Instrumentation & Measurement Magazine 24-6 - 41

detectors are the most frequently used for detecting both the
presence and the movement of people inside the rooms because
of their good sensibility and low-cost.
Behavioral identification can be improved also by measuring
data coming from home openings, such as doors and
windows. Either mechanical or magnetic switches are commonly
employed in smart homes to detect the opening and
closure of doors and windows. Acquired data can, of course,
be employed for optimizing energy consumption. As an example,
smart homes often feature intelligent systems that are
able to automatically disable the heating system in that rooms
where some window has been opened.
Nowadays, another typology of sensors is spreading out
quickly. Several domestic appliances embed complex digital
systems that feature sensors. Such systems are employed first
for enhancing functionalities, performance and security of the
device. As an example, several cooking hobs and ovens have
fire and temperature sensors. Many of such smart appliances
also provide internet or wireless connectivity (Bluetooth)
which allows the users to remotely control them. Smart homes
can take advantage of these integrated sensors to retrieve additional
information of the occupants' behavior. As an example,
it is possible determine if a person is cooking by monitoring
hob sensors.
Other sensors can be, instead, embedded in the infrastructures
of the building. As an example, switch sensors and plug
sensors can be attached to the electric grid to detect when and
where the occupants are switching on/off lights or attaching
loads to some plug. It is also common to employ smart lamps
which include a Wi-Fi interface for controlling them and detecting
their status. Other sensors can be connected to the
plumbing system so that it is possible to determine if the person
is using water. All of this information can greatly improve
the reliability of the behavior classifiers to provide both activity-related
and positional data. Moreover, proximity or tactile
sensors can be easily attached on several items and tools. This
way, it is possible to understand if anyone is using or moving a
specified object inside the home.
A completely different category of sensors is represented
by the wearable sensors. Such sensing devices are employed
in several applications, including sport and fitness, medical
rehabilitation and healthcare. The development of these
sensors is extremely difficult because several critical constraints
exist in such applications. First, these sensors should
be comfortable to wear, and they should not provide any impairment
to subject movements. Subsequently, they have to
be very small and lightweight. Moreover, they must have a
wireless capability because wires attached to the body for
long periods are surely uncomfortable. Severe issues are related
to the safety and power supply of such devices which
usually employ small rechargeable batteries or energy harvesting
techniques.
Nevertheless, wearable sensors are quickly spreading,
and several technologies have been specifically developed
for the healthcare field [20]. As an example, inertial sensors,
force, pressure and strain gauge sensors can be successfully
September 2021
employed to determine the posture and the physical activity
of a person. Such sensors can also be employed for acquiring
bio-physical activities and parameters. As an example, ballistocardiogram
is a non-invasive technique which allows one to
detect breath and heart rate [21] by employing special piezoelectric
flexible films in contact with the body by monitoring
the small movements and deformations due to breathing and
the heart beating. These sensors are therefore able to provide
very useful data about the health and the physical activity of
the subject.
Self-powered sensors [12] are also very important in this
kind of application. They have a very advantageous characteristic:
they do not require a power supply or a battery to
properly operate. Therefore, they are perfect to be embedded
in clothes or any object of common use. Radio-Frequency Identification
tags (RFID) belong to this category. In the simplest
form, they are passive resonant circuits which are enabled
when they enter into proximity of a suitable tag reading system.
Such sensors can be successfully employed for object
proximity or for person identification. As an example, each
person living in the smart home can easily wear an RFID. Several
readers, deployed in different key points of the home, are
able to identify each person during his movement around the
home. This can be a valid solution for behavior classification
because it allows the classifier to properly associate data incoming
from the sensors with the specific person.
Cognitive and Well-being Assessment
Systems
Taking advantage of the already established technologies
employed in smart homes, it is possible to develop several
applications for the cognitive and well-being assessment of
people who, either due to old age or impaired capabilities, require
a certain level of assistance [15], [22]-[24].
Interest in these systems has been constantly increasing in
the last years, because of the several advantages they can provide
both to the assisted people and to the caregivers. First,
such technologies allow mildly-impaired people to maintain
an independent life without the needing to be continuously or
partially assisted by anyone. Moreover, these systems also permit
relatives and healthcare personal to remotely monitor the
health condition of the patients and undertake the required actions
only when actually required. As a consequence, there is a
reduction in overall costs for the healthcare system and an improvement
in the safety and well-being of patients, who can be
engaged in their daily activities without requiring external assistance.
Most cognitive and well-being assessment systems
monitor the daily behaviors of the occupants to detect possible
anomalies. As an example, if a person, who typically awakens
at a certain time and takes breakfast, suddenly changes
his behavior, it may indicate a health problem that needs to be
investigated.
Apparently, there are only minimal differences between
the technologies employed in smart homes and ones employed
for healthcare. Both of them need sensors deployed
in the environment to acquire sufficient information so that
IEEE Instrumentation & Measurement Magazine
41
Instrumentation & Measurement Magazine 24-6

Table of Contents for the Digital Edition of Instrumentation & Measurement Magazine 24-6
