Recent progress on artificial intelligence-enhanced multimodal sensors integrated devices and systems

Journal of Semiconductors, Journal Year: 2025, Volume and Issue: 46(1), P. 011610 - 011610

Published: Jan. 1, 2025

Abstract Multimodal sensor fusion can make full use of the advantages various sensors, up for shortcomings a single sensor, achieve information verification or security through redundancy, and improve reliability safety system. Artificial intelligence (AI), referring to simulation human in machines that are programmed think learn like humans, represents pivotal frontier modern scientific research. With continuous development promotion AI technology Sensor 4.0 age, multimodal is becoming more intelligent automated, expected go further future. this context, review article takes comprehensive look at recent progress on AI-enhanced sensors their integrated devices systems. Based concept principle technologies algorithms, theoretical underpinnings, technological breakthroughs, pragmatic applications fields such as robotics, healthcare, environmental monitoring highlighted. Through comparative study dual/tri-modal with without using (especially machine learning deep learning), highlight potential performance, data processing, decision-making capabilities. Furthermore, analyzes challenges opportunities afforded by offers prospective outlook forthcoming advancements.

Language: Английский

Achieving a High-Output Direct-Current Droplet Triboelectric Generator via Synergistic Effects of a Dual Switch and Electric Double Layer

Nano Letters, Journal Year: 2025, Volume and Issue: unknown

Published: March 6, 2025

Droplet triboelectric generators (D-TENGs) have garnered significant attention for harvesting raindrop energy but face challenges such as low output performance and alternating current (AC) output. This study proposes a high-performance direct (DC) D-TENG with dual-switch (DS) structure (DS-DC-D-TENG) that synergizes effects electric double layers (EDL) to generate DC pulses. Remarkably, using 0.1 mM NaCl droplets, the DS-DC-D-TENG achieves record-breaking short-circuit of 75 μA polymer-based DC-D-TENGs. The physical mechanism is elucidated through an equivalent circuit model finite element method (FEM) simulation. Unlike conventional designs, it directly charges capacitors without rectifier, powers integrated systems temperature humidity sensing display, can be used self-powered droplet counter measure number frequency, showcasing its application potential. work provides novel insights into design future applications

Language: Английский

A Signal Amplitude-Insensitive Triboelectric Touch Panel with a Significantly Reduced Signal Channel and Deep-Learning-Enhanced Robustness

ACS Applied Materials & Interfaces, Journal Year: 2024, Volume and Issue: 16(42), P. 57843 - 57850

Published: Oct. 9, 2024

The self-powered triboelectric touch panel has garnered considerable research attention due to its potential reduce system energy consumption and applications in human–machine interfaces, e-skin, the Internet of Things. Current methods for achieving triboelectric-based positioning an M × N detection pixel array typically require signal amplitude comparison across at least + channels, thereby limiting lightweight design possibilities. In contrast, our novel "resistor ladder" approach necessitates only 4 channels positioning. This method leverages a lookup table correlating positions with ratios from different rendering it insensitive significantly enhancing robustness. We fabricated transparent using PET tribomaterial, where surface roughness was enhanced through plasma treatment. successfully demonstrated 128 taps within sliding predefined table. To further enhance device robustness, 2D convolutional neural network implemented, which achieved impressive accuracy 97.7% even under artificially introduced defects. study represents initial exploration amplitude-insensitive methods, reducing number required robustness panels.

Language: Английский