image: The systematic framework for the proposed multi-liquid identification is depicted in Fig. 1. The system comprises a space-time-coding metasurface illuminated by a feeding pyramidal horn antenna, a receiving monopole antenna, a personal computer (PC), and a universal software radio peripheral (USRP) device. The metasurface is dynamically controlled by a field-programmable gate array (FPGA)-based circuit, while the USRP device interfaces with the feeding horn, the receiving antenna, and the PC. An orthogonal frequency division multiplexing (OFDM) signal is generated on the PC, modulated by the USRP, and transmitted via the feeding horn antenna. The metasurface, governed by the FPGA control circuit, transforms the incident signal into multiple radiation beams, each characterized by distinct frequencies and propagating along predefined directions. These beams individually illuminate the liquids positioned along their respective paths. The signals scattered by the multiple liquids are captured by the monopole antenna and subsequently transmitted to the PC for analysis. Notably, AI classification algorithms are implemented on the PC to accurately identify and distinguish the different liquids.
Credit: ©Science China Press
Professor Shi's team has long been dedicated to metasurface technology and its applications. This research was inspired by two pressing real-world needs: the potential threat of explosive liquids to public safety and the efficiency limitations of conventional liquid security screening technologies. "In practical scenarios like airport security checks, rapid and accurate identification of hazardous liquids has always been a major challenge," Professor Shi stated in an interview. "We hope our breakthrough in metasurface technology can provide more efficient solutions for public safety." To tackle these critical challenges, the team pioneered a groundbreaking integration of three cutting-edge technologies: spatiotemporally programmable metasurface architecture, software-defined radio platforms, and advanced artificial intelligence classification algorithms. This synergistic combination successfully breaks through the fundamental constraint of conventional detection systems that were limited to single-target recognition with low accuracy.
"Our team is the first to apply the harmonic modulation characteristics of space-time-coding metasurfaces to liquid detection," Professor Shi emphasized during the interview. "This interdisciplinary innovation not only solves the accuracy and efficiency bottlenecks of traditional detection methods but, more importantly, provides a completely new technological pathway for public safety." The research team developed a specific space-time-coding scheme that enables the metasurface to generate multiple independent harmonic detection channels. They also created an AI-based CSI (Channel State Information) feature extraction algorithm for precise liquid identification and significantly improved anti-interference capabilities through system architecture optimization. Experimental results demonstrate that the system achieves near-perfect accuracy in simultaneous detection of multiple liquids, with notable advantages in cost-effectiveness and ease of deployment, offering a revolutionary technological solution for hazardous liquid detection.
"This technology is just the first step in our intelligent metasurface sensing platform," Professor Shi explained. "Going forward, we will focus on three key breakthroughs: first, integrating the system with millimeter-wave technology to achieve molecular-level liquid identification with higher precision; second, developing miniaturized chip integration solutions for embedding in mobile devices; and third, building cloud-based intelligent analysis networks to provide real-time security monitoring for large public spaces." The research team plans to expand applications into biomedical fields (such as non-invasive glucose monitoring) and environmental monitoring while collaborating with industry partners to promote industrialization. Dai Shihan added, "This will establish an important milestone in translating metasurface technology from the laboratory to practical applications."
See the article:
MetaDetection: Wireless intelligent detection of multiliquids by using space-time-coding metasurface
https://doi.org/10.1093/nsr/nwaf245
Journal
National Science Review