The development process begins by designing a compact embedded hardware unit consisting of an ESP32 controller, flame sensor, smoke sensor, temperature sensor, relay driver, and miniature water or gas-based extinguishing mechanism. Each sensor is calibrated independently to establish reliable operating ranges before integrating them into a unified fire detection circuit. Instead of responding to a single sensor, the embedded software evaluates multiple environmental parameters simultaneously to improve decision accuracy.
A decision engine is then implemented inside the microcontroller using threshold comparison and timing logic. The controller continuously filters noisy sensor readings, confirms persistent abnormal conditions, and determines whether the situation represents a genuine fire emergency. Once confirmed, the relay activates the extinguishing mechanism while simultaneously recording the event with timestamps and sensor values.
The cloud communication layer is developed using MQTT, enabling each intelligent extinguisher to publish telemetry data securely to a central monitoring server. A backend application processes incoming messages, stores them in a database, and generates device-specific activity logs. Rather than displaying only alarms, the monitoring dashboard also presents sensor calibration status, communication health, battery condition, maintenance schedules, and suppression statistics for every installed unit.
Finally, extensive validation is performed by simulating different environmental scenarios including smoke without flames, heat without combustion, and actual fire conditions. The system response is analysed to minimise false positives while ensuring rapid activation during genuine emergencies. Performance metrics such as detection time, communication latency, suppression activation time, and system reliability are documented to evaluate overall effectiveness before deployment.
