ABSTRACT Fire protection is critical in crowded places with complex buildings because it decreases the chance of life‐threatening occurrences. Traditional systems, which often depend on static exit signs or single danger detection, are not designed to manage the dynamic nature of flames and crowd movement. This paper introduces an IoT‐based intelligent evacuation system that addresses this gap through the People Management and Navigation Scheme (PMNS), a novel algorithmic framework for real‐time decision‐making. The PMNS dynamically synthesizes data from smoke sensors, temperature and humidity sensors, and infrared people‐counting sensors, all of which are deployed on microcontrollers. The system's core contribution is its ability to make integrated decisions based on location‐specific sensor thresholds, preventing false alarms and enabling context‐aware responses. Upon detecting a threshold breach, the system activates LED strip actuators via relay circuits and initiates people counting. It intelligently manages congestion by deactivating LEDs at overcrowded exits and redirecting occupants to safer, less congested alternatives. A custom‐developed web dashboard facilitates a hybrid control strategy, allowing for both autonomous operation and manual supervisory intervention. Real‐time monitoring and alerts are provided to mobile and desktop interfaces. This paper presents the design and thorough validation of a low‐cost, scalable prototype that achieves rapid response times of less than 1 s, precise fire detection, and highly accurate people counting (97.33%). By bridging the gap between theoretical studies and practical implementation, this system demonstrates that sophisticated, adaptive evacuation management is feasible with low‐cost components, marking a significant advancement in intelligent building safety systems.
Khan et al. (Tue,) studied this question.