A Deep Learning-Enabled Skin-Inspired Pressure Sensor for Complicated Recognition Tasks with Ultralong Life

Area of Science:

• Materials Science
• Wearable Technology
• Biomimetic Engineering

Background:

• Flexible full-textile pressure sensors are crucial for direct integration into clothing.
• Challenges remain in achieving high sensitivity, wide detection range, and long working life.
• Complex tasks require intricate sensor arrays prone to damage and extensive data processing.

Purpose of the Study:

• To develop a simple, skin-inspired full-textile pressure sensor.
• To achieve high performance metrics including sensitivity, detection range, and durability.
• To enable complex task recognition using a single, integrated sensor.

Main Methods:

• Fabrication of a full-textile pressure sensor using a dip-and-dry approach.
• Incorporation of signal transmission, protective, and sensing layers.
• Development of an artificial Internet of Things (IoT) system for sensor application.

Main Results:

• Achieved high sensitivity (2.16 kPa-1) and an ultrawide detection range (0-155.485 kPa).
• Demonstrated impressive mechanical stability over 1 million loading/unloading cycles.
• Successfully performed complex tasks like handwriting digit and human activity recognition with high accuracy using a single sensor.

Conclusions:

• The skin-inspired full-textile sensor offers a promising route for electronic textiles.
• The developed sensor exhibits excellent performance and durability.
• Potential applications include human-machine interaction and advanced human activity detection.