. [CrossRef]
94. Lalande, F. Modelling of the Induced Strain Actuation of Shell Structures. Ph.D. Thesis, Virginia Tech.,
Blacksburg, VA, USA, 1995.
95. Tseng, K.K.; Wang, L. Smart piezoelectric transducers for in situ health monitoring of concrete.
Smart Mater. Struct. 2004, 13, 1017. [CrossRef]
96. Yang, Y.; Lim, Y.Y.; Soh, C.K. Practical issues related to the application of the electromechanical impedance
technique in the structural health monitoring of civil structures: II. Numerical verification. Smart Mater. Struct.
2008, 17, 035009. [CrossRef]
Sensors 2018, 18, 1307 18 of 18

97. Hamzeloo, S.R.; Shamshirsaz, M.; Rezaei, S.M. Damage detection on hollow cylinders by electro-mechanical
impedance method: Experiments and finite element modeling. C. R. Méc. 2012, 340, 668–677. [CrossRef]
98. Rugina, C.; Toader, A.; Giurgiutiu, V.; Ursu, I. The electromechanical impedance method for structural health
monitoring of thin circular plates. Proc. Rom. Acad. Seri. A Math. Phys. Tech. Sci. Inf. Sci. 2014, 15, 272–282.
99. Djemana, M.; Hrairi, M.; Al Jeroudi, Y. Using Electromechanical Impedance and Extreme Learning Machine
to Detect and Locate Damage in Structures. J. Nondestruct. Eval. 2017, 36, 39. [CrossRef]
100. Giurgiutiu, V.; Zagrai, A.; Bao, J. Damage identification in aging aircraft structures with piezoelectric wafer
active sensors. J. Intell. Mater. Syst. Struct. 2004, 15, 673–687. [CrossRef]
101. Wait, J.R.; Park, G.; Sohn, H.; Farrar, C.R. Plate damage identification using wave propagation and impedance
methods. Proc. SPIE 2004, 5394, 53–65.
102. Monnier, T.; Jayet, Y.; Guy, P.; Baboux, J.C. Aging and damage assessment of composite structures using
embedded piezoelectric sensors. In Review of Progress in Quantitative NDE; Plenum Press: New York, NY,
USA, 2000; p. 19.
103. An, Y.K.; Sohn, H. Integrated impedance and guided wave based damage detection. Mech. Syst. Signal Process.
2012, 28, 50–62. [CrossRef]
104. Gulizzi, V.; Rizzo, P.; Milazzo, A.; Ribolla, E.L.M. An integrated structural health monitoring system based
on electromechanical impedance and guided ultrasonic waves. J. Civ. Struct. Health Monit. 2015, 5, 337–352.
[CrossRef]
105. Nasrollahi, A.; Ma, Z.; Rizzo, P. Nondestructive assessment of waveguides using an integrated
electromechanical impedance and ultrasonic waves approach. Health Monit. Struct. Biol. Syst. 2017,
10170, 101701C.
106. Ellenberg, A.; Branco, L.; Krick, A.; Bartoli, I.; Kontsos, A. Use of unmanned aerial vehicle for quantitative
infrastructure evaluation. J. Infrastruct. Syst. 2014, 21, 04014054. [CrossRef]
107. Ham, Y.; Han, K.K.; Lin, J.J.; Golparvar-Fard, M. Visual monitoring of civil infrastructure systems via
camera-equipped Unmanned Aerial Vehicles (UAVs): A review of related works. Vis. Eng. 2016, 4, 1.
[CrossRef]
108. Flammini, F.; Pragliola, C.; Smarra, G. Railway infrastructure monitoring by drones. In Proceedings of the
International Conference on Electrical Systems for Aircraft, Railway, Ship Propulsion and Road Vehicles
& International Transportation Electrification Conference (ESARS-ITEC), Toulouse, France, 2–4 November
2016; pp. 1–6.
109. Reagan, D.; Sabato, A.; Niezrecki, C. Unmanned aerial vehicle acquisition of three-dimensional digital image
correlation measurements for structural health monitoring of bridges. Nondestructive Characterization and
Monitoring of Advanced Materials, Aerospace, and Civil Infrastructure. Int. Soc. Opt. Photonics 2017, 10169.
[CrossRef]
110. Na, W.S.; Baek, J. Impedance-based non-destructive testing method combined with unmanned aerial vehicle
for structural health monitoring of civil infrastructures. Appl. Sci. 2016, 7, 15. [CrossRef]
111. Space X. Hyperloop Alpha. Available online: http://www.spacex.com/sites/spacex/files/hyperloop_
alpha-20130812.pdf (accessed on 24 April 2018).

© 2018 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/).