Wearable Flexible Sensors: A Review
Item Type Article
Authors Nag, Anindya; Mukhopadhyay, Subhas Chandra; Kosel, Jürgen
Citation Nag A, Mukhopadhyay SC, Kosel J (2017) Wearable Flexible
Sensors: A Review. IEEE Sensors Journal: 1–1. Available: http://
dx.doi.org/10.1109/jsen.2017.2705700.
Eprint version Post-print
DOI 10.1109/jsen.2017.2705700
Publisher Institute of Electrical and Electronics Engineers (IEEE)
Journal IEEE Sensors Journal
Rights (c) 2017 IEEE. Personal use of this material is permitted.
Permission from IEEE must be obtained for all other users,
including reprinting/ republishing this material for advertising or
promotional purposes, creating new collective works for resale
or redistribution to servers or lists, or reuse of any copyrighted
components of this work in other works.
Download date 10/08/2022 00:07:57
Link to Item http://hdl.handle.net/10754/623772
1530-437X (c) 2016 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSEN.2017.2705700, IEEE
Sensors Journal
Wearable Flexible Sensors: A Review
Anindya Nag
1*
, Subhas Chandra Mukhopadhyay
1
and Jürgen Kosel
2
Abstract— The paper provides a review on some of the
significant research work done on wearable flexible sensors
(WFS). Sensors fabricated with flexible materials have been
attached to a person along with the embedded system to
monitor a parameter and transfer the significant data to the
monitoring unit for further analyses. The use of wearable
sensors has played a quite important role to monitor
physiological parameters of a person to minimize any
malfunctioning happening in the body. The paper categorizes
the work according to the materials used for designing the
system, the network protocols and different types of activities
that were being monitored. The challenges faced by the current
sensing systems and future opportunities for the wearable
flexible sensors regarding its market values are also briefly
explained in the paper.
Keywords- Wearable flexible sensors, physiological
parameter, wireless sensor network, artificial skins, strain
sensors.
I. INTRODUCTION
The advent of sensors in the application world has
revolutionized the quality of human life. Earlier what it took
hours to study or monitor an event can be addressed in
minutes or seconds with the help of sensing systems. The
dynamic use of sensors has led to the ever growing
modification of the existing sensors. They have been used
for different sectors like gas sensing [1, 2], environmental
monitoring [3, 4], monitoring constituents in food products
like meat [5], beverages [6, 7], etc. to name a few. But
monitoring of physiological parameters is one of the most
important applications of sensors as it helps to develop a
model regarding human behavior. Each attribute can be
studied individually to understand the anomalies faced by a
patient and can be counteracted on.
Sensors can be broadly classified into two categories,
flexible [8] and non-flexible [9]. The former one is
fabricated of materials which are malleable to a certain
extent without changing its properties, whereas the later one
is rigid and made of brittle materials. The non-flexible
sensors have been developed earlier among which the
sensors with silicon substrates are the most common ones.
Even though these sensors find a vast field of applications,
there are certain disadvantages like stiffness, intransigency,
etc.
Anindya Nag and Prof. Subhas Chandra Mukhopadhyay are with Faculty
of Science and Engineering, Macquarie University, Sydney, Australia.
(Email: anindya1991@gmail.com, subhas.mukhopadhyay@mq.edu.au)
Prof. Jürgen Kosel is with Computer Electrical and Mathematical Sciences
and Engineering Division, King Abdullah University of Science and
Technology, Saudi Arabia (Email: jurgen.kosel@kaust.edu.sa).
These disadvantages are prominent especially when the
sensing system is associated with monitoring physiological
parameters of a person or any application which involves
prominent stress on the sensor, thus damaging the sensor.
These results in choosing an alternate approach where the
sensor can be dynamically used thus negate any
inconvenience for the person or protecting the sensor from
damaging while using it on a bendable object. Apart from
this, low fabrication cost, light weight, better mechanical
and thermal properties are some of the advantages which
make the use of flexible sensors a better approach.
Wearable sensors have revolutionized the way the activities
of a person are being monitored [10]. They provide the
information accurately and efficiently regarding the
behavior and actions of a person. In today’s world, wearable
sensors are used in many sectors like medical, security,
communication, etc. Figure 1 shows a schematic of a
monitoring system to sensing the physiological parameters
like heart rate and respiratory rate of a person and transmit
the data wirelessly to the cloud via any information gateway
[11]. This is a quick and efficient system because any
abnormality in the transmitted data can generate a
notification to the healthcare or family members.
Fig. 1: Schematic representation of the use of wireless wearable sensors for
physiological parameter monitoring [11].
The paper has been divided into seven sub-sections.
