Professional Documents
Culture Documents
Abstract—A substantial amount of growth is being achieved by databases, or the cloud. Based on the analysis performed on
businesses through IoT-based services. The emergent of small elec- the collected data, specific actions are taken according to the
tronic devices capable of computing, which are commonly known as services that are required. There are a variety of IoT sensors,
wearables in IoT domain has proven to have huge impact in people’s
life. Theses wearables are capable of collecting vital information including smart sensors, actuators, and wearable sensors. We will
about a person’s activities and behaviours regularly. This makes discuss sensors in the context of wearable devices in section
them suitable for many applications in health monitoring, fitness, IV. Using multiple communication technologies, heterogeneous
arXiv:2304.09861v1 [cs.HC] 12 Apr 2023
sports, education and some industry related applications. To this IoT objects can be connected to provide specific smart services.
end, in this paper, we aim to provide a general review on IoT-based The different communication technologies used in the context
wearables, the sensors adopted for several categorized wearables, the
communication technologies adopted and the most widely adopted of wearable devices will be discussed in detail in section V. In
data processing techniques for wearables. Furthermore, we present IoT, processors, microcontrollers, System-on-Chips (SoCs), and
the challenges faced for wide adoption of wearables and the future Field-Programmable Gate Arrays (FPGAs) comprise the ”brain”
research directions. and computational power. Arduino, UDOO, FriendlyARM, Intel
Index Terms—Internet of things, wearables, communication tech- Galileo, Raspberry PI, Gadgeteer, BeagleBone, Cubieboard, Z1,
nologies, data analytic WiSense, Mulle, and T-Mote Sky are some of the hardware
platforms designed to run IoT applications.
The emergent of small electronic devices capable of computing,
I. I NTRODUCTION
which are commonly known as wearables in IoT domain has
become worth their weight in gold when they are integrated into
a true IoT system. As a result, most research papers currently
published in the literature connect wearable devices to the Internet
in one of two ways. Either the wearable devices send data to the
cloud or to an Internet server for offline processing or some of
the data processing is done locally on the wearable device. Having
integrated IoT platforms and addressing many issues pertaining to
data ownership, data sharing policies, privacy, and safety will fully
realize the prevalence of wearable devices.
Smart watch Smart jewelry Smart wristband
In many cases, a mobile network is needed to support these
wearable devices. Data is also a concern when dealing with
sensors that collect many readings. In works [14]–[18] the authors
create specialized networks that can handle the needs of wearable
devices. These networks also address several concerns such as
security and data storage. In, [19] shows that edge nodes can also
Smart glasses Smart shoes Smart belts play an important role with IoT devices and data collection. It can
reduce load on the server and save storage space.
TABLE I: Summary of the Categorized Wearables and the relavant related works
Type of Technol-
Range Data Rates Frequency Band (s) Topology
ogy
ZigBee 10 to 100 meters 250kbps 2GHz Star, ad hoc, and mesh
WiFi 10 to 100 meters 6.75Gbps 2.4GHz, 5GHz Point Hub
2.402GHz to Point to point, point to
Bluetooth 10 to 100 meters 2.1Mbps
2.408GHz multi-point
169 MHz (Asia),
868MHz (Europe)
LoRaWan 15 to 20 Kilometers 250bps to 5.5kbps Star
91MHz (North
America)
Using pretested templates for each type of fall and comparing device can help measure and study kinematics in a sport setting.
angular velocities and angles between falls and normal activities, Examples of the sensors being used in a device can be found in
[30] propose a fall detection framework. In [31], deep learning [44] and [45] where the sensors are used for their advantages in
and activity characteristics are used to develop a novel method capturing certain ranges of motions. The data is then captured
for detecting human falls on furniture using scene analysis that and analyzed by their respective data analysis programs. Sensors
adopts Region-based Convolutional Neural Network (R-CNN). in IoT are very violate especially in Wearables, many sensors
A fall detection system incorporated tri-axial accelerometers, have more than one application, for example IMU sensors have
gyroscopes, and magnetometers to obtain fall parameters was applications in wearables outside of sports. For example, in [46]
proposed by [32]. Based on a reverse approach, the authors in [33] IMU sensors are used to create 3D maps, an application that is
propose a novel autonomous fall detection system. The proposed used outside of the scope of motion.
