Design and Experimental Evaluation of an ESP32-Based IoT Early Warning System for LPG Gas Leakage

Abdulelah Ahmed (Faculty of Computer Science, Universitas Singaperbangsa Karawang, Indonesia)
Aji Primajaya (Faculty of Computer Science, Universitas Singaperbangsa Karawang, Indonesia)

Jurnal Janitra Informatika dan Sistem Informasi
Vol. 6 No. 1 (2026): April - Jurnal Janitra Informatika dan Sistem Informasi

This study aims to design and evaluate an IoT-based early warning system for LPG gas leaks to enhance household safety. The experimental methodology involved system design using an ESP32 microcontroller and an MQ2 gas sensor as core components, integrated with local alarms (LED, buzzer) and remote notifications via the Blynk platform. The analysis was performed through controlled testing with 30 experimental trials to measure system response time, detection accuracy, and notification success rate. The results indicate the system achieved an average response time of 1.4 seconds (SD = 0.2 s), a detection accuracy of 96.7%, and a notification success rate of 100%. The system operates stably and proves effective as an early warning solution, with a practical recommendation for sensor installation near floor level based on the density properties of LPG.