Design and Development of an Embedded Fire Detection System using Neural Networks


  • A. Hassini Algeria


Artificial Neural Network, Embedded System, Fire Detector, Sensors.


The failure to respond to a fire alarm is a prominent threat to the safety of human and property in the fire area and the nearby area. One of the reasons for the lack of response to a fire alarm is the tendency to ignore most alarms due to the repeated false alarms. This study proposes an intelligent fire detection system based on the detection of smoke density, temperature, and carbon monoxide in the environment where the fire happens. The optical smoke detector represents the main device to detect fire by activating the other sensors to analyze data from the environment. The fire detection algorithm of this system is based on an ANN (artificial neural network). The emergency case is controlled by the ATMEGA16 microcontroller and confirmed by alarm, which subsequently sends SMS with Specify fire location via Global System for Mobile Communications (GSM). This proposed system proves, through its hardware and software design, a remarkable improvement of system performances in terms of the low cost of the design by dividing the structure into two units; fire detection unit and setup unit, in terms of low power consumption by controlling the working time of each element, and other features such as the calibration, where it allows the system to adapt in their environment. This system detects fire after analyzing chemical and physical features during a period of fewer than 30 seconds to trigger the alarm, including preheat time of the gas sensor. The results show that the proposed system can distinguish between fire and no fire situation, where some situations can be considered as a false alarm when using the smoke alarm, whereas the results of our system were similar to the expected states; therefore, false alarm will be reduced or eliminated.




How to Cite

OUANID, M., & A. Hassini. (2021). Design and Development of an Embedded Fire Detection System using Neural Networks. Journal of Telecommunication, Electronic and Computer Engineering (JTEC), 13(2), 9–15. Retrieved from