system uses a camera attached to the subject rather than static
sensors at fixed locations. Therefore, monitoring is not limited to
areas where the sensors are located but includes all areas where B. Sensors found in Healthcare
the subject travels. [34] presents a novel system that uses a smart When it comes to healthcare there is a long list of sensors that
camera to detect falls and classify activities around the waist. are found in wearables. Wearables in health care can be used to
An EEG-based BCI prototype is proposed [35] to detect whether monitor or treat patients, which are powered by different types
an on-site worker is sleep-deprived or not in an elegant manner. of sensors. Sensors already discussed [41][42][43] also have use
Modified safety helmets with a discreetly placed signal acquisition in healthcare since they deal with the study of motion with the
device are worn by workers to ensure their safety. Table I presents human body [40].
a summary of the categorized wearables and the relevant related Looking at table III there are many different sensors found
works published in the literature. in different wearable devices for healthcare. Many sensors are
present in many different devices, and sometimes are even com-
IV. S ENSORS IN I OT- BASED W EARABLE bined with other sensors from other categories of wearables. Look-
Sensors are a crucial part of IoT-based wearable, they allow for ing at the airflow sensor, it has a straightforward purpose within
the data collection of a certain action that can then be analyzed healthcare, to help a patient breathe. They can be found mostly
and measured. They are many different sensors, and each single in ventilators[48] as well as devices for oxygen therapy[50]. A
sensor can have multiple different applications. pressure sensor have an unexpectedly large impact on devices,
they are used in ventilators, oxygen therapy, sleep therapy[51], and
automation of drug infusion[52]. Another sensor that is common
A. Sensors found in Sports throughout healthcare is oxygen sensors, which measure a patients
Wearables that are used in sports to gain metrics and about an oxygen level. The sensor can be found in devices that attach to
athlete’s performance are categorized under Inertial Measurement a patient’s finger [54]. EEG sensors are also used in healthcare,
Unit (IMU) sensors and help with the study of motions, kinematics they are used in research for the human brain to help gain a better
[40] The sensors that are categorized under IMUs are magnetome- understanding and can also be used to help diagnose any issues.
ter, accelerometer, and gyroscopes. Magnetometer helps determine Glucose sensors mostly benefit patients with diabetes, the sensor
orientation of an object which in sports has many applications, the has become non-invasive and implemented on IoT devices such as
sensors themselves are small require low energy to power and can smartwatches that can automatically send alerts to let the patient
communicate over I2C or SPI [41]. Accelerometers can measure know any imbalances are occurring. The ECG sensor might be the
movement in a given direction, they can be as precise to measure most know sensor in healthcare, it is present in every hospital and
movement such as wrist tilt [42]. Gyroscopes measure orientation almost all smart watches/ring[59][60][61] feature an ECG sensor.
as well as angular velocity, it also has the ability to maintain The ECG measures heart rate and can collect information about a
orientation as well [43]. Combining the sensors into a single patient throughout the day to see if any irregularities are occurring.
6
C. Sensors found in Safety with everything between loss prevention to navigation. For more
Wearables in the area of safety are also finding applications, local mapping that does not need communication to satellite, like
the scenarios range from at home to work environment safety. somewhere in a house or small community, accelerometers are
Wearable devices are fitted with sensors that can check the envi- used to measure distance between certain points to create a map.
ronment, make sure certain equipment is being handled properly, This comes into play in applications with smart homes, as well
and prevent and kind of injury. Taking a look at [62] the authors as mobile devices to help predict a user’s habits.
used different sensors in a wearable device to determine harmful
environmental conditions. The sensors used in their wearable are
temperature, C02, UV, and CO [63][64][65][66]. When it comes
to safety most sensors can be implemented in wearables, for
sports, health, and tracking and localization they can be added as
preventative measures. For example, in sports, accelerometers can
be used to detected impacts which can cause injuries, in healthcare
ECG sensors can be used to detect abnormal heart rates, and in
tracking sensors can be implemented to prevent someone from
getting lost.
