UNIT I INTRODUCTION TO WEARABLE SYSTEMS AND

SENSORS
Wearable Systems- Introduction, Need for Wearable Systems, Drawbacks
of Conventional Systems for Wearable Monitoring, Applications of
Wearable Systems, Types of Wearable Systems, Components of wearable
Systems. Sensors for wearable systems-Inertia movement sensors,
Respiration activity sensor, Inductive plethysmography, Impedance
plethysmography, pneumography, Wearable ground reaction force
sensor.

Chapter 1
Wearable Systems- Introduction
Wearable technology is any kind of electronic device designed to be worn
on the user's body. Such devices can take many different forms, including
jewelry, accessories, medical devices, and clothing or elements of
clothing. The term wearable computing implies processing or
communications capabilities, but in reality, the sophistication among
wearables can vary.

The most sophisticated examples of wearable technology

include artificial intelligence (AI) hearing aids, Google Glass and
Microsoft's HoloLens, and a holographic computer in the form of a virtual
reality (VR) headset. An example of a less complex form of wearable
technology is a disposable skin patch with sensors that transmit patient
data wirelessly to a control device in a healthcare facility.

Need for wearables

Fundamentally, wearables can perform the following basic functions or
unit operations
• Sense
• Process (Analyze)
• Store
• Transmit
• Apply (Utilize)
Of course, the specifics of each function will depend on the application
domain and the
wearer, and all the processing may occur actually on the individual or at
a remote location
(e.g., command and control center for first responders, fans watching the
race, or viewers
enjoying the mountaineer’s view from the Mount Everest base camp).

Figure 2 is a schematic representation of the unit operations associated

with obtaining and
processing situational data using wearables. For example, if dangerous
gases are detected by a wearable on a first responder, the data can be
processed in the wearable and an alert issued. Simultaneously, it may be
transmitted to a remote location for confirmatory testing and the results –
along with any appropriate response (i.e., put on a gas mask) – can be
communicated to the user in real-time to potentially save a life . This same
philosophy can also be used by an avid gamer who might change his
strategy depending on what “weapons” are available to him and how his
opponents are performing. Each of these scenarios requires personalized
mobile information processing, which can transform the sensory data into
information and then to knowledge that will be of value to the individual
responding to the situation. While wearables are being used in many
fields, as discussed, this chapter will focus primarily on wearables in the
healthcare domain. Wearables provide an unobtrusive way to
longitudinally monitor an individual – not just during the day but, over
the individual’s life-time. Such an expansive view of the individual will
be valuable in detecting changes over time and help in early detection of
problems and diseases leading to preemptive care and hence, a better
quality of life. Inferring the potential of wearables in other application
domains should be straightforward and can be accomplished by
instantiating the fundamental principles and concepts presented here.
Attributes of wearables
A sensor is defined as “a device used to detect, locate, or quantify energy
or matter, giving a signal for the detection of a physical or chemical
property to which the device responds”. Not all sensors are necessarily
wearable, but all wearables, as discussed earlier and shown in Figure 2,
must have sensing capabilities. The key attributes required of an ideal
wearable are shown in Figure 4. From a physical standpoint, the wearable
must be lightweight and the form factor should be variable to suit the
wearer. For instance, if the form factor of the wearable to monitor the vital
signs of an infant prone to sudden infant death syndrome prevents the
infant from (physically) lying down properly, it could have significant
negative implications. The same would apply to an avid gamer – if the
form factor interferes with her ability to play “naturally,” the less likely
that she would be to adopt or use the technology. Esthetics also plays a
key role in the acceptance and use of any device or technology. This is
especially important when the device is also seen by others i.e., the
essence of fashion. Therefore, if the wearable on a user is likely to be
visible to others, it should be esthetically pleasing and, optionally, even
make a fashion statement while meeting its functionality. In fact, with
wearables increasingly becoming an integral part of everyday lives, the
sociological facets of the acceptance of wearables open up exciting
avenues for research. Ideally, a wearable should become such an integral
part of the wearer’s clothing or accessories that it becomes a “natural”
extension of the individual and “disappears” for all intents and purposes.
It must have the flexibility to be shape-conformable to suit the desired
end-use; in short, it should behave like the human skin. The wearable must
also have the multifunctional capability and be easily configurable for the
desired end-use application. Wearables with single functionality (e.g.,
measuring just the heart rate) are useful, but in practical applications,
more than one parameter is typically monitored; and, having multiple
wearables – one for each function or data stream – would make the
individual look like a cyborg and deter their use even if the multiple data
streams could be effectively managed. The wearable’s responsiveness is
critical, especially when used for real-time data acquisition and control
(e.g., monitoring a first responder in a smoke-filled scene). Therefore, it
must be “always on.” Finally, it must have sufficient data bandwidth to
enable the degree of interactivity, which is key to its successful use.
Finally, wearables can be classified based on their field of application,
which can range from health and wellness monitoring to position tracking.
“Information processing” is listed as one of the application areas because
many of these traditional functions such as processing e-mail can now be
done on a wearable in the form of a wristwatch. It is important to note that
not all the classes are mutually exclusive. For instance, a wearable can be
multifunctional, active, noninvasive, and be reusable for health
monitoring. The proposed taxonomy serves two key functions: First, it
helps in classifying the currently available wearables so that the
appropriate ones can be selected depending upon the operating
constraints; second, it helps in identifying opportunities for the design and
development of newer wearables with performance attributes for sp ecific
areas that need to be addressed.
Taxonomy for wearables
Figure 5 shows the proposed taxonomy for wearables. To begin with, they
can be classified as a single function or multifunctional. They can also be
classified as invasive or noninvasive. Invasive wearables (sensors) can be
further classified as minimally invasive, those that penetrate the skin
(subcutaneous) to obtain the signals, or as an implantable, such as a
pacemaker. Implantable sensors require a hospital procedure to be put into
place inside the body. Noninvasive wearables may or may not be in
physical contact with the body; the ones not in contact could either be
monitoring the individual or the ambient environment (e.g., a camera for
capturing the scene around the wearer or a gas sensor for detecting
harmful gases in the area). Noninvasive sensors are typically used in
systems for continuous monitoring because their use does not require
extensive intervention from a healthcare professional.
Wearables can also be classified as active or passive depending upon
whether or not they need the power to operate; pulse oximetry sensors fall
into the former, while a temperature probe is an example of a passive
wearable that does not require its power to operate. Yet another view of
wearables is the mode in which the signals are transmitted for processing
wired or wireless. In the former, the signals are transmitted over a physical
data bus to a processor; in the wireless class of wearables, the
communications capability is built into it, which transmits the signals
wirelessly to a monitoring unit.
Components of wearable Systems.
The main components of wearable devices are as follows

