Dr Farzad Tofigh - University of Technology Sydney

Accurate estimation of crowd distribution is fundamental to the safe and efficient operation of public buildings, transport infrastructure, and large indoor venues. Existing approaches predominantly rely on vision-based systems or device-dependent wireless technologies, both of which introduce significant limitations relating to privacy, deployment complexity, scalability, and operational cost. This research investigates a new paradigm for device-free crowd sensing based on distributed electromagnetic-field monitoring and collaborative estimation. 

The project exploits distributed metamaterial-inspired sensing structures to measure subtle perturbations in ambient electromagnetic fields caused by human presence and movement. Unlike conventional sensing modalities, metamaterial sensors provide a low-cost, passive and privacy-preserving means of observing environmental changes without requiring wearable devices, active participation or line-of-sight. By translating spatial variations in electromagnetic energy into informative sensing signatures, these sensors enable continuous monitoring of occupancy and crowd dynamics in complex indoor environments. The proposed system uses a network of spatially distributed sensing nodes to collaboratively estimate crowd density and spatial distribution. Each node observes local field perturbations, while the network collectively reconstructs broader patterns of crowd distribution. This architecture enables scalable, robust and privacy-preserving monitoring in large indoor environments where communication, computation, power and deployment resources may be limited. 

By combining metamaterial-enabled wireless sensing, RF signal processing and distributed intelligence, the research aims to establish a new generation of environment-agnostic crowd-monitoring systems capable of near-real-time operation. The resulting methodology will provide a foundation for scalable sensing analytics from sparse measurements, with potential applications in smart buildings, critical infrastructure, transport hubs and future smart city environments. More broadly, the work advances distributed sensing and estimation methods for extracting useful spatial information from limited data.