1) ZigBee: ZigBee technology is built on the IEEE 802.15.4 example, for tracking and localization wearable can feature an
standard. Resource-constrained environments and devices with accelerometers to track distance traveled but a wearable from
limited power provide the perfect environment for this kind of healthcare application could be using an accelerometers to track
technology to thrive. In spite of being standardized in 2003, the motion of a patient for physical therapy.
it was originally conceived in 1998 [69]. ZigBee has a range Machine learning can play an important role in data analysis
of approximately ten to a hundred meters and consumes an for wearables. It can help learn a person’s habits and find what is
insignificant amount of power. ZigBee uses the direct sequence normal for the user. In [73] it comes in the form of personalized
spread spectrum (DSSS) technique and adopts the ad hoc, mesh medicine, where a machine learning algorithm can learn about
and star topology. In ZigBee, there are three types of devices: the a patient and then keep checks on the patient for irregularities
coordinator (ZC), the router (ZR), and the end-device (ZED) [70]. outside of their normal. The same type of concept can be used in
2) Wi-Fi: A wireless communication standard based on all applications of wearables. Another example of data analysis
IEEE802.11, wireless fidelity (Wi-Fi) is one of the most popular comes from [74] where a wearable is being used to help diagnose
wireless technologies. In order to access radio channels, Wi-Fi early stages of Parkinson’s disease. The sensors collect data which
uses the CSMA/CA protocol. With a range up to 10 meters and is then analyzed and compared to the early signs of the disease.
a high -power consumption, Wi-Fi is a highly efficient wireless
network. There are four types of configurations available with this
network: infrastructure, ad hoc, bridge, and repeater, which has a
maximum data transfer rate of 6.75Gbps and 140MHz of channel
capacity for a very high price [71].
3) Bluetooth: Wireless personal area networks (WPANs) with
high security can be formed using Bluetooth, a proprietary open
standard developed for mobile and fixed devices. Using a master-
slave (client-server) model, Bluetooth is a packet-oriented protocol
that is connection-oriented, packet-based. Bluetooth were limited
by the initial specifications’ high-power consumption, long com-
munication set-up time, and high transmission latency (about 100
ms). Due to these limitations, Bluetooth Low Energy (BLE) was
developed to overcome them. BLE is designed specifically for
applications relating to health, sports, and fitness applications.
4) LoRaWan: IoT networks have become increasingly inter-
ested in LoRaWan (low-power wide-area networks) because it is
suitable for LPWAN (low-power wide-area networks). LoRaWan
is very suitable for wireless sensor networks due to its robustness
and range. Frequency Shift Keying (FSK) or Chirp Spread Spec-
trum (CSS) are the main modes of communication used in LoRa
radios. A node and a gateway are the two main types of devices
in this technology. Gateways are connected to thousands of nodes
at once, each sending and receiving information from the gateway
in a LoRaWan setup.
In the wearable device domain, Bluetooth Classic (standardized
but no longer maintained by IEEE 802.15.1), Bluetooth Smart and
Wi-Fi have become the most common standards [72] used for
connectivity to the Internet. A comparative analysis of the different
Fig. 8: System used for collecting data and analyzing data from
adopted communication technologies used in IoT-based wearables
a sensor [74]
in terms of range, data rates, frequency bands and topology is
presented in Table II.
In many of the systems, the data analysis doesn’t take place on
the device itself. The computation needed cannot be met, there
B. Data Analytics edge nodes and cloud -based systems are put in place to analyze
Data analytics has a large application within wearables in IoT, the data. However, this doesn’t come without challenges, many
since many new methods of machine learning are being developed, forms of data analytics require a large amount of data also many
they can be used within these applications. In a system there is the sensors are collecting data constantly throughout the entire day.
device with its many sensors taking constant measurements it is Over time this accumulates into terabytes of data that can fill up
the analytics that take the measurements and make them a readable storage drive quickly. Another challenge that must be addressed
output. IoT-based Wearables are classified in to four applications, in privacy, the data that is being collected about a person’s
health applications, sports applications, safety applications, track- habits/lifestyle is very detailed and can be shared between different
ing and localization applications. Wearables from each application companies, making privacy an issue. In [75] an approach is taken
can differ slightly in design and include the same sensors, it is with data analytics to unlabeled the data to help privacy. By doing
how the data is analyzed that differs from each application. For this they are not disclosing where the data is coming from and
8
keeping to anonymous, the only competent that is receiving the C. Trust-related Issues for Medical Use Cases
labelled data is the algorithm doing the analysis. There is lack of trust in the sensitive data produced by consumer
wearables in patients monitoring applications. For example, data
related to heart rate, pulse rate, and other health metrics is sensed
by wearables using consumer hardware. Therefore, physicians
are reluctant to use these data for critical diagnostics because
they much rely on the accuracy of the hardware used in making
the wearables as well as the accuracy of using it properly. A
critical challenge would therefore be to improve health-related
data accuracy cheaply.
