<![CDATA[Soybean (Glycine max (L.) Merr.) is a strategic food commodity in Indonesia, with national demand reaching 2.5 million tons annually, while domestic production fulfills only approximately 36% of this requirement. A critical determinant of soybean productivity is the effectiveness of Rhizobium japonicum bacterial inoculation, which performs biological nitrogen fixation in root nodules. Inoculation effectiveness is highly sensitive to soil conditions—including pH, moisture, temperature, and nitrogen content—that vary spatially and temporally across agricultural fields. Farmers' limited access to real-time soil data results in uniform inoculant dosing without consideration of field variability, leading to inefficient bioinoculant utilization and suboptimal inoculation outcomes. This study aimed to design and implement an IoT-based Rhizobium inoculation Decision Support System (DSS) model using the Mamdani Fuzzy Inference System (FIS) method. The system integrates four soil sensors—SEN0169 for pH, Capacitive Soil Moisture v1.2 for moisture, DS18B20 for temperature, and an NPK RS485 Modbus sensor for nitrogen content—all connected to an ESP32-WROOM-32D microcontroller serving as an edge processor. The Mamdani Fuzzy inference engine was developed with 4 input variables, 14 fuzzy sets, and 25 IF-THEN rules formulated in collaboration with agronomy experts, employing Centroid (Center of Gravity) defuzzification to generate inoculation dose recommendations within the range of 0–200 g/ha. The system output is executed by a peristaltic pump via PWM signals, while sensor data and Fuzzy decisions are transmitted to a cloud server via the MQTT protocol and monitored through a web dashboard and mobile application. System validation against assessments from three agronomy experts across 50 soil condition scenarios yielded an agreement rate of 82.3%. Average decision latency from sensor reading to pump activation was 1.38 seconds, with a field uptime of 96.4% over 30 days of testing. The proposed system demonstrates significant potential for improving bioinoculant use efficiency and supporting sustainable enhancement of national soybean productivity.]]>
Copyrights © 2026
