Instrumentation & Measurement Magazine 24-5 - 48

frequency of 100 Hz for the whole system. Depending on the
region at which the textile is positioned, it can measure body
temperature, interaction pressures/forces between the user
and the environment as well as identify different activities of
daily life such as gait, sitting and squatting, where there is the
additional possibility of detecting the gait cadence.
The temperature responses indicate the feasibility of the
optical fiber-embedded smart textile to measure such parameters,
which can be performed when the user is not moving.
As another possibility, the temperature response can be decoupled
from the sensors' responses due to the dynamic
movement by the frequency characteristics of each signal,
where the dynamic movements of the user have higher frequency
components than the temperature behavior. Thus, the
cross-sensitivity reduction can be achieved by filtering the sensors'
responses in different windows, using low-pass filters for
temperature and high-pass for dynamic movements. This approach
can also be used to evaluate the interaction pressure
between the user and the environment, where the smart textile
can provide pressure mapping when the textile region is
in contact with environment, e.g., a chair and mattress. As the
gait is a periodic movement, the cadence (defined as the steps
per minute/second) can also be obtained from the frequency
response, where the peak frequency of the sensor response in
frequency domain (obtained through Fast Fourier Transform)
is related to the cadence (in cycles per second). Moreover,
the suitability of measuring angular movements in multiple
planes of the multiplexed system (Fig. 3) can also be used on
the identification of activities (Fig. 6), where the time- and frequency-domain
signals are used in conjunction with machine
learning techniques [10].
For activity detection, the smart textile is positioned on
the lower back of the user, who is asked to walk, sit and squat,
whereas the frequency and time responses of the sensor array
are analyzed by means of the Principal Components Analysis
(PCA). The principal components from PCA are analyzed
in conjunction with a clustering technique such as k-means,
resulting in the identification of the different activities proposed
in [10]. The obtained results show the feasibility of the
intensity variation-based multiplexed system on the remote
assessment of activities and health condition of the users using
a wearable system that is integrated in clothing and does not
inhibit user's natural movements. Such a system can be integrated
with an IoT module for remote health monitoring, and
due to its scalability, a higher number of sensors can be used
for measurements of additional parameters (such as respiration
and heart rates) or for the assessment of multiple regions
in the user.
Conclusions and Discussion
As other quasi-distributed sensing approaches have become
more popular, their widespread implementation surpasses
the use of intensity variation-based sensors due to advantages
such as their multiplexing capabilities and wavelength-encoded
data (immunity to deviations on light source power).
The multiplexing technique for intensity variation-based
48
sensors discussed here addresses the drawbacks of this sensor
approach, where a quasi-distributed sensor system is obtained
with a self-referencing scheme in which the light source power
deviations are compensated. The multiplexed system can be
used in multipoint and multiparameter assessment as well
as the movement analysis in multiple planes. The discussed
applications for health assessment and movement analysis,
including the 3D-printed insole, POF smart carpet and smart
textile, indicate a new avenue for optical fiber-embedded systems,
where low cost, portable and transparent systems can be
used in clothing and accessories of daily life to monitor the user's
conditions.
The advantages of the multiplexing technique for intensity
variation-based sensors are related to the low cost and
portability of the device when compared with other optical
fiber sensing technologies. Comparing with conventional
approaches for movement analysis, the optical fiber sensors
can be readily incorporated in textiles and in flexible
structures, where their higher flexibility does not inhibit the
natural movements of the users, which, in conjunction with
the small dimensions of the device, lead to a higher transparency
between the sensor system and the user. In addition, the
multiplexing capabilities of the presented intensity variation
sensing method enable multiple sensors to be embedded in a
structure (such as the ones presented in the insole and smart
textile approach) with the possibility of increasing the number
of sensors (or measurement regions) without a large increase
on the cables and components for the signal acquisition.
As the multiplexing technique uses the intensity variation
sensor, a drawback of the sensor system is the sensitivity
to light source power deviations, which leads to the necessity
of using a self-referencing technique, where an additional
photodetector is positioned on the end of the fiber (Fig. 1). In
addition, the LED is side-coupled to the fiber, which leads to
the necessity of creating machined regions (lateral sections)
on the fiber to enable the light coupling with the fiber's core.
Nevertheless, future research points towards the use of doped
fibers with increased scattering or fluorescence in predefined
wavelength region to increase the sensor sensitivity, where
such a fluorescence region can also be used as a reference light
source power compensation (as shown in [23]).
The possibility of remote monitoring and ambient-assisted
living are key features on the Healthcare 4.0 systems. Therefore,
the approach herein discussed can have a major impact
on the development of widely available systems for remote
healthcare, where there is also the possibility of embedding
the sensor system in the user's clothes for assessment of multiple
parameters in a distributed configuration. In addition, the
multiplexed intensity variation sensor system can also be integrated
with accessories and components of daily life such as
chairs and mattress to provide non-wearable remote monitoring
as well as ambient-assisted living.
References
[1] " Summary World Report On Disability, " World Health
Organization, pp. 1-24, 2011.
IEEE Instrumentation & Measurement Magazine
August 2021
Instrumentation & Measurement Magazine 24-5

Table of Contents for the Digital Edition of Instrumentation & Measurement Magazine 24-5