D. Privacy-related Issues
It is possible that privacy breaches could occur as a result of
an exchange of personal data between wearables and IoT hubs,
including vital health signals, dosage, and location. It is typical for
wearable IoT devices to operate in broadcast mode so that other
network nodes can discover them easily. As a result, unauthorised
users can intercept data in the form of eavesdropping attacks,
which leads to privacy violation. A number of questions remain
unanswered concerning the effective protection of users’ privacy.
This is an interesting research problem that is on-going not only
in the wearable domain but in other security-related domains.
Fig. 9: System used for privacy data processing [75]
VII. C ONCLUSION
From fitness and sport to health monitoring, wearable devices
VI. C HALLENGES AND F UTURE R ESEARCH D IRECTIONS are becoming increasingly popular. In this paper, we provided a
There are several challenges in the domain of IoT-based wear- comprehensive review of the most important research efforts from
ables. To fully utilize the benefits of wearables devices in the the literature in IoT-based wearables. We categorized the wear-
community, these challenges need to be addressed urgently. ables according to their applicable applications. Additionally, the
sensors, communication technologies and data analytic techniques
A. Battery-related Issues adopted in the IoT-based wearables is investigated and presented
Since batteries have limited working times, wearable devices by surveying multiple papers published in the literature. Addi-
often have a short sustainable working time, which causes in- tionally. the challenges as well as the future research directions
convenience to people’s daily lives. Therefore, when designing in IoT-based wearables is presented. In terms of communication
wearables, special considerations need to be taken into account technologies, Bluetooth Classic, Bluetooth Smart and Wi-Fi are
to minimize human interaction and ensure that the batteries last the most common standards adopted for wearables connectivity
for many hours without needing to be replaced or recharged. For to the Internet.
example, there are a variety of systems available, such as those
that use low power consumption or techniques that harvest energy R EFERENCES
such as thermo-electric and piezo-electric that can be utilized in [1] L. D. Xu, W. He, and S. Li, “Internet of things in industries: A survey,”
wearable domain. IEEE Transactions on Industrial Informatics, vol. 10, no. 4, pp. 2233–2243,
2014.
[2] M. Chui, J. Manyika, and M. Miremadi, “Four fundamentals of workplace
B. Communication-related Issues automation,” McKinsey Quarterly, vol. 29, no. 3, pp. 1–9, 2015.
[3] A. Al-Fuqaha, M. Guizani, M. Mohammadi, M. Aledhari, and M. Ayyash,
Currently, most wearable devices connect to the Internet pri- “Internet of things: A survey on enabling technologies, protocols, and
marily through proxy devices such as smartphones or personal applications,” IEEE Communications Surveys & Tutorials, vol. 17, no. 4,
computers. For example, most fitness tracking wearable applica- pp. 2347–2376, 2015.
[4] N. Niknejad, W. B. Ismail, A. Mardani, H. Liao, and I. Ghani, “A
tions still communicate with clouds through smartphones. This comprehensive overview of smart wearables: The state of the art literature,
shows a gap in the adoption of direct communication with the recent advances, and future challenges,” Engineering Applications of
internet, which hinders the development of some specific delay Artificial Intelligence, vol. 90, p. 103529, 2020. [Online]. Available:
https://www.sciencedirect.com/science/article/pii/S0952197620300348
sensitive applications in the wearable domain. Potential reasons [5] Forecast unit shipments of wrist-worn wearables worldwide from 2019
for this lack of support may be the absence of secure direct to 2024. [Online]. Available: https://healthtechinsider.com/2020/10/26/
communication devices built into wearable operating systems or idc-forecasts-double-digit-weables-growth-through-2024/
[6] D. Nahavandi, R. Alizadehsani, A. Khosravi, and U. R. Acharya,
the slower development of third-party applications for wearables. “Application of artificial intelligence in wearable devices: Opportunities and
However, the demand for wearable devices capable of direct challenges,” Computer Methods and Programs in Biomedicine, vol. 213, p.