1. Control
Wearable-specific microcontrollers are small, so as to be comfortable and
discrete. On the other hand, the distinctive shapes and colors can function
as a decorative element. Several of the boards available are hand-
washable (minus the power source). Read the documentation carefully.

2. Input/Output
In place of pins, these boards have metal eyelets which you can loop
conductive thread through to sew soft circuit connections. Some boards
also have snaps — or eyelets large enough to solder on snaps — for easy
removal.

3. Conductive Textiles
A material containing metals, such as silver or stainless steel, through
which an electrical current can flow is said to be conductive. Wearable
systems can make use of these materials in a variety of ways, such as:
 Thread for making circuits
 Fabric for capacitive touch sensors
 Hook-and-loop for switches

4. Sensors
Sensors gather information about the environment, the user, or both.
Examples of the former include light, temperature, motion (ACC), and
location (GPS). Examples of the latter include heart rate (ECG), brain
waves (EEG), and muscle tension (EMG). A few wearable
microcontrollers have basic sensors onboard. Other manufacturers offer a
range of external sensor modules that connect to the main board.

5. Power
When scoping out a wearable design one of the first things to consider is
the power requirement. Do you just want to illuminate a few LEDs, or do
you want to run a servomotor? Boards with an integrated holder for a
lithium coin battery are nice for low-power projects that need to be self-
contained. However, boards with a standard JST connector (with or
without a circuit to charge LiPo batteries) are more versatile.

6. Actuators
One generic way to describe a wearable system is: In response to X, where
X is the input from a sensor, Y happens. Actuators such as LEDs, buzzers
or speakers, and servomotors are what make things happen.