Followed by the introduction given in section I, the
materials used to fabrication wearable flexible sensors are
briefly given in section II. Then, some of the standard
classes of sensing types covered by wearable flexible
sensors are described in section III. Then the sensor
networks and the types of activities being monitored are
given in section IV and V respectively. Finally, the
challenges faced by the current systems and future
opportunities of wearable flexible systems are given in
1530-437X (c) 2016 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSEN.2017.2705700, IEEE
Sensors Journal
section VI. Section VII provides the conclusion for the
paper.
II. MATERIALS FOR WEARABLE FLEXIBLE SENSORS
The material used for fabrication of sensors is decided from
some factors like the application of the sensor, its
availability, total cost of manufacturing, etc. Organic
electronics is one prime sector in the material side which
has been substantially cultivated for the manufacture of
flexible wearable devices [12]. Some of the prospects in the
used of organic devices for flexible wearable devices is
shown in figure 2. These types of sensors have been used in
the manufacturing of thin film transistors, ionic pumps,
polymer electrodes, etc. Organic and large area electronics
(OLAE) [13] is a process to develop electronic devices
printed in thin layers using functional inks. The substrates
used for these operations are main PET and PEN due to
their transparency and lower cost compared to other organic
polymers. OLAE process is currently used to develop
wearable health and medical devices. Use of PDMS [14,
15], PEN [16], PI [17], P(VDF-TrFE) [18], Parylene [19]
and Polypyrrole [20] have been commonly done to develop
flexible sensors [21] for different applications. The
electrode part of the sensor has been developed from
different conducting materials like carbon-based
nanomaterials and metallic nanoparticles. The carbon
compounds include graphene [22-24], carbon nanotubes
(CNTs) [25, 26], carbon fibers [27], etc. Among the metallic
nanoparticles, silver [28, 29], gold [30, 31] and nickel [32]
are some of the most commonly used ones in flexible
wearable sensors.
Fig. 2: Pictorial representation of the different prospects of wearable
flexible devices using organic electronics [12].
There are different kinds of techniques with which the
flexible sensors are developed. The dimensions of the final
products dictate the procedure used to make the sensor
prototype. Photolithography [33], screen-printing [34],
inkjet printing [35], laser cutting [36] are some of the
common ones. The raw materials used in developing these
sensors depend on the applications for which the properties
of the material vary. Polydimethylsiloxane (PDMS) [37],
Polyethylene terephthalate (PET) [38], Polyethylene
naphthalate (PEN) [39], Polyimide (PI) [40] are some of the
insulating substrates commonly used to develop flexible
sensors. The difference in these polymeric materials lies in
their Young’s modulus, refractive index, etc. There are
some conductive polymers like poly(3,4-
ethylenedioxythiophene) polystyrene sulfonate (PEDOT:
PSS), Polyacetylene, polyaniline are some of the examples
of conducting polymers which conduct electricity due to
their lower band gap compared to their insulating
counterparts. These polymers are mainly used in developing
solar cells, batteries; liquid crystal displays (LCDs), etc.
Carbon nanotubes [41], silver [42], gold [43] and copper
nanoparticles [44], are some of the materials used for
fabricating the electrodes in flexible sensors. Among CNTs,
different sensing devices were developed with Single-
Walled Carbon Nanotubes (SWCNTs) and Multi- Walled
Carbon Nanotubes (MWCNTs). These two types have been
used accordingly in different based on their respective
applications.
III. TYPES OF SENSING USING WEARABLE FLEXIBLE SENSORS
The wearable flexible sensors have been employed to
various kinds of sensing in everyday life. These
implementations vary with the structure and properties of
the sensors. Some of the common types of sensing
performed with the flexible sensors have been described in
this section.
Electrochemical sensing [45] is one of the most common
types of flexible sensing that has been performed over the
years. The flexible sensors, with their unique chemical and
electronic properties have been an excellent choice to carry
out different types of biochemical sensing. Some of the
common types of electrochemical sensing include
monitoring of glucose [46-48], pH [49-52], cholesterol [53,
54], etc. The glucose and pH sensors have been developed
from CNTs [55] due to their curvature sidewalls and
hydrophobic nature which provides a strong interaction
through π-bonding. Some of the sensors [56] have used a
layer-by-layer (LBL) structure to give it a more sturdy
structure. Two kinds of polymers, PDDA and PET, were
used to develop the substrate. The SWCNTs, being used as
electrodes, were functionalized with –COOH group to
increase the oxidative nature of the electrodes. Along with
glucose sensing, these sensors provided high sensitivity
towards monitoring of pH between the pH values of 5 to 9.
Figures 3(a) and 3(b) represent the shows the flexibility and
dimension of the sensor respectively.
