ESP-based fire and gas leak detection with automated response

Abstract

Safety hazards for LPG in institutional kitchens include leaking gas, flames, heating of gas tanks, and malfunction of devices. At present, the Bindura University of Science Education Kitchen uses a manual monitoring system, where one detects flames, gas smell, and changes in body temperature which may be difficult to employ during times of heavy workload, especially if the kitchen is unoccupied. This paper provides details about the design and implementation of internet of things (IoT) system for automatic detection and reaction to fire and gas leakages using ESP32. This IoT system employs the DHT22, MQ-6, MQ-9, and the flame sensors. Normally closed solenoid valve shuts off the gas in case of any hazards, including gas leaks, presence of smoke, and fire. The 5V pump is used to pour water for firefighting purposes in case of presence of fire or smoke. The 5V fan keeps room and gas tank cool, while the servo motor moves the nozzle of the pump towards the source of fire or smoke. Results from controlled system testing have shown that there was 100% accuracy of flame detection after over 100 tests. The measuring accuracy of the DHT22 was within ±0.5°C while its stability was ±1°C. The error of the MQ-6 sensor was 3% while the stability was ±35 units whereas the error of MQ-9 sensor was 5% while the stability was ±28 units. On average, it takes 2 seconds for activation of all safety measures in automatic mode. Such a cost-effective system offers many advantages in terms of safety in LPG systems.