7. Networking
To communicate with smart devices, the internet, or other wearable
systems, you need wireless connectivity. In addition to Wi-Fi and
Bluetooth, wearable-friendly options include:
 BLE, which has lower power consumption than classic Bluetooth, a
range of 50m, and a data transmission rate up to 1 Mbps
 NFC, a radio frequency field with a range of approximately 20cm and
data transmission rate up to about 400 Kbps

Types of Wearable Systems

1.Smartwatches:

These days, the watches are tech-enabled. They double up as a fitness

tracker, and sleep monitor in addition to being the classic time-keeping
device. Smartwatches provide us with many other features including
enabling us to make & attend phone calls and check messages. Some
watches have the feature of playing FM radio or audio & video files with
a Bluetooth headset. They generally connect to the smartphone via an app
and act as a supporting device. They are often referred to as a ‘Wearable
Computer’ on your wrist because of the bundle of features that can use
through the touchscreen.

2.FitnessTrackers:

Fitness Trackers are among the wearable technology devices wearable on

the wrist. Fitness trackers were primarily launched to perform the function
of pedometer, i.e. counting the number of steps but they have evolved to
become an overall health monitor since then. They perform various
functions including tracking your heartbeat, monitoring your sleep,
calories burned, and other metrics. They share the data to the app on the
smartphone. In toto, they make a perfect health tracker. Some devices are
enabled to regularly share the information on the metrics of the wearer to
their physicians to keep them informed and help early detection of any
issue.

3.SmartJewelry:

Jewelry no more acts like pieces of ornaments on your neck or hand, they
have become smart. Smart Jewelry are those wearables like necklaces,
wrist bands, bracelets, or rings that are tech-enabled to help you track your
steps, track monitor your heartbeat & sleep, and some even notify you of
incomingcalls.

4.GameSimulators:

The rise of VR in gaming has given rise to many wearable devices that
simulate an environment and make the experience more realistic,
engrossing, and adventurous. The devices include VR Headsets (also
called Head-Mounted Displays or HMDs) that create a visual simulation
and bands that come with built-in sensors to detect your movements.
These bands enable you to control your movements through hand
gestures.

5.SmartClothing:

The advancement of technology with IoT has fostered many inventions

including Smart Clothes. Smart clothes are also popularly known as E-
Textile as they come integrated with electronic devices that measure the
health metrics of the wearer. Smart clothes help measure health-related
aspects like heart rate, respiration rate, sleep, the body temperature, and
provide you with that information. Smart clothing also includes smart
shoes that examine your health, steps, fatigue, and collect other metrics to
help you improve health and prevent injury.

6.SmartGlasses:

Ranging from simple smart glasses that are equipped with Bluetooth
wireless music and hands-free calling to the glasses that can live stream
videos to take photos, to advanced smart glasses that are AR-enabled to
give you an immersive experience, these smart glasses are the of eyewear.
Smart glasses can enable the user to read text messages and reply to them
hands-free. Smart glasses by some companies are equipped with features
like internet access and browsing through voice commands.

7.HeartbeatTrackers&BloodPressureMonitors:

There are fitness trackers for a specific use case like monitoring the
heartbeat or regularly measuring the blood pressure. These devices help
track the metrics among the people who suffer from related diseases. The
fitness trackers record and provide the measurements to the wearer
regularly. Some devices are enabled to share the data with the physician.

8.SmartEarbuds:
New to enter the wearable technology market are earbuds. Though
Bluetooth earbuds are existing for a while now, they aren’t considered
among wearable technology because they do not collect and send data.
But some companies are making earbuds smart. Smart earbuds have a
built-in gyroscope, GPS, and compass. The sensors in the earbuds relay
the information to the smartphone, which enables it to know your
direction and movement. Hence, the smart earbuds are equipped to
provide directions in real-time.

9.SmartContactLens:

Smart Contact Lens is among the recent inventions made possible with
IoT. The smart contact lenses currently available in the market are helpful
for medical reasons. It helps monitor eyes for various diseases like
Diabetes, Glaucoma, and cataracts. It helps in the treatment of
farsightedness. Apart from medical reasons, some companies are working
on smart contact lenses that are AR-enabled, work on solar power, and
capture and store images and videos. Smart lenses are among the
implantable devices.
Applications of wearables
Currently other applications within healthcare are being explored, such
as:

