A Secure Federated IoT-Based Multi-Sensor Framework for Intelligent Fire Detection and Alarm Systems in Smart Buildings

Fire detection remains a critical component of modern smart infrastructure due to the increasing risks of fire- related incidents in urban and industrial environments. Traditional fire detection systems that rely on single-sensor mechanisms often suffer from delayed responses, false alarms, and poor adaptability in dynamic conditions [1]. To address these limitations, this study proposes a novel AIoT-enabled multi-sensor fire detection and alarm framework that integrates temperature, smoke, gas, and flame sensors with the ESP32 microcontroller and NB-IoT communication technology [3], [10], [11]. The system utilizes cloud-based analytics and threshold-driven decision logic to ensure accurate, real-time alerts while maintaining low energy consumption [2], [5]. Multi-sensor data fusion techniques enhance detection precision by combining complementary sensor inputs to minimize false positives and improve reliability [4], [9]. The prototype was experimentally validated under both fire and non-fire conditions, demonstrating an overall detection accuracy of 94% and reducing false alarms by approximately 30% compared to conventional single-sensor systems [6], [15]. This research provides a scalable, energy-efficient, and intelligent solution suitable for deployment in smart buildings and urban safety networks [7], [8], [13]. The proposed framework paves the way for future integration with predictive analytics, federated learning, and wireless power solutions to further enhance proactive fire prevention and safety management [12], [14].