Page 111 - Read Online

Page 20 of 34 Xi et al. Soft Sci 2023;3:26 https://dx.doi.org/10.20517/ss.2023.13

applications or as an external electrical stimulation device.

Other self-powered sensors
In addition to PENGs and TENGs, there are some self-powered technologies that can also be used in
etc.
wearable sensors, such as thermoelectric material sensors, solar material sensors,

Thermoelectric materials play an important role in self-powered wearable sensors. The thermoelectric
material is a material that can directly convert thermal energy into electrical energy. When one end of a
thermoelectric sensor is stimulated by heat, the thermoelectric effect generates a voltage, enabling the
[42]
conversion of thermal energy to electrical energy . Therefore, the application of thermoelectric materials
in self-powered wearable sensors can realize the self-power of the sensor and avoid the failure of the sensor
due to the exhaustion of the battery. At the same time, thermoelectric sensors can also convert the heat
energy generated by the body itself into electrical energy, thereby realizing the recycling of human energy.
He et al. prepared a new self-supporting self-powered temperature and strain sensor using a simple droplet
casting method . The composite membrane can maintain stable thermoelectric performance after 1,000
[133]
washes and can withstand repeated bending and stretching. It can successfully detect temperature changes
and strain deformation under self-powered conditions. Wang et al. have developed a self-powered wearable
ultraviolet index detector, which is achieved by using a small flexible thermoelectric generator as the power
[26]
source . Wearable ultraviolet detectors can successfully self-power by collecting human heat.
Solar materials can directly convert the energy in sunlight into electrical energy, so the application of solar
[200]
materials in self-powered wearable sensors can realize the self-power of the sensors . The solar sensor can
drive the sensor by absorbing light energy and converting it into electrical energy in different indoor or
outdoor environments so that the sensor has better adaptability and usability. Solar energy materials also
have the characteristics of environmental protection and energy saving, realizing environmental protection
and energy recycling, which is of great significance in a society that is increasingly concerned about
environmental protection and sustainable development. Choi et al. proposed a new wearable self-powered
pressure sensor based on the integration of piezoelectric transmission microporous elastomer (PTME) and
thin film organic solar cell (OSC) . PTME shows that in response to the applied pressure, the micropores
[201]
gradually close with compression, resulting in a change in transmittance. The unique optical properties of
PTME enable OSC to respond to changes in current caused by pressure, which can be used for detecting the
bending/stretching of human fingers, among other applications.

Machine learning
Machine learning is becoming increasingly popular in self-powered IoT sensors as it enables intelligent
decision-making without the need for continuous manual input. In self-powered IoT sensors, machine
learning algorithms can be used to analyze data from various sensors and make predictions or decisions
based on this data. A key benefit of using machine learning in self-powered IoT sensors is that it can make
them more autonomous and self-sufficient. By using machine learning algorithms to analyze data and make
decisions, these sensors can reduce the need for manual intervention, which is particularly important in
applications where access to manual maintenance or resources may be limited. Tan et al. developed a
[202]
machine learning-based device . The device can realize gesture recognition with a full keyboard and
multi-command input. It is based on machine learning algorithms, with a maximum accuracy of 92.6%
when recognizing 26 letters. It can be seen that machine learning algorithms have transformed the complex
signals collected by sensors into meaningful and practical results. Another advantage of using machine
learning in self-powered IoT sensors is that it can help improve their accuracy and reliability. By analyzing a
large amount of data and identifying patterns, machine learning algorithms can make highly accurate