Artificial Intelligence and Multi-Sensor Fusion Based Universal Fire Detection System for Smart Buildings Using IoT Techniques

Abstract

The present work focuses on the development of a fire prevention system for buildings. The first subsystem identifies electrical fire causes. The system integrates nonvisual and visual fire detection techniques. The nonvisual approach is based on a fuzzy logic model that fuses electric wire current and temperature information. The visual technique is based on an artificial intelligence approach for fire detection. A deep learning model is deployed on an embedded platform, the Raspberry Pi 3B+. The transfer learning approach applies to a MobileNet deep learning model. The second subsystem identifies the general causes of fire and the necessary actions to be taken to prevent it. The developed fire detection system aims at the early detection of fire with fewer false positives.