Other type of electrochemical sensing represent the
monitoring of cholesterol, which is a lipid formed in the cell
membranes of animals. These types of sensors have been
manufactured with both SWCNTs and MWCNTs integrated
with sol-gels [57]. LBL method has also been employed
with the structuring of these sensors to integrate assemble
different materials in a compact way [58]. So, these types of
sensors have been developed with techniques like screen-
printing [59], spin-coating [60], where a separate membrane
1530-437X (c) 2016 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSEN.2017.2705700, IEEE
Sensors Journal
of enzymes like cholesterol esterase [59], cholesterol
oxidase [61] had been immobilized on the sensing surface.
Fig. 3: Glucose/pH sensors developed from PET/PDDA and CNTs. (a)
Sensors subjected to bending to show their flexibility. (b) Comparison of
the size of an individual sensor to a coin to represent its dimension [56].
Pressure [62, 63] and strain [64, 65] sensors are one of the
most standardized applications of flexible sensors. Different
kinds of piezoresistive and piezoelectric sensors have been
developed till date to monitor various physiological
parameters by using them as bandages, gloves, etc. [66].
Figure 4 shows one such type formed from vertically
aligned SWCNTs and PDMS as electrodes and substrate
respectively. These types of sensors vary regarding gauge
factor (GF) and % of the tensile and compressive strain they
can sustain without reaching the breaking point. Some of the
pressure sensors [67] had been manufactured as electronic
bandages where the electrodes were developed by an
agglomeration of two nanoparticles. The usage of more than
of conductive material allowed the sensor to be used in
different mediums.
Fig. 4: Flexible and stretchable strain sensors used for physiological
parameter monitoring. The sensors are fixed on a (a) bandage (b) glove,
and (c) knee to determine the movement of the respective organs in terms
of the change in electrical resistance [66].
These pressure sensors are also used for tactile sensing [68,
69] and artificial intelligence [21, 70]. Some of the strain
sensors [71] developed and tested in the laboratory had
provided a change in conductivity up to a strain of 300%
having a GF of 50. These sensors were based on a
nanocomposite of polyurethane (TPU) and MWCNTs with
nano-fibrillated cellulose (NFC) as fillers.
Biomedical signal monitoring is another sector which has
been worked up with wearable flexible electronic devices
[72]. Monitoring of metabolites on the skin was done by
sensors with ion-electron potentiometric transducers
developed from SWCNTs [73]. Oppositely charged multi-
layered films of MWCNTs were used to establish chemo-
resistive sensors [74]. The detection of sodium (Na
+
) and
potassium (K
+
) ions was detected using a sensor designed
with Cu/PI flexible electronic layer attached to an antenna
for wireless transmission of data to an Android smartphone
[75]. Monitoring of saliva for bacterial infection on tooth
enamel had been done using graphene nanosensors. These
sensors were connected to inductive coil antenna patterned
with interdigital electrodes [76]. Flexible Organic
electrochemical transistors (OECTs) are another type of
sensors used for testing of saliva by converting biochemical
signals to electrical signals. They are developed with a
PANI/Nafion – graphene bilayer film [77]. These transistors
were also developed by the lamination of polypropylene
films and amorphous silicon thin-film transistors on plasma-
enhanced PI substrates. These sensors were used as pressure
sensors and in large area sensor skins [78].
Magnetic field sensors [79] are one category developed
using inorganic functional nano-membranes with polymeric
foils. A linear array of 8 sensors was formed to work on the
principle of Hall Effect to achieve high bulk sensitivity. A
wearable electronic nose [80] was also developed with a
sensor array prepared from a nanocomposite of CNTs and
PEN. Hydrogel systems along with electrophysiological
sensors [81] were prepared with a spin coated and a
thermally cured layer of PI on top of a layer of Poly (methyl
methacrylate) (PMMA). The electrodes were formed with a
bilayer of electron beam evaporated Cr and Au. These
fabricated devices were applied for ECG, stress-strain
measurements along with other biomedical devices [82].
Interestingly, even alloys were used in WFS to develop
biometric sensors [83]. Thin film thermocouples like Sb
2
Te
3
and Bi
2
Te
3
along with Kapton substrate were used to
fabricate a low power, flexible micro-thermoelectric
generator. The device is proposed to be used in Ambient
Assistant Living (AAL) applications.
