Implementasi Sistem Notifikasi Listrik dan Peringatan Pemadaman Lampu Berbasis IoT Pada Ruang Server BMKG

Abstract

Kestabilan pasokan listrik merupakan faktor kritis untuk menjaga operasional ruang server BMKG dari risiko kerusakan perangkat dan kehilangan data akibat pemadaman tak terdeteksi. Untuk mencegah keterlambatan penanganan cadangan daya (UPS), Proyek Akhir ini mengimplementasikan sistem monitoring listrik berbasis Internet of Things (IoT) menggunakan mikrokontroler ESP32, sensor PZEM-004T, komunikasi redundan, notifikasi Telegram Bot, dan alarm lokal. Hasil pengujian menunjukkan akurasi sensor yang tinggi dengan Mean Absolute Percentage Error (MAPE) tegangan sebesar 0,642% dan arus 4,474%. Sistem peringatan dini merespons secara cepat dengan rata-rata pengiriman notifikasi Telegram sebesar 9,90 detik, serta dilengkapi fitur filter false alarm pada unit receiver yang menunda pemicuan alarm hingga rata-rata 5,16 menit. Implementasi selama 14 hari berturut-turut di Ruang Server BMKG mencatat nol kejadian pemadaman dengan efisiensi daya internal alat hanya sebesar 2 Watt, di mana keandalan peringatan daruratnya telah divalidasi melalui simulasi pemutusan daya manual. Lebih lanjut, analisis regresi linier menunjukkan stabilitas tegangan pada rata-rata 231,51 Volt dengan koefisien determinasi (R^2) 0,8633, yang membuktikan kuatnya fenomena voltage drop di mana setiap kenaikan tegangan 1 Volt diikuti penurunan daya terukur sebesar 67,25 Watt. Keberhasilan implementasi sistem ini secara efektif menjamin ketersediaan informasi status kelistrikan secara real-time guna mendukung keamanan operasional perangkat server dalam jangka panjang.

The stability of the power supply is a critical factor in maintaining BMKG server room operations against the risk of hardware damage and data loss due to undetected power outages. To prevent delays in managing backup power (UPS), this Final Project implemented an Internet of Things (IoT)-based electricity monitoring system utilizing the ESP32 microcontroller, PZEM-004T sensor, redundant communication, Telegram Bot notifications, and a local audio alarm. Test results indicated high sensor accuracy with a Mean Absolute Percentage Error (MAPE) of 0.642% for voltage and 4.474% for current. The early warning system responded swiftly with an average Telegram notification delivery time of 9.90 seconds, supported by a false alarm filter on the receiver unit that delays the alarm trigger by an average of 5.16 minutes. A 14-day continuous implementation in the BMKG Server Room recorded zero power outage incidents with a highly efficient internal device power consumption of only 2 Watts, where the emergency alert reliability was successfully validated through manual power disconnection simulations. Furthermore, linear regression analysis demonstrated voltage stability at an average of 231.51 Volts with a coefficient of determination (R^2) of 0.8633, statistically proving the voltage drop phenomenon where every 1 Volt increase is followed by a 67.25 Watt decrease in measured power. The successful implementation of this system effectively ensures the availability of real-time electricity status information to support the long-term operational security of the server equipment.