 Applications for monitoring of glucose, alcohol, and lactateor blood

oxygen, breath monitoring, heartbeat, heart rate and
its variability, electromyography (EMG), electrocardiogram (ECG)
and electroencephalogram (EEG), body temperature, pressure (e.g. in
shoes), sweat rate or sweat loss, levels of uric acid and ions – e.g. for
preventing fatigue or injuries or for optimizing training patterns,
including via "human-integrated electronics"
 Forecasting changes in mood, stress, and health
 Measuring blood alcohol content
 Measuring athletic performance
 Monitoring how sick the user is
 Detecting early signs of infection
 Long-term monitoring of patients with heart and circulatory problems
that records an electrocardiogram and is self-moistening
 Health Risk Assessment applications, including measures
of frailty and risks of age-dependent diseases
 Automatic documentation of care activities
 Days-long continuous imaging of diverse organs via a
wearable bioadhesive stretchable high-resolution ultrasound imaging
patch or e.g. a wearable continuous heart ultrasound imager (potential
novel diagnostic and monitoring tools)
 Sleep tracking
 Cortisol monitoring for measuring stress
 Measuring relaxation or alertness e.g. to adjust their modulation or to
measure efficacy of modulation techniques

 Epidermal skin technology. According to Science Daily, the

Terasaki Institute for Biomedical Innovation invented wearable
electronic skin for monitoring health. A next-generation of
wearables, this ultra-thin e-skin patch can be attached to the wearer's
chest area along with a small wireless transmitter by using water
spray and can be worn for up to a week. It is sensitive enough to
pick up and record electro signals, such as heartbeats and muscle
movements, which can be sent to healthcare providers via the cloud
so they can monitor the user's vitals remotely. This powerful
wearable is a steppingstone for monitoring chronic illnesses such as
heart failure and diabetes.

 Health monitoring. People use wearable technology to track and

receive notifications for their heart rate and blood pressure, watch
their calorie intake or manage their training regimens. The COVID-
19 pandemic boosted the use of wearable technology, as consumers
gained a broader awareness of personal hygiene and taking
 precautions to prevent the spread of infections. Apple, for instance,
updated its Cardiogram app by introducing a new sleeping beats-
per-minute feature that monitors heart rate fluctuations for COVID-
19 patients.

 Entertainment and gaming. The gaming and entertainment

industries were the first to adopt VR headsets, smart glasses and
controllers. Popular VR head-mounted displays, such as Oculus
Quest, Meta Quest and Sony PlayStation VR, are used for all types
of entertainment purposes, including gaming, watching movies and
virtual traveling.

 Fashion and smart clothing. Clothing known as smart clothing,

or intelligent fashion, has been gaining wide popularity over the past
few years. Smart jackets, such as Levi's jacket made with Google's
Project Jacquard technology whose threads are composed of
electrical fibers, enable the wearer to answer calls, play music or
take photos right from their sleeves. Smartwatches, wristbands,
smart shoes and smart jewelry are also popular examples of
wearable technology.

 Military. These wearables include technology that tracks soldiers'

vitals, VR-based simulation exercises and sustainability technology,
such as boot inserts that estimate how well the soldiers are holding
their equipment weight and how terrain factors can affect their
performance.

 Sports and fitness. Sports use wearable athletic devices that are
either built into the fabric of the sports apparel or are incorporated
into sports equipment, such as bats and balls. The GPS and
Bluetooth-linked devices relay real-time data to coaches for analysis
through connected electronic devices such as laptops. Besides
wearable athletic devices, familiar wearable technology such as
Fitbit, Apple Watch, Garmin, Samsung Galaxy Watch and Polar are
 used extensively to track various areas of the player's health and
performance metrics.

Advantages of Wearable Technology

 Rapid data results can help drive improvements. Having

immediate data to make decisions and drive improvements may
be helpful, rather than waiting for more formal or detailed
assessments.
 Detailed data can supplement loss analysis and loss
trends. Additional data can help identify specific trends in your
claims history.
 Can help build a business case for senior management. It can
be challenging to help senior management make decisions or
determine if some of your funding should be spent on
improvements. The data from wearable technology devices can
help support your business case for that spend.
 Data from wearable sensors offers promising job risk analysis
and evaluation opportunities for safety and ergonomics
practitioners. Most ergonomic assessments or evaluations
require additional time to observe and manually collect data.
Having instant data can save time and expedite ergonomic
assessments or evaluations.
 Enhance employee wellness programs. More organizations are
starting to promote wellness programs for employees. Some
wearable technology devices can assist with easily tracking
wellness program data that could supplement or support your
efforts.

