<![CDATA[The risk of fire due to liquefied petroleum gas (LPG) leaks poses a serious threat to the operations of small and medium-sized snack food enterprises (MSMEs). This study aims to design and build an early warning and automatic protection system against LPG gas leaks and fire indications based on the Internet of Things (IoT). The system is controlled by an ESP32 microcontroller, integrating an MQ-2 sensor to detect gas concentration, a DHT22 sensor for environmental temperature monitoring, and an IR Flame sensor to detect fire radiation. Decision-making in the system is implemented using the Mamdani Fuzzy Logic method, which processes input data through 18 rule bases to classify environmental conditions into Safe, Alert, and Danger statuses. At the protection level, the system is equipped with a relay to cut off the electrical supply, a solenoid valve to stop the gas supply, and a buzzer as a warning alarm. Real-time data monitoring is carried out via the ThingSpeak dashboard, supported by a Micro SD module as a local data logger. The results of testing over seven days show that the Fuzzy Logic integration operated optimally without triggering any false alarms (zero false-trigger). Immediately upon detection of a Danger condition, the actuators successfully responded to the commands to cut off the current and close the gas valve in less than one second, with a 100% success rate. The application of IoT and Fuzzy Logic technology is proven to provide reliable, efficient, and flexible mitigation to improve the safety standards of MSME kitchens.]]>
Copyrights © 2026
