Tourist overcrowding significantly affects visitors’ experiences, local residents’ quality of life, and environmental sustainability. Traditional monitoring approaches based on static sensors are limited, especially during temporary events with unpredictable crowd dynamics. To address these challenges, a flexible and adaptive solution for real-time crowd monitoring is being developed through ongoing funded projects and will be presented in this tutorial. The system employs low-cost, scalable sensors capable of detecting mobile devices through Wi-Fi signals, while effectively handling MAC address randomization. These sensors support communication over multiple radio access networks (multi-RAN), dynamically optimizing connectivity and enabling deployment in suboptimal locations. To enhance detection accuracy, the system includes an adaptive calibration mechanism using drone- based sensitivity mapping and automatic parameter optimization. A semi-supervised machine learning fingerprinting approach is also under development to ensure resilience to device updates and new models, supported by data gathered in a controlled Faraday cage environment. Ongoing pilot deployments during temporary public events organized by the Municipality of Lisbon are validating the system’s flexibility and ease of use. To support broader adoption, the project will produce comprehensive documentation, tutorials, and instructional videos, facilitating operational autonomy for municipalities, law enforcement, and civil protection authorities.

--