Disadvantages of Wearable Technology

 Requires a time commitment to review and analyze data. A
team or committee may need to review the large amount of data
that is generated from the devices.
 Requires financial commitments and planning. You may need
senior management or finance team approval prior to the full
implementation stage. The cost of wearable technology depends
on how many employees and locations are involved.
 Devices could lead to distraction. For many employees, wearing
this device for an entire shift can be distracting, especially if the
device has haptic feedback or vibration reminders.
 Data security and privacy could be compromised with legal,
financial, and personal consequences. An information
technology (IT) department will need to ensure the data generated
from the devices is secured for authorized individuals and ensure
proper consent is obtained from each individual whose data is
being collected.
 Devices could lead to over-trust or under-trust. This could be
challenging when reviewing all the data to determine realistic
trends. Sometimes this results in trusting or not trusting all the
reviewed data before making any decisions or improvements.
Sensors for Wearable Systems
Introduction
When designing wearable systems to be used for physiological and
biomechanical parameters monitoring, it is important to integrate sensors
easy to use, comfortable to wear, and minimally obtrusive. Wearable
systems include sensors for detecting physiological signs placed on-body
without discomfort, and possibly with capability of real-time and
continuous recording. The system should also be equipped with wireless
communication to transmit signals, although sometimes it is opportune to
extract locally relevant variables, which are transmitted when needed.
Most sensors embedded into wearable systems need to be placed at
specific body locations, e.g. motion sensors used to track the movements
of body segments, often in direct contact with the skin, e.g. physiological
sensors such as pulse meters or oximeters. However, it is reasonable to
embed sensors within pieces of clothing to make the wearable system as
less obtrusive as possible. In general, such systems should also contain
some elementary processing capabilities to perform signal pre-processing
and reduce the amount of data to be transmitted. A key technology for
wearable systems is the possibility of implementing robust, cheap
microsystems enabling the combination of all the above functionalities in
a single device. This technology combines so-called micro-electro-
mechanical systems (MEMS) with advanced electronic packaging
technologies. The former allows complex electronic systems and
mechanical structures (including sensors and even simple motors) to be
jointly manufactured in a single semiconductor chip. A generic wearable
system can be structured as a stack of different layers. The lowest layer is
represented by the body, where the skin is the first interface with the
sensor layer. This latter is comprised of three sub-layers: garment and
sensors, conditioning and filtering of the signals and local processing. The
processing layer collects the different sensor signals, extracts specific
features and classifies the signals to provide high-level outcomes for the
application layer. The application layer can provide the feedback to the
user and/or to the professional, according to the specific applications and
to the user needs. Recent developments embed signal processing in their
systems, e.g. extraction of heart rate, respiration rate and activity level.
Activity classification and more advanced processing on e.g. heart signals
can be achievable exploiting miniaturization and low-power consumption
of the systems. Examples of data classification are[1, 2, 3]: classification
of movement patterns such as sitting, walking or resting by using
accelerometer data [4] or ECG parameters such as ST distance extracted
from raw ECG data [5, 3]; another example is the estimation of the energy
consumption of the body [6, 7]; in [8] the combined use of a triaxial
accelerometer and a wearable heart rate sensor was exploited to accurately
classify human physical activity; estimation of upper limb posture by
means of textile embedded flexible piezo resistive sensors [9]. Examples
of integrated systems for health monitoring are in [10, 11]. In the
following paragraphs, two classes of sensors which can be easily
integrated into wearable systems are reported and described. More
specifically, inertial sensors to monitor biomechanical parameters of
human body and sensors to capture physiological signs are addressed,
describing the operating principles and indicating the possible fields of
application.
Sensors for Wearable Systems
Biomechanical Sensors
Biomechanical sensors are thought to be used to record kinematic
parameters of body segments. Knowledge of body movement and gesture
can be a means to detect movement disturbances related to a specific
pathology or helpful to contextualize physiological information within
specific physical activities. An increasing of heart rate, for example, could
be either due to an altered cardiac behavior or simply because the subject
is running.
Inertial Movement Sensors
Monitoring of parameters related to human movement has a wide range
