Dependable Fire Detection System with Multifunctional Artificial Intelligence Framework.

Area of Science:

• Artificial Intelligence
• Computer Science
• Smart City Technology

Background:

• Traditional fire detection systems lack accuracy due to simple sensors and struggle with environmental changes.
• Existing machine learning and rule-based methods are often static and cannot adapt to dynamic environments.
• Network latency in legacy systems hinders prompt fire detection, crucial for smart city safety.

Purpose of the Study:

• To propose a novel fire detection system integrating a multifunctional AI framework and a data transfer delay minimization mechanism.
• To enhance the accuracy and speed of fire detection for improved smart city safety.
• To address limitations of current systems in terms of adaptability and promptness.

Main Methods:

• Development of a multifunctional artificial intelligence framework comprising multiple machine learning algorithms and an adaptive fuzzy algorithm.
• Implementation of Direct-MQTT based on Software-Defined Networking (SDN) to mitigate traffic congestion issues.
• Verification of system performance through accuracy and delay time measurements.

Main Results:

• The proposed fire detection system achieved an accuracy exceeding 95%.
• End-to-end delay, including transfer and decision times, was reduced by an average of 72%.
• The system demonstrated superior performance compared to existing fire detection solutions.

Conclusions:

• The developed AI-driven fire detection system offers a significant improvement in accuracy and speed.
• The integration of SDN-based Direct-MQTT effectively minimizes data transfer delays.
• This system provides a robust solution for enhancing the safety and responsiveness of smart cities.