SISTEM PENDETEKSI DINI KESUBURAN TANAH BERBASIS IoT MENGGUNAKAN SENSOR KELEMBAPAN DAN WARNA

Soil fertility is a crucial factor in supporting agricultural productivity. Conventional methods for determining soil fertility, such as visually observing soil color or checking moisture by touch, tend to be subjective and inaccurate. This research develops an IoT-based early detection system for soil fertility using a soil moisture sensor and a TCS3200 color sensor. Data from both sensors is processed using a Mamdani-type Fuzzy Inference System (FIS) to classify soil fertility into three categories: fertile, moderately fertile, and infertile. The device is controlled by a NodeMCU ESP8266 microcontroller connected to a Wi-Fi network to transmit measurement results in real time to Google Sheets and display them on a 0.96-inch OLED display. Testing was conducted through sensor calibration and field measurements on various soil samples. Calibration results demonstrated high sensor accuracy with an average moisture sensor error of 1.9%. Field testing demonstrated the system's ability to accurately classify soil fertility based on the measured color and moisture parameters. This research is expected to help farmers monitor land conditions efficiently, objectively, and in real time, thereby supporting the implementation of small-scale precision agriculture.
Keywords: IoT, soil fertility, moisture sensor, TCS3200 color sensor, fuzzy Mamdani