of applications. In the medical field, motion analysis tools are widely used
both in rehabilitation and in diagnostics. In the multimedia field, motion
tracking is used for the implementation of life like videogame interfaces
and for computer animation. Standard techniques enabling motion
analysis are based on stereo-photogrammetric, magnetic and
electromechanical systems. These devices are very accurate but they
operate in a restricted area and/or they require the application of obtrusive
parts on the subject body. On the other hand, the recent advances in
technology have led to the design and development of new tools in the
field of motion detection which are comfortable for the user, portable and
easily usable in non-structured environments. Current prototypes realized
by these emergent technologies utilize micro-transducers applied to the
subject body (as described in the current paragraph) or textile-based strain
sensors. The first category, instead, includes devices based on inertial
sensors (mainly accelerometers and gyroscopes) that are directly applied
on the body segment to be monitored. These sensors can be realized on a
single chip (MEMS technology) with low cost and outstanding
miniaturization. Accelerometers are widely used for the automatic
discrimination of physical activity and the estimation of body segment
inclination with respect to the absolute vertical. Accelerometers alone are
not indicated for the estimation of the full orientation of body segments.
The body segment orientation can be estimated by using the combination
of different sensors through data fusion techniques (Inertial Measurement
Units, IMU). Usually, tri-axial accelerometers (inclination), tri-axial
gyroscopes (angular velocity), magnetometers (heading angle) and
temperature sensors (thermal drift compensation) are used together. Main
advantages of using accelerometers in motion analysis are the very low
encumbrance and the low cost. Disadvantages are related to the possibility
of obtaining only the inclination information in quasi-static situations (the
effect of the system acceleration is a noise and the double integration of
acceleration to estimate the segment absolute position is unreliable).
Accelerometers are widely used in the field of wearable monitoring
systems, generally used in the monitoring of daily life activities (ADL).
Physical activity detection can be exploited for several fields of
application, e.g. energy expenditure estimation, tremor or functional use
of a body segment, assessment of motor control, load estimation using
inverse dynamics techniques [26, 27] or artificial sensory feedback for
control of electrical neuromuscular stimulation [28, 29, 30]. Usually,
three-axial accelerometers are used. They can be assembled by mounting
three single-axis accelerometers in a box with their sensitive axes in
orthogonal directions or using a sensor based on one mass [31]. An
accelerometer measures the acceleration and the local gravity that it
experiences. Considering a calibrated tri-axial accelerometer (the
accelerometer signal (y) contains two factors: one is due to the gravity
vector (g) and the other depends on the system inertial acceleration (a),
both of them expressed in the accelerometer reference frame :The
inclination vector (z) is defined as the vertical unit vector, expressed in
the accelerometer coordinate frame [4]. In static conditions, only the
factor due to gravity is present and the inclination of the accelerometer
with respect to the vertical is known. In dynamic conditions, the raw
accelerometer signal does not provide a reliable estimation of the
inclination, since the inertial acceleration is added to the gravity factor.
This estimation error grows as the subject movements become faster (e.g.
running, jumping). Many algorithms have been developed and tested to
perform a reliable estimation of the subject body inclination: most of them
use low pass filters with very low cut-off frequency in order to extract
z[4] (i.e. introducing a considerable time delay), others implement more
complex techniques which use a model-based approach mainly based on
Kalman filter techniques. An example of integration of these sensors in a
garment was developed in the frame of the Proetex project (FP6-2004-
IST-4-026987), which aimed at using textile and fibre based integrated
smart wearables for emergency disaster intervention personnel. The
ProeTEX motion sensing platform is used to detect long periods of user
immobility and user falls to the ground and it is realized by means of two
tri-axial accelerometer modules. One accelerometer is placed in the higher
part of the trunk (collar level) in order to detect inactivity and falls to the
ground. The second sensor is placed in the wrist region and its aim is to
achieve more accuracy in inactivity detection, since an operator can move
his arms while his trunk is not moving. The core of the motion sensor is
the processing algorithm described in, which allows to perform a reliable
estimation of the body inclination even in the case of intense physical
activity such as running or jumping. This algorithm allows a good
estimation of subject activities and generated fall alarms with very high
sensitivity and extremely low level of false positives.
Respiration Activity sensor
The most challenging vital sign to accurately record during continuous
