Pembuatan Prototipe Sistem Monitoring Kesuburan Tanah Berbasis Internet Of Things

Abstract

Perubahan iklim dan fluktuasi mikroklimat pada lahan pertanian terbuka menuntut adanya pengelolaan kualitas tanah yang presisi, khususnya pada komoditas rentan seperti tanaman tomat yang membutuhkan kelembaban dan nutrisi spesifik agar terhindar dari kecacatan fisik. Untuk mengatasi masalah tersebut, penelitian ini merancang prototipe sistem monitoring kesuburan tanah berbasis Internet of Things (IoT) menggunakan mikrokontroler ESP32 dan metodologi SDLC Waterfall. Secara teknis, sistem ini mengakuisisi metrik unsur hara (NPK) menggunakan sensor Modbus RS485, serta derajat keasaman (pH), kelembaban, suhu, dan intensitas cahaya. Seluruh data dikonversi ke format JSON dan ditransmisikan via protokol MQTT secara asinkron setiap 5 detik untuk divisualisasikan secara real-time melalui website monitoring smart farming berbasis React.js dan Node.js. Keandalan instrumen divalidasi melalui perhitungan Mean Absolute Percentage Error (MAPE) dengan rata-rata tingkat akurasi kumulatif mencapai 94,93%, mengonfirmasi presisi akuisisi data. Selanjutnya, analisis regresi linear berganda membuktikan secara empiris bahwa parameter iklim mikro dan sifat fisik tanah memiliki intervensi signifikan terhadap ketersediaan hara. Secara spesifik, Uji T menunjukkan bahwa kadar Nitrogen sangat dipengaruhi oleh Suhu Udara (nilai Sig. < 0,001), sementara ketersediaan Fosfor dan Kalium sangat dipengaruhi oleh Kelembaban Tanah dan Intensitas Cahaya (nilai Sig. < 0,001). Melalui integrasi perangkat keras dan analisis data tervalidasi ini, prototipe berhasil menyajikan data empiris untuk mencegah stres pada tanaman tomat dan memberikan justifikasi objektif guna mendukung efisiensi otomasi sistem irigasi presisi.

Climate change and microclimate fluctuations in open agricultural fields require precise soil quality management, particularly for vulnerable crops such as tomatoes, which require specific moisture levels and nutrients to avoid physical defects. To address these issues, this study designed a prototype Internet of Things (IoT)-based soil fertility monitoring system using an ESP32 microcontroller and the Waterfall SDLC methodology. Technically, the system acquires nutrient (NPK) metrics using an RS485 Modbus sensor, as well as pH, moisture, temperature, and light intensity. All data is converted to JSON format and transmitted asynchronously via the MQTT protocol every 5 seconds for real-time visualization through a smart farming monitoring website built with React.js and Node.js. Instrument reliability was validated through Mean Absolute Percentage Error (MAPE) calculations, with an average cumulative accuracy rate of 94.93%, confirming the precision of data acquisition. Furthermore, multiple linear regression analysis empirically demonstrated that microclimate parameters and soil physical properties have a significant impact on nutrient availability. Specifically, the t-test showed that nitrogen levels are strongly influenced by air temperature (Sig. value < 0.001), while phosphorus and potassium availability are strongly influenced by soil moisture and light intensity (Sig. value < 0.001). Through the integration of hardware and this validated data analysis, the prototype successfully provided empirical data to prevent stress in tomato plants and offered objective justification to support the efficiency of precision irrigation system automation.