Estimating the People Density in Restricted Spaces Using Probe Requests for Wi-Fi Connections

Estimating the density of people in a restricted space (such as a classroom, a waiting room, or a carriage of a train or subway) aids both users and administrations in understanding the dynamics within their spaces and managing and organising them effectively. Not all restricted spaces can be equipped with people-counting solutions, and having the exact number of people present is not always necessary. This article presents a potential solution for estimating, non-intrusively, the number of people in a restricted space, assuming that the majority of individuals carry a cell phone. The method estimates the number of occupants by leveraging the capability of a device equipped with Wi-Fi (e.g., an Espressif ESP32) to intercept probe requests used to establish a Wi-Fi connection. Each single device generates these requests, which are associated with its virtual MAC address that changes over time. By computing the number of requests with different MAC per unit of time over an extended duration (e.g., tens of seconds), an approximation of the space's occupancy can be derived. The efficacy of this approach was tested in two distinct real environments: subway carriages and classrooms at the Milan Polytechnic, demonstrating the viability of the proposed solution.