monitoring is the respiratory activity due to the fact that the signals are
affected by movement artifacts and filtering or feature recognition
algorithms are not very effective. Monitoring of respiratory activity
involves the collection of data on the amount and the rate at which air
passes into and out of the lungs over a given period of time. In literature,
there are several methods to do this, both directly, by measuring the
amount of air exchanged during the respiration activity, and indirectly, by
measuring parameters physically correlated to breathing, such as changes
in thorax circumference and/or cross section, or trans-thoracic impedance.
Direct methods are based on a spyrometer that measures directly the
airflow in the lung exchanged during inspiration and expiration, but of
course it cannot be integrated into a wearable system because it employs
a mouthpiece, which could interfere with the freedom of movements,
disrupting the normal breathing pattern during measurement, thus causing
discomfort for the user. Indirect methods exploit displacements of the
lung that are transmitted to the thorax wall and vice versa, and therefore
measurements of chest-abdominal surface movements can be used to
estimate lung volume variation. In literature, a number of devices have
been used to measure rib cage and abdominal motion including mercury
in rubber strain gauges, linear differential transducers, magnetometers,
and optical techniques, but almost all cannot be comfortably integrated
into a wearable system. For reference only, it is worthwhile citing a more
sophisticated technique, called stereo photogrammetry, which makes it
possible to estimate the three-dimensional coordinates of points of the
thorax, estimating therefore volume variations. Nevertheless, this system
presents a considerable drawback in that it is cumbersome, extremely
expensive, and can only be used in research environments or in laboratory
applications. Indirect techniques that can be implemented in wearable
systems are respiratory inductive plethysmography, impedance
plethysmography, piezo resistive and/or piezoelectric pneumography.
These systems are minimally invasive and do not interfere with physical
activity.
Inductive Plethysmography
The inductive plethysmography method for breathing monitoring consists
of two elastic conductive wires placed around the thorax and the abdomen
to detect the cross sectional area changes of the rib cage and the abdomen
region during the respiratory cycles. The conductive wires are insulated
and generally sewn in a zig-zag fashion onto each separate cloth band.
They can be considered as a coil and are used to modulate the output
frequency of a sinewave current produced by an electric oscillator circuit.
As a matter of fact, the sinewave current generates a magnetic field, and
the cross-sectional area changes due to the respiratory movements of the
rib cage and of the abdomen determine a variation of the magnetic field
flow through the coils. This change in flow causes a variation of the self-
inductance of each coil that modulates the output frequency of the
sinusoidal oscillator. This relationship allows for monitoring the
respiratory activity by detecting the frequency change in the oscillator
output signal. For accurate volumetric measurements using RIP, it is
assumed that the cross-sectional area within the rib cage and the abdomen
coil, respectively, reflects all of the changes occurring within the
respective lung compartment, and further that the lung volume change is
the sum of the volume changes of the two compartments. Under optimal
situations, lung volume can be approximated with an error less than 10%.
Impedance Plethysmography
This technique consists of injecting a high frequency and low amplitude
current through a pair of electrodes placed on the thorax and measuring
the trans-thoracic electrical impedance changes [40]. As a matter of fact,
there is a relationship between the flow of air through the lungs and the
impedance change of the thorax. The measurements can be carried out by
using either two or four electrode configurations. Electrodes can be made
of fabric and integrated into a garment or,even, embedded into an
undershirt. It is worthwhile noting that by measuring the trans-thoracic
electrical impedance it is possible to non-invasively monitor, in addition
to breathing rate also tidal volume, functional residual capacity, lung
water and cardiac output.
Pneumography Based on Piezoresistive Sensor
Piezoresistive pneumography is carried out by means of piezoresistive
sensors that monitor the cross-sectional variations of the rib cage. The
piezoresistive sensor changes its electrical resistance if stretched or
shortened and is sensitive to the thoracic circumference variations that
occur during respiration. Piezoresistive sensors can be easily realized as
simple elastic wires or by means of an innovative sensorized textile
technology. It consists of a conductive mixture directly spread over the
fabric. The lightness and the adherence of the fabric make the sensorized
garments truly unobtrusive and uncumbersome, and hence comfortable
for the subject wearing them. This mixture does not change the
mechanical properties of the fabric and maintains the wearability of the
garment. Figure 1.3 shows where the two conductive wires or bands could
