Page 38 - Read Online

Zhu et al. Soft Sci 2024;4:17 https://dx.doi.org/10.20517/ss.2024.05 Page 25 of 38

Niu et al. built an intelligent material sensing system by building a multilayer perceptron (MLP) neural
network and making it learn the collected capacitive response data of the all-fabric bionic (AFB) e-skin to
[211]
different materials in proximity and pressure modes . The system can accurately discriminate nine
materials with fuzzy morphology and smooth surfaces by the differences in dielectric constants and
hardness of the materials, with an average accuracy of 96.6%.

[212]
Kim et al. presented a novel e-skin system integrated with a deep neural network . It has a fine laser-
induced crack structure to detect small deformations, captures dynamic motion without creating a sensor
network, and, in combination with a deep neural network, a single skin sensor can discriminate the motion
of the corresponding body part.

Tan et al. built a bioinspired spiking multisensory neural network (MSeNN) that integrates vision, touch,
[213]
hearing, simulated smell and taste with cross-modal learning via ANNs . Through distributed multi-
sensors and bionic layered architecture design, it not only senses, processes, and memorizes multimodal
data but also fuses multisensory data at both hardware and software levels.

APPLICATIONS OF E-SKINS
Scholars have been working hard to expand the application areas of e-skins. Here, we will briefly introduce
the applications of e-skins, which can be simply divided into two major application fields: (1) wearable
electronics and healthcare and (2) intelligent machinery. The former is the field of most concern. It is worth
mentioning that there are not very clear boundaries between different applications.

Wearable electronics and healthcare
E-skins are invented to mimic human skin, making them naturally suitable for use in the healthcare and
[36]
wearable fields . The application environments often require the corresponding e-skin devices to possess
various qualities such as non-toxicity, harmlessness, high flexibility, strong biocompatibility, chemical
stability, comfort, high adhesion, and reconfigurability. Materials such as gels, soft metals (mainly precious
metals with high chemical stability in physiological environments, such as Ag and Pt), PDMS, PI, and
[5]
cellulose are widely used . Typical application scenarios include detection and recognition of human limb
movements, continuous monitoring of signals such as electrocardiogram (ECG), electromyogram (EMG),
electroencephalogram (EEG), etc., and detection of human body fluids (sweat, water vapor, excretory fluids,
etc.) .
[214]
As shown in [Figure 13A], Gao et al. fabricated a composite hydrogel [poly(vinyl alcohol)/microfiber
composite hydrogel (PVA/MF-CH) and PVA/MF/Glycerol-CH (PVA/MF/Gly-CH)] composed of PVA
hydrogel and PU microfiber network embedded using electrospinning and spin-coating . In addition to
[215]
the common excellent ionic conductivity, freezing resistance, and dehydration resistance, their thickness
(< 5 μm), Young’s modulus (~0.04 MPa), and tensile strength (~6 MPa) are adjustable, so they possess
better mechanical compatibility with human organs and tissues, which is of great significance in complex
application environments, such as wearable and implantable bioelectronic devices, and are suitable for
healthcare applications. Finally, the flexible electrodes prepared with them demonstrated their potential for
long-term monitoring of electromyographic (EMG) bio-signals. Sun et al. proposed a green and sustainable
one-pot synthesis method by in situ photopolymerization of polymerizable deep eutectic solvents (PDES) -
treated cellulose pulp, and the prepared cellulose-based Ion conductors (ICs), PDES/cellulose microfibers
(CMFs), exhibited very high stretchability (3,210% ± 302%), high ionic conductivity (0.09 S·m ), high
-1
toughness (13.17 ± 2.32 MJ·m ), strong self-healing ability, good stability, and compatibility with human
-3
skin . The detection of typical human strain signals and ECG signals using flexible electrodes prepared
[216]