communication with the internet is on the rise [76]. This is a 106541, 2022. [Online]. Available: https://www.sciencedirect.com/science/
article/pii/S0169260721006155
potential area of research to explore the design of wearable devices [7] What is the Internet of Things (IoT). [Online]. Available: https:
with such capability. //www.tibco.com/reference-center/what-is-the-internet-of-things-iot
9
[8] J. Pourbemany, Y. Zhu, and R. Bettati, “A survey of wearable devices pairing [31] W. Min, H. Cui, H. Rao, Z. Li, and L. Yao, “Detection of human falls
based on biometric signals,” 2021. on furniture using scene analysis based on deep learning and activity
[9] S. D. Mamdiwar, A. R, Z. Shakruwala, U. Chadha, K. Srinivasan, and characteristics,” IEEE Access, vol. 6, pp. 9324–9335, 2018.
C.-Y. Chang, “Recent advances on iot-assisted wearable sensor systems [32] P. Pierleoni, A. Belli, L. Palma, M. Pellegrini, L. Pernini, and S. Valenti,
for healthcare monitoring,” Biosensors, vol. 11, no. 10, 2021. [Online]. “A high reliability wearable device for elderly fall detection,” IEEE Sensors
Available: https://www.mdpi.com/2079-6374/11/10/372 Journal, vol. 15, no. 8, pp. 4544–4553, 2015.
[10] S. R. Haghi Mostafa, Thurow Kerstin, “Wearable devices in medical [33] K. Ozcan, S. Velipasalar, and P. K. Varshney, “Autonomous fall detection
internet of things: Scientific research and commercially available devices,” with wearable cameras by using relative entropy distance measure,” IEEE
hir, vol. 23, no. 1, pp. 4–15, 2017. [Online]. Available: http://www. Transactions on Human-Machine Systems, vol. 47, no. 1, pp. 31–39, 2016.
e-sciencecentral.org/articles/?scid=1075870 [34] K. Ozcan, A. Mahabalagiri, and S. Velipasalar, “Automatic fall detection and
[11] Y. Xue, “A review on intelligent wearables: Uses and risks,” Human Behavior activity classification by a wearable camera,” in Distributed Embedded Smart
and Emerging Technologies, vol. 1, no. 4, pp. 287–294, 2019. [Online]. Cameras. Springer, 2014, pp. 151–172.
Available: https://onlinelibrary.wiley.com/doi/abs/10.1002/hbe2.173 [35] S. R. Dhole, A. Kashyap, A. N. Dangwal, and R. Mohan, “A novel helmet
[12] R. Sanchez-Iborra, “Lpwan and embedded machine learning as enablers for design and implementation for drowsiness and fall detection of workers on-
the next generation of wearable devices,” Sensors, vol. 21, no. 15, 2021. site using eeg and random-forest classifier,” Procedia Computer Science, vol.
[Online]. Available: https://www.mdpi.com/1424-8220/21/15/5218 151, pp. 947–952, 2019.
[13] J. Passos, S. I. Lopes, F. M. Clemente, P. M. Moreira, M. Rico-González, [36] J.-L. Reyes-Ortiz, L. Oneto, A. Samà, X. Parra, and D. Anguita, “Transition-
P. Bezerra, and L. P. Rodrigues, “Wearables and internet of things (iot) aware human activity recognition using smartphones,” Neurocomputing, vol.
technologies for fitness assessment: A systematic review,” Sensors, vol. 21, 171, pp. 754–767, 2016.
no. 16, 2021. [Online]. Available: https://www.mdpi.com/1424-8220/21/16/ [37] J. Qi, P. Yang, M. Hanneghan, S. Tang, and B. Zhou, “A hybrid hierarchical
5418 framework for gym physical activity recognition and measurement using
[14] Y. Bello, E. Figetakis, A. Refaey, and P. Spachos, “Continuous integration and wearable sensors,” IEEE Internet of Things Journal, vol. 6, no. 2, pp. 1384–
continuous delivery framework for sds,” in 2022 IEEE Canadian Conference 1393, 2018.