be applied.
Plethysmography Based on Piezoelectric Sensor
This method is based on a piezoelectric cable or strip which can be simply
fastened around the thorax, thus monitoring the thorax circumference
variations during the respiratory activity. A possible implementation can
be a coaxial cable whose dielectric is a piezoelectric polymer (p(VDF-
TrFE)), which can be easily sewn in a textile belt and placed around the
chest. The sensor is sensitive to the thorax movements and produces a
signal directly proportional to the thorax expansion in terms of charge
variation, which was converted in an output voltage proportional to the
charge by means of a charge amplifier. A suitable local processor can
enable implementation of the Fast FourierTransform in real time and
extraction of the breathing rate.
Wearable GRF Sensor System
The quantitative analysis of gait variability using kinematics and kinetic
characterizations can be helpful to medical doctors in monitoring
patients’ recovery status in clinical applications. Moreover, these
quantitative results may help to strengthen their confidence in the
rehabilitation. Walking speed, stride length, the centre of mass (CoM)
and the centre of pressure (CoP) have been considered as factors in the
evaluation of walking gait. According to one study on slip type falls,
friction force was used to draw up important safety criteria for detecting
safe gait, so the transverse components of ground reaction force (GRF)
may provide important information for quantifying gait variability. Many
kinds of stationary systems such as force plates and instrumented
treadmill devices are available to measure CoP and triaxial GRF. Because
a stationary force plate cannot measure more than one stride, in studies of
continuous walking, a complex system consisting of many force plates
and a data fusion method must be constructed . Therefore, the force plate
technology probably imposes some constraints on our ability to measure
human movement and is not feasible for measurements in everyday
situations. An instrumented treadmill or dynamometric platform formed
by laying two force plates under a treadmill can overcome some
limitations of the system with distributed multiple force plates in
successive measurements of the GRF for gait evaluation. However, a
guide used to constrain the direction of the foot is necessary to ensure
that subjects walk along a straight line, because if a human body segment
motion analysis system is not available for a simultaneous measurement
of the foot orientation, any technique based on force plates conventionally
requires subjects to walk along a pre-defined specific path. Although gait
variability can be assessed in straight walking , gait analysis
concentrating solely on straight-line walking or running may not
adequately interpret gait variability, because turning or walking direction
changes probably have effects on extrinsic gait variability . To overcome
such limitations of stationary devices in GRF measurement, many
researchers are developing wearable sensors attached to shoes . Pressure
sensors have been widely used to measure gaits and the distributed
vertical component of GRF and to analyze the loading pattern on the
plantar soft tissue during the stance phase of gait , but in these systems
the transverse components of GRF (friction forces) which are one of the
main factors leading to falling, have been neglected. By fixing two
externally mounted sensors beneath the front and rear boards of a special
shoe, researchers have developed an instrumented shoe for ambulatory
measurements of CoP and triaxial GRF in successive walking trials, and
the application of the instrumented shoe to estimate joint moments and
powers of the ankle was introduced in. The mounted sensor itself, having
a height of 15.7 mm, increases the height and weight of the shoe, and
affects normal walking gait. Moreover, its application study was
restricted to human kinetics analysis using the spatio temporal
measurements of GRF and CoP.

A wearable GRF sensor system was constructed using five small

triaxial force
sensors (USL06-H5-500N-C, weight: 15 g, size: 20 mm × 20 mm × 5
mm). The GRF and CoP measured using the wearable sensor system were
expressed in a global coordinate system which was located on the
interface between the instrumented shoe and the ground. The origin of
the global coordinate system was fixed to a point around the anatomical
centre of the ankle when the sensor system was worn on the foot. The x-
axis was chosen to represent the anterior-posterior direction on the
interface plane contacting with the floor, which was based on landmarks
from the shoe. The z-axis was made vertical, and the y-axis was chosen
such that the resulting global coordinate system would be right-handed.
By mounting the five triaxial sensors on an aluminum plate beneath the
shoe, we can accurately align all five local coordinate systems defined
for each triaxial sensor with the global coordinates. Fxi, Fyiand Fzi ( I =
1, 2, 3, 4 and 5) indicate triaxial forces measured by the distributed five
triaxial sensors, and (xi, yi) is defined as the position of each triaxial
sensor, for example, (x5, y5) indicates the position of the sensor placed
on the forefoot. The total weight of the sensor shoe is about 300 g, and the
shoe size is 250 mm.