Insect Antennae Inspired Tactile Sensor and Perception Algorithm

Abstract

Tactile sensors are essential in robotics, smart devices, and human–machine interaction. Among them, skin-like planar tactile sensors support adaptive gripping and manipulation but face major challenges such as complex wiring, limited deformability and mechanical robustness. As supplement, organisms often rely on cilia-like tactile organs that exhibit high sensitivity, mechanical compliance, and omnidirectional perception. For instance, the cockroach antenna, with distributed mechanoreceptors and graded stiffness, effectively extracts environmental features from continuous contact and remains functional even after physical damage.

Inspired by this, this dissertation presents the development of an artificial antenna-like tactile sensor (E-Antenna) with bio-competitive performance in fundamental sensing and tactile-induced environmental perception, aiming to advance tactile sensing technologies in robotics. The E-Antenna adopts a slender, high-aspect-ratio structure made of flexible elastic material with embedded magnetic particles, enabling a non-wiring magnetic sensing approach. This design ensures excellent mechanical compliance and robustness under extreme deformations, including 1800° twist, 224% stretch, 360° bending, large compression, and punctures, making it highly effective in active interaction and environmental exploration.

To realize directional tactile perception, the magnetic particles are axially magnetized, producing distinct responses to stimuli from different directions. The sensor achieves a directional recognition standard deviation as low as 0.13°, with a maximum of only 1.76°, outperforming biological antennae by a factor of 17. Furthermore, theoretical modeling of load-deflection and deflection-magnetism relationships guides the implementation of two optimization strategies: segmented flexibility and partial magnetization. These strategies enhance sensitivity and signal-to-noise ratio by 220%, and extend sensing range by 73%, enabling the E-Antenna to detect fine environmental features critical for reliable tactile perception.

To facilitate integration into various robots, we adopt a modular and scalable design strategy for the E-Antenna. Four prototypes of different scales were implemented across diverse platforms. A centimeter-scale E-Antenna was mounted on a mobile robot to enable vision-free navigation along curved walls with a tracking deviation of just 0.2 mm. A millimeter-scale version was installed beneath a mobile robot for ground texture recognition, achieving 97% accuracy. For dexterous manipulation, a larger E-Antenna was embedded in a brush attached to a robotic arm, enabling conformal brushing on serpentine surfaces with a stable force variance of only 0.34 N. In biomedical applications, a miniaturized E-Antenna was integrated into a toothbrush to provide precise multi-dimensional force feedback and enable brushing pattern recognition for habit assessment. Additionally, we developed a portable E-Antenna-based spirometer to support long-term pulmonary monitoring and daily at-home breathing training for patients.

In summary, this dissertation imitates insect's antennae and proposes a flexible electronic antenna with sensitive omnidirectional sensing capability and excellent mechanical robustness. By integrating bioinspired design, theoretical modeling, optimized design strategies, and AI-powered tactile-induced perception algorithms, the proposed tactile sensory system achieves bio-competitive tactile sensitivity, remarkable mechanical robustness, and advanced environmental perception capabilities. Its effectiveness has been validated through three robotic applications and one biomedical application, offering valuable insights into tactile-induced environmental perception and interaction. These findings pave the way for significant advancements in both robotic and healthcare applications.

Date of Award	31 Jul 2025
Original language	English
Awarding Institution	City University of Hong Kong
Supervisor	Lu LIU (Supervisor) & Yajing SHEN (Supervisor)