Sistem Smart Irrigation Stand-alone Berbasis Energi Surya dengan Pendekatan Manajemen Daya Adaptif

Abstract

Laporan akhir ini mengevaluasi fungsi perangkat, pola telemetri, dan konsistensi manajemen daya pada sistem irigasi skala pot yang dikembangkan dalam kegiatan magang di BRIN. Sistem menggunakan ESP32, panel surya, solar charge controller, aki, dua INA219, sensor lingkungan, dua sensor kelembapan media, relay, pompa, dan Blynk. Keputusan penyiraman menggunakan kelembapan pot otomatis dan tegangan aki dengan pembatasan durasi serta jeda operasi. Pengujian menggunakan dua pot kacang hijau dengan penyiraman otomatis dan manual. Data 1 Juli–6 Agustus 2026 mencakup 11.423 rekaman dalam 39 sesi. Penggunaan satu aki 6 V pada 1–7 Juli dibedakan dari konfigurasi dua aki seri 12 V pada 9 Juli–6 Agustus. Analisis menggunakan statistik deskriptif, perbandingan kelembapan berpasangan, dan pemeriksaan kode Energy Management System (EMS) terhadap ambang program tanpa interpolasi data kosong. Sebanyak 3.178 dari 3.192 pasangan tegangan aki dan kode EMS (99,56%) sesuai dengan aturan program. Tersedia 3.222 pasangan kelembapan media, tetapi banyak pembacaan mencapai batas relatif 100%. Dokumentasi uji fungsi memperlihatkan aliran air, sedangkan status pompa aktif tidak tersedia pada histori yang diekspor. Hasil mendukung keterlaksanaan pemantauan dan konsistensi klasifikasi EMS berbasis tegangan, bukan besarnya penghematan. Efisiensi air, efisiensi energi, dan peningkatan pertumbuhan tidak dihitung karena pengukuran pendukungnya tidak tersedia.

This final project evaluated device functions, telemetry patterns, and power-management consistency in a pot-scale irrigation system developed during an internship at BRIN. The system used an ESP32, solar panels, a solar charge controller, batteries, two INA219 modules, environmental sensors, two media-moisture sensors, a relay, a pump, and Blynk. Watering decisions used automatic-pot moisture and battery voltage with operating-duration and cooldown limits. Two mung bean pots received automatic and manual watering. The dataset covered 1 July–6 August 2026 and contained 11,423 records in 39 sessions. The single 6 V battery used on 1–7 July was distinguished from the two series-connected batteries used as a nominal 12 V supply on 9 July–6 August. Analysis comprised descriptive statistics, paired moisture comparisons, and checks of Energy Management System (EMS) codes against program thresholds without interpolating missing data. Of 3,192 valid voltage-code pairs, 3,178 (99.56%) agreed with the rules. There were 3,222 paired moisture observations, although many reached the relative 100% limit. Functional-test documentation showed water delivery, whereas active pump status was absent from the exported history. The findings support monitoring implementation and voltage-based EMS classification consistency, rather than quantified savings. Water-use efficiency, energy efficiency, and growth improvement were not calculated because supporting measurements were unavailable.