on Electrical and Computer Engineering (CCECE), 2022, pp. 406–410. [38] R. C. Shit, S. Sharma, D. Puthal, and A. Y. Zomaya, “Location of things
[15] Y. Bello, A. R. Hussein, M. Ulema, and J. Koilpillai, “On sustained zero (lot): A review and taxonomy of sensors localization in iot infrastructure,”
trust conceptualization security for mobile core networks in 5g and beyond,” IEEE Communications Surveys & Tutorials, vol. 20, no. 3, pp. 2028–2061,
IEEE Transactions on Network and Service Management, vol. 19, no. 2, pp. 2018.
1876–1889, 2022. [39] S. Tomic, M. Beko, and R. Dinis, “Rss-based localization in wireless sensor
[16] Y. Bello, A. A. Abdellatif, M. S. Allahham, A. R. Hussein, A. Erbad, networks using convex relaxation: Noncooperative and cooperative schemes,”
A. Mohamed, and M. Guizani, “B5g: Predictive container auto-scaling for IEEE Transactions on Vehicular Technology, vol. 64, no. 5, pp. 2037–2050,
cellular evolved packet core,” IEEE Access, vol. 9, pp. 158 204–158 214, 2014.
2021. [40] M. Rana and V. Mittal, “Wearable sensors for real-time kinematics analysis
[17] Y. Bello, M. S. Allahham, A. Refaey, A. Erbad, A. Mohamed, and in sports: A review,” IEEE Sensors Journal, vol. 21, no. 2, pp. 1187–1207,
N. Abdennadher, “B5g: Predictive container auto-scaling for cellular evolved 2021.
packet core,” in 2021 IEEE International Conference on Communications [41] Bosch Sensortec, “BMA150-Magnetometer,” https://www.bosch-sensortec.
Workshops (ICC Workshops), 2021, pp. 1–7. com/products/motion-sensors/magnetometers-bmm150/, accessed: 2022-7-
[18] J. Singh, Y. Bello, A. R. Hussein, A. Erbad, and A. Mohamed, “Hierarchical 23.
security paradigm for iot multiaccess edge computing,” IEEE Internet of [42] ——, “BMA400-Accelerometer,” https://www.bosch-sensortec.com/
Things Journal, vol. 8, no. 7, pp. 5794–5805, 2021. products/motion-sensors/accelerometers/bma400/, accessed: 2022-7-23.
[19] E. Figetakis and A. Refaey, “Uav path planning using on-board ultrasound [43] ——, “BMG250-Gyroscope,” https://www.bosch-sensortec.com/products/
transducer arrays and edge support,” in 2021 IEEE International Conference motion-sensors/gyroscopes-bmg250/, accessed: 2022-7-23.
on Communications Workshops (ICC Workshops), 2021, pp. 1–6. [44] Zepp Labs, “Zepp-Golf,” http://www.zepplabs.com/en-us/golf/
[20] IEC TC 124 Strategic Business Plan. [Online]. Available: https://www.iec. zepp-sensor-specifications/, accessed: 2022-7-23.
ch/public/miscfiles/sbp/124.pdf [45] Kinexon, “Kinexon Perform,” https://kinexon.com/solutions/basketball/, ac-
[21] IEC TC 124 Strategic Business Plan. [Online]. Available: https://www.iec. cessed: 2022-7-23.
ch/dyn/www/f?p=103:103:0::::FSP ORG ID,FSP LANG ID:12601,25 [46] C. Li, L. Yu, and S. Fei, “Real-time 3d motion tracking and reconstruction
[22] S. Seneviratne, Y. Hu, T. Nguyen, G. Lan, S. Khalifa, K. Thilakarathna, system using camera and imu sensors,” IEEE Sensors Journal, vol. 19, no. 15,
M. Hassan, and A. Seneviratne, “A survey of wearable devices and chal- pp. 6460–6466, 2019.
lenges,” IEEE Communications Surveys & Tutorials, vol. 19, no. 4, pp. 2573– [47] Honeywell, “AWM3000 series airflow sensors,” https://sps.honeywell.
