Li Xianghui, Zhang Jiayi, Lin Zihan, Li Junhao, Wang Xinchen, Hong Weiqiang, Li Zhaobin, Zhou Ziyuan, He Xiaoying, Zhang Rui, Hao Jianhong, Shao Yupeng, Deng Feifei, Zhao Yunong, Guo Xiaohui
Key Laboratory of Intelligent Computing and Signal Processing of Ministry of Education, School of Integrated Circuits, Anhui University, Hefei 230601, China.
ACS Appl Mater Interfaces. 2025 Aug 28. doi: 10.1021/acsami.5c15896.
In advanced robotics and human-machine interfaces, there is a critical demand for flexible sensors that can bridge the gap between noncontact perception and physical interaction. Integrating noncontact magnetic sensing for proximity detection with contact-based pressure sensing for tactile feedback in a single device is a key approach to meeting this demand. However, achieving high performance in both modalities is challenging due to a fundamental trade-off: materials and structures optimized for high pressure sensitivity are often compromised by the integration of magnetic components required for field detection, and vice versa. To address the above issues, this paper proposes a rabbit leg-inspired flexible pressure-magnetic sensor (RF-PMS) designed to provide a highly integrated, performance-balanced solution for advanced human-machine interaction. The sensor's high performance is rooted in its unique design: a rabbit-leg bioinspired microstructure enables efficient stress concentration to significantly enhance sensitivity, while a composite of NdFeB nanoparticles in a silicone rubber (SR) matrix facilitates highly sensitive magnetic field detection and novel ternary encoding through field-induced changes in both permittivity and electrode distance, featuring three key advantages. First, the integration of biomimetic microstructures with multiwalled carbon nanotubes (MWCNTs)/NdFeB/silicone rubber (SR) nanocomposites enables simultaneous detection of pressure and magnetic fields. Second, the sensor exhibits a sensitivity of 1.0648 kPa (0-1 kPa), with an ultralow detection limit of 7 Pa. Third, a ternary signal encoding system (-1, 0, + 1) supports contactless information encryption. Fabricated via precision 3D printing, the RF-PMS demonstrates a fast response time of 62 ms and good mechanical durability (>4,500 cycles). For practical applications, it enables robotic object classification through grasp-induced capacitive signals, achieving 95.2% accuracy via a CNN-based framework. Additionally, the device supports secure data transmission using a programmable 4 × 4 magnetic array. This compact, multifunctional platform addresses key limitations in current flexible sensors and opens new opportunities for next-generation wearable devices and human-machine interfaces.
在先进机器人技术和人机界面领域，对能够弥合非接触感知与物理交互之间差距的柔性传感器有着迫切需求。将用于接近检测的非接触磁传感与用于触觉反馈的基于接触的压力传感集成在单个设备中，是满足这一需求的关键途径。然而，由于一个基本的权衡，要在两种模式下都实现高性能具有挑战性：为高压灵敏度优化的材料和结构，往往会因集成用于磁场检测所需的磁性元件而受到影响，反之亦然。为了解决上述问题，本文提出了一种受兔腿启发的柔性压力 - 磁传感器（RF - PMS），旨在为先进的人机交互提供高度集成、性能平衡的解决方案。该传感器的高性能源于其独特设计：受兔腿启发的微观结构能够实现高效的应力集中，从而显著提高灵敏度，而钕铁硼纳米颗粒在硅橡胶（SR）基体中的复合材料则有助于实现高灵敏度的磁场检测，并通过场致介电常数和电极距离变化实现新颖的三元编码，具有三个关键优势。首先，仿生微观结构与多壁碳纳米管（MWCNTs）/钕铁硼/硅橡胶（SR）纳米复合材料的集成，使得能够同时检测压力和磁场。其次，该传感器的灵敏度为1.0648 kPa（0 - 1 kPa），检测极限低至7 Pa。第三，三元信号编码系统（-1, 0, +1）支持非接触式信息加密。通过精密3D打印制造的RF - PMS响应时间快，为62 ms，且具有良好的机械耐久性（>4500次循环）。在实际应用中，它能够通过抓握引起的电容信号实现机器人物体分类，通过基于卷积神经网络（CNN）的框架实现95.2%的准确率。此外，该设备支持使用可编程4×4磁阵列进行安全数据传输。这个紧凑的多功能平台解决了当前柔性传感器的关键局限性，并为下一代可穿戴设备和人机界面开辟了新机遇。
