A SMART SYSTEM OF HORTICULTURAL PLANT RECOMMENDATIONS USING PARAMETER-BASED RULE-BASED METHODS LAND AND THE INTERNET OF THINGS

SISTEM CERDAS REKOMENDASI TANAMAN HORTIKULTURA MENGGUNAKAN METODE RULE-BASED BERBASIS PARAMETER TANAH DAN INTERNET OF THINGS

Abstract

This research designs and develops an Internet of Things (IoT)-based intelligent system
for recommending horticultural vegetable crop selection using a rule-based expert system
with forward chaining inference technique. Unlike previous studies that generally
focused solely on soil condition monitoring, this system integrates the simultaneous
measurement of seven soil parameters with automated decision-making capabilities to
guide users in precise crop selection. The system utilizes an RS485 Modbus RTU-based
7-in-1 NPK sensor to measure soil moisture, temperature, electrical conductivity (EC),
pH, Nitrogen (N), Phosphorus (P), and Potassium (K). An ESP32 microcontroller
processes the sensor data through a rule base covering 15 vegetable commodities, and
distributes recommendation notifications along with fertilizing guidance via a Telegram
Bot, while all data are synchronized to a MySQL database. Testing was conducted on
three soil conditions: yard soil, compost soil, and agricultural field soil. Results confirm
that the system performs stable data acquisition across all seven parameters, and the
forward chaining logic demonