Page 172 - Read Online

Wang et al. Soft Sci. 2025, 5, 28 https://dx.doi.org/10.20517/ss.2025.11 Page 23 of 29

electronics, pressure sensor arrays have shown great application potential in health monitoring, intelligent
interaction and equipment. However, continuous optimization and innovation are needed for their
practical application. Based on the current research, challenges, and solutions discussed, the following three
future development directions are proposed:

(1) Development of high-performance materials and optimization of complex structures

The selection of materials and the design of structures have been the key aspects of concern in the
fabrication of sensor devices. Material selection directly affects the performance of flexible sensor arrays,
including sensitivity, durability, and stability. At the material level, the future development of flexible sensor
arrays can further focus on the design and application of multifunctional composite materials, such as
conductive polymers (e.g., PEDOT: PSS) and two-dimensional materials (e.g., MXene, graphene) or liquid
metals (e.g., Ga-In alloys) composite, to simultaneously improve conductivity, flexibility, and
environmental stability. Besides, “adaptive materials” that respond to environmental changes such as
temperature, humidity, or electrical stimulation can be used to build sensor units with dynamic tuning
capabilities. For example, materials that automatically adjust their mechanical fit or electrical conductivity
under different external conditions are expected to increase the robustness of sensors in complex scenarios.
Meanwhile, self-healing materials are a key development for achieving long-term stability. It can realize
rapid repair after micro-damage and prolong device lifetime. In addition, biodegradability and green
manufacturing properties of materials will be the focus of future development, especially in wearable
medical devices, which require good biocompatibility and sustainability.

In terms of structure, more attention should be paid in the future to the combination of stretchability and
structural programmability. On the one hand, by introducing mechanical configurations such as negative
Poisson’s ratio structure, helical electrode design, island-bridge structure, etc., the working reliability of the
device under large deformation environment can be significantly improved; on the other hand,
programmable morphological designs such as origami/paper-cutting structure and pneumatically driven
structure are utilized to realize the precise fit and functional stability of the flexible device on 3D surfaces.

(2) Intelligent data processing

High-density integrated sensor devices tend to be limited by data post-processing. Therefore, intelligent
data processing methods are also more important in the future development of the device. In many research
works, intelligent data processing has become an indispensable core link to realize high-performance
applications. Since the sensor arrays usually contain tens or even hundreds of pixel units, they generate a
huge amount of high-dimensional, multi-noise signals in practical applications. These signals not only have
significant temporal and spatial redundancy, but are also susceptible to interference from environmental
variables such as temperature, humidity, and mechanical fatigue, thus affecting sensing accuracy and system
stability. One of the future directions is to be deeply integrated with artificial intelligence (AI). By
embedding neural networks directly into the sensor system, real-time feature extraction and signal
classification can be realized at the front end. It can also significantly reduce data transmission and
processing delays. This “sense-computing” strategy is especially critical for real-time feedback scenarios
such as wearable devices and robotic skin. On the other hand, multiple information fusion technology will
become an important way to improve the intelligence level of the sensing system. By fusing multimodal
signals such as pressure, temperature, strain, capacitance, etc., comprehensive recognition of complex
external stimuli can be realized. For example, neural networks and other methods can be used to process
time series data to realize complex gesture recognition, gait analysis and other functions.