Exploring quantum sensing for fine-grained liquid recognition

TL;DR

This paper introduces a quantum wireless sensing system that significantly improves fine-grained liquid recognition accuracy by reducing thermal noise, enabling precise identification of liquids and pH levels non-intrusively.

Contribution

The paper presents a novel quantum receiver for wireless sensing that eliminates traditional electronic components, enhancing sensing granularity and accuracy for liquid recognition tasks.

Findings

Recognizes 17 liquids with over 99.9% accuracy.

Achieves 99.0% accuracy in pH measurement at 0.1 granularity.

Outperforms conventional sensing methods in fine-grained liquid recognition.

Abstract

Recent years have witnessed the use of pervasive wireless signals (e.g., Wi-Fi, RFID, and mmWave) for sensing purposes. Due to its non-intrusive characteristic, wireless sensing plays an important role in various intelligent sensing applications. However, limited by the inherent thermal noise of RF transceivers, the sensing granularity of existing wireless sensing systems are still coarse, limiting their adoption for fine-grained sensing applications. In this paper, we introduce the quantum receiver, which does not contain traditional electronic components such as mixers, amplifiers, and analog-to-digital converters (ADCs) to wireless sensing systems, significantly reducing the source of thermal noise. By taking non-intrusive liquid recognition as an application example, we show the superior performance of quantum wireless sensing. By leveraging the unique property of quantum receiver, we propose a novel double-ratio method to address several well-known challenges in liquid recognition, eliminating the effect of liquid volume, device-target distance and container. We implement the quantum sensing prototype and evaluate the liquid recognition performance comprehensively. The results show that our system is able to recognize 17 commonly seen liquids, including very similar ones~(e.g., Pepsi and Coke) at an accuracy higher than 99.9\%. For milk expiration monitoring, our system is able to achieve an accuracy of 99.0\% for pH value measurements at a granularity of 0.1, which is much finer than that required for expiration monitoring.