Smart Fire Guardian: A Hybrid Architecture for Fire Detection and Suppression Using Image Processing and Sensor Fusion in Enclosed Industrial Environments
Smart Fire Guardian: A Hybrid Architecture for Fire Detection and Suppression Using Image Processing and Sensor Fusion in Enclosed Industrial Environments
Shaahin G. Shirazi1 Mani Filoo2
1) Electronics Department, Allameh Helli 1+1 Educational Complex, NODET (National Organization for Development of Exceptional Talents), Iran - Email:
2) Electronics Student, Allameh Helli 1+1 Educational Complex, NODET, Iran - Email:
IOI Code
:
AAHD-ABKCZG
Publication :
National Conference on Information Technology, Nanotechnology, Artificial Intelligence and Technological Futures Studies(itnaf.ir)
Abstract :
Fire incidents in enclosed industrial environments—particularly in warehouses storing flammable materials such as textiles, wood, plastics, and chemicals—represent one of the most critical challenges in industrial safety and risk management. In such environments, rapid initial detection and timely response to fire play a decisive role in minimizing financial losses, preventing disruptions to production processes, and ensuring personnel safety. Conventional fire alarm systems, which primarily rely on smoke and heat sensors, often face limitations under certain operational conditions, including delayed detection, sensitivity to environmental variations, and the occurrence of false alarms. In recent years, significant advances in computer vision, deep learning, and intelligent monitoring systems have enabled the development of more advanced approaches for early detection of fire and smoke. In these approaches, data obtained from surveillance cameras and environmental sensors are analyzed simultaneously, and data fusion techniques are employed to enable more accurate decision‑making regarding fire incidents. This paper presents an analytical review of major fire detection methods based on image processing, environmental sensors, and hybrid systems. Based on the findings of this review, a conceptual framework named “Smart Fire Guardian” is introduced for rapid fire detection and early suppression in enclosed industrial environments. The proposed framework is designed around local processing, offline operation capability, emergency power support, and the integration of automated or drone‑based fire suppression units.
Keywords : fire detection image processing sensor fusion computer vision deep learning industrial safety