IV. SENSOR NETWORKS FOR WEARABLE FLEXIBLE SENSORS
Real-time applications of the monitoring of different
physiological parameters are significantly dependent on the
sensor network used to monitor and transfer the recorded
data. After processing the received data in the analog and
digital division of the signal conditioning circuit, the data is
transferred from the sensor node to the monitoring unit via
router for further analysis. A schematic diagram for the
transmission of data from the sensor to the monitoring is
shown in figure 5. The selection of a particular
1530-437X (c) 2016 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSEN.2017.2705700, IEEE
Sensors Journal
communication network depends on the cost of set-up,
power consumption, the number of sensor nodes, the range
of trans-reception, etc. Table 1 shows the comparison of
some network protocols standardized by IEEE [84]. Among
them, Bluetooth has been the most reasonable one due to its
cheaper installation cost, less hardware, and high
compatibility. That’s why; substantial research work has
been done on developing Bluetooth integrated health care
systems [85-87]. Apart from the mentioned protocols in
Table 1, there are some other networks with which data
transmission for different biomedical flexible systems takes
place. SHIMMER uses a Chipcon radio transceiver and 2.4
GHz Rufa™ antenna [88]. Apart from this, there are other
network remote technologies like Sun SPOT, IRIS,
Mica2/MicaZ, Telos [89].
Fig. 5: Schematic diagram of the transmission of data from the sensor to
the monitoring unit.
Table 1: Network protocols standardized by IEEE [84].
Standard
ZigBee
(IEEE
802.15.4)
Bluetooth
(IEEE
802.15.1
WPAN)
Wi-Fi
(IEEE
802.11
WLAN)
Wi-Max
(IEEE
802.11
WWAN)
Range
(m)
100
10
5000
15000
Data rate
(kbps)
250-500
1000-
3000
1000-45000
75000
Band-
width
(GHz)
2.4
2.4
2.4,3.7 and
5
2.3, 3.5
and 3.5
Network
Topology
Star,
Mesh, and
Cluster
trees
Star
Star, Tree,
P2P
Star, Tree,
and P2P
Applicati
ons
Wireless
Sensors
(Monitori
ng and
Control)
Wireless
Sensors
(Monitori
ng and
Control)
PC based
Data
acquisition,
Mobile
Internet
Mobile
Internet
Among these, Telos was developed by UC, Barkley which
used an IEEE 802.15.4 complaint radio claiming to use one-
tenth of power compared to previous mote platforms [90].
Radio frequency (RF) is another network protocol which is
used by different flexible acoustic resonators for data
transmission [91]. For example, ECG monitoring systems
have used Tmote Sky platform which has an 802.15.4 radio
interface at 250 Kbps [92]. Wireless physiological
management system (WPMS) was introduced [93] which
defines carrying the real-time physiological measurement
data wirelessly from the medical sensors to the processing
unit. The probable applications for this technique are in drug
delivery systems like chemotherapy, diabetic insulin
therapy, AIDS therapy [94]. The schematic diagram of the
hardware architecture of the wireless sensor node for
WPMS is shown in figure 6 [93].
Fig. 6: Schematic diagram of the hardware architecture for the sensor node
for WPMS [93].
Another network protocol called Wearable Based Sensor
Networks (WBSNs), based on IEEE 802.15.4 was
introduced that had different probable applications like the
ECG-based system, a wearable platform for light, audio,
motion and temperature sensing [95]. Toumaz
Technologies, UK devised a wireless system-on-chip
integrated system where the transceiver operates between
862-870 MHz and 902-928 MHz ISM bands in European
and North American countries respectively [96]. Research
projects with antennas and RF systems integrated into
clothes have also been progressed working on Body Area
Network (BAN) where the low powered devices would be
surface mounted on the clothing in a fixed position [97].
BAN is categorized into three categories: off-body, on-
body, and in-body [93, 98]. Battery operated systems was
another option that was considered where the developed
system would be powered by a battery integrated into the
system [99, 100]. The advantage of using self-powered
systems [101-103] is that the battery or the power unit of the
wireless system does not have to be replaced every time the
charging-discharging cycle gets over.
![Fig. 1: Schematic representation of the use of wireless wearable sensors for physiological parameter monitoring [11].](/figures/fig-1-schematic-representation-of-the-use-of-wireless-jgslcaxr.webp)
![Fig. 11: Flexible Printed Circuit Board developed for in situ perspiration analysis [160].](/figures/fig-11-flexible-printed-circuit-board-developed-for-in-situ-1o5dl3wq.webp)
![Fig. 10: (a) Schematic diagram showing the pH sensitive chip along with the LED and photodiode. (b) Place of the sensor on a person [139].](/figures/fig-10-a-schematic-diagram-showing-the-ph-sensitive-chip-hlibu47k.webp)
![Fig. 2: Pictorial representation of the different prospects of wearable flexible devices using organic electronics [12].](/figures/fig-2-pictorial-representation-of-the-different-prospects-of-2uu55vs4.webp)
![Fig. 12: An overview of different applications of carbon-based Lithium-ion Batteries (LIBs) [167].](/figures/fig-12-an-overview-of-different-applications-of-carbon-based-3cnn13w3.webp)
![Fig. 13: Schematic diagram for the POC diagnostics using wearable flexible sensors based on different substrates [178].](/figures/fig-13-schematic-diagram-for-the-poc-diagnostics-using-3flqr4si.webp)