BIT: Battery-free, IC-less and Wireless Smart Textile Interface and Sensing System

TL;DR

BIT tackles the challenge of wearable textiles that require no batteries, ICs, or connectors by using near-field electromagnetic coupling between a reader and a textile receiver coil to wirelessly power and read sensors. The method extends resonant sensing to $N$-parallel series $RLC$ circuits, enabling resistive, capacitive, and inductive sensing with concurrent operation of up to three sensors while accounting for transmission-line effects and coil misalignment. A mathematical representation of the impedance $Z(f)$ and a three-step sensor-value estimation algorithm are developed and validated by simulations and a user study, achieving average sensor-estimation accuracies above 90% and interaction-classification accuracy around 93%. The approach reduces embedded electronics, improves manufacturability and sustainability of smart textiles, and supports flexible deployment on garments and accessories.

Abstract

The development of smart textile interfaces is hindered by the inclusion of rigid hardware components and batteries within the fabric, which pose challenges in terms of manufacturability, usability, and environmental concerns related to electronic waste. To mitigate these issues, we propose a smart textile interface and its wireless sensing system to eliminate the need for ICs, batteries, and connectors embedded into textiles. Our technique is established on the integration of multi-resonant circuits in smart textile interfaces, and utilizing near-field electromagnetic coupling between two coils to facilitate wireless power transfer and data acquisition from smart textile interface. A key aspect of our system is the development of a mathematical model that accurately represents the equivalent circuit of the sensing system. Using this model, we developed a novel algorithm to accurately estimate sensor signals based on changes in system impedance. Through simulation-based experiments and a user study, we demonstrate that our technique effectively supports multiple textile sensors of various types.

Figures (10)

• Figure 1: (a) The equivalent circuit of a traditional resonant sensor system. (b) The equivalent circuit of our entire system, which consists of a reader circuit and a smart textile interface circuit. The interface includes a receiver coil ($L_r$) connected in parallel with multiple transmission and sensor circuits. Each transmission line circuit contains parasitic resistance ($R_{line}$), inductance ($L_{line}$), and capacitance ($C_{line}$), connected to a sensor circuit with a resistor (r), capacitor (c), and inductor (l) in series. The reader circuit includes a voltage exciter, internal load (R), and transmitter coil ($L_t$), with parasitic capacitance ($C_{SMA}$). Impedance is measured by capturing voltage at point (n) and applying the voltage divider rule.
• Figure 2: The prototype used in our experiment to validate the accuracy of our equivalent circuit model. The reader was constructed using a NanoVNA NanoVNA connecting to a standard NFC transmitter coil (39mm by 42mm with 4 turns) MFRC-522coil through a SubMiniature version A (SMA) connector SMAconnector. For the smart textile interface, each sensor circuits consisted of a resistor, an inductor and a capacitor, connecting to an embroidered receiver coil (39mm by 39mm with 5 turns) in parallel via transmission line with lengths of 100mm, 200mm, 300mm and same gap of 10mm. Detailed physical parameters and electrical attributes of components are illustrated in the figure. The coupling factor of the two coils was measured to be around 0.53 using a 2-port VNA to measure mutual inductance Jeon2019. The fabrication process is the same as described in Section \ref{['sec_implementation_of_smart_textile_interface']}.
• Figure 3: The ground truth S11 values (blue and orange) and the predicted results using our model (green and red).
• Figure 4: The capacitance, inductance and resistance of the twisted transmission lines shown by the length.
• Figure 5: The absolute impedance spectrum of the entire system and the supposed resonant frequencies (red line) of each sensor circuit used in the model validation. The green lines illustrate the value of Eq. \ref{['eq_rf']} calculated by the two sensor circuits with lower resonant frequency. Their intersections with the measured impedance spectrum accurately represent the resonant frequencies of the sensor circuits. Note that the resonant frequency of the third sensor circuit is 30.9MHz, which exceeds 30MHz and is not plotted.