2620, 2017. com/ca/en/products/advanced-sensing-technologies/healthcare-sensing/
[23] F. J. Dian, R. Vahidnia, and A. Rahmati, “Wearables and the internet of things airflow-sensors/amplified/awm3000-series, accessed: 2022-7-23.
(iot), applications, opportunities, and challenges: A survey,” IEEE Access, [48] Hillrom, “Life 2000 Ventilation,” https://www.hillrom.com/en/products/
vol. 8, pp. 69 200–69 211, 2020. life2000-system/life2000-overview/, accessed: 2022-7-23.
[24] C. Nave and O. Postolache, “Smart walker based iot physical rehabilitation [49] NXP, “MPX-5010,” https://www.nxp.com/products/
system,” in 2018 International Symposium in Sensing and Instrumentation sensors/pressure-sensors/differential-gauge-up-to-10-kpa/
in IoT Era (ISSI). IEEE, 2018, pp. 1–6. 0-to-10kpa-differential-and-gauge-integrated-pressure-sensor:MPXx5010,
[25] G. Yang, J. Deng, G. Pang, H. Zhang, J. Li, B. Deng, Z. Pang, J. Xu, accessed: 2022-7-23.
M. Jiang, P. Liljeberg et al., “An iot-enabled stroke rehabilitation system [50] Inogen, “G3-System,” https://www.inogen.com/products/g3-systems/, ac-
based on smart wearable armband and machine learning,” IEEE journal of cessed: 2022-7-23.
translational engineering in health and medicine, vol. 6, pp. 1–10, 2018. [51] Philips, “DreamStation-2,” https://www.usa.philips.com/healthcare/product/
[26] B. Xu, L. Xu, H. Cai, L. Jiang, Y. Luo, and Y. Gu, “The design of an m- HCEUX520H15C/dreamstation-2-auto-cpap-advanced-auto-cpap, accessed:
health monitoring system based on a cloud computing platform,” Enterprise 2022-7-23.
Information Systems, vol. 11, no. 1, pp. 17–36, 2017. [52] R. Dudde, T. Vering, G. Piechotta, and R. Hintsche, “Computer-aided
[27] B. Xu, L. Da Xu, H. Cai, C. Xie, J. Hu, and F. Bu, “Ubiquitous data accessing continuous drug infusion: setup and test of a mobile closed-loop system
method in iot-based information system for emergency medical services,” for the continuous automated infusion of insulin,” IEEE Transactions on
IEEE Transactions on Industrial informatics, vol. 10, no. 2, pp. 1578–1586, Information Technology in Biomedicine, vol. 10, no. 2, pp. 395–402, 2006.
2014. [53] ProtoCentral, “NMAX30100,” https://protocentral.com/product/
[28] A. Ghorbel, S. Bouguerra, N. B. Amor, and M. Jallouli, “Cloud based pulse-pulse-ox-heart-rate-sensor-based-on-max30102/, accessed: 2022-
mobile application for remote control of intelligent wheelchair,” in 2018 14th 7-23.
International Wireless Communications & Mobile Computing Conference [54] Insignia, “NS-PSOXMW1,” https://www.insigniaproducts.com/pdp/
(IWCMC). IEEE, 2018, pp. 1249–1254. NS-PSOXMW1/6438023, accessed: 2022-7-23.
[29] M. O. Agyeman and A. Al-Mahmood, “Design and implementation of [55] NeuroSky, “MindWave,” https://www.sparkfun.com/products/14758, ac-
a wearable device for motivating patients with upper and/or lower limb cessed: 2022-7-23.
disability via gaming and home rehabilitation,” in 2019 Fourth International [56] Muse, “Muse-2,” https://choosemuse.com/muse-2/, accessed: 2022-7-23.
Conference on Fog and Mobile Edge Computing (FMEC). IEEE, 2019, pp. [57] Dexcom, “G6 CGM,” https://ca.store.dexcom.com/en-CA/
247–252. dexcom-g6/g6-sensor-single/STS-GS-008.html?gclid=
[30] M. Nyan, F. E. Tay, and E. Murugasu, “A wearable system for pre-impact fall EAIaIQobChMI967Ig6qd-QIVZYJbCh0NKASBEAQYASABEgJvRPD
detection,” Journal of biomechanics, vol. 41, no. 16, pp. 3475–3481, 2008. BwE, accessed: 2022-7-23.
10