A Wearable Tactile Sensor Array for Large Area Remote Vibration Sensing in the Hand

TL;DR

This paper introduces a wearable, high-resolution tactile sensor array inspired by human hand anatomy, capable of remotely sensing vibrations across the entire hand during natural interactions, advancing tactile sensing technology.

Contribution

The authors developed a 126-channel flexible sensor array that mimics hand anatomy, enabling real-time, high-resolution remote vibration sensing during manual activities.

Findings

Accurately captures whole hand tactile signals during interactions.

Demonstrates high-resolution vibration sensing matching human tactile frequency range.

Enables remote sensing of tactile signals across the hand.

Abstract

Tactile sensing is a essential for skilled manipulation and object perception, but existing devices are unable to capture mechanical signals in the full gamut of regimes that are important for human touch sensing, and are unable to emulate the sensing abilities of the human hand. Recent research reveals that human touch sensing relies on the transmission of mechanical waves throughout tissues of the hand. This provides the hand with remarkable abilities to remotely capture distributed vibration signatures of touch contact. Little engineering attention has been given to important sensory system. Here, we present a wearable device inspired by the anatomy and function of the hand and by human sensory abilities. The device is based on a 126 channel sensor array capable of capturing high resolution tactile signals during natural manual activities. It employs a network of miniature three-axis sensors mounted on a flexible circuit whose geometry and topology were designed match the anatomy of the hand, permitting data capture during natural interactions, while minimizing artifacts. Each sensor possesses a frequency bandwidth matching the human tactile frequency range. Data is acquired in real time via a custom FPGA and an I$^2$C network. We also present physiologically informed signal processing methods for reconstructing whole hand tactile signals using data from this system. We report experiments that demonstrate the ability of this system to accurately capture remotely produced whole hand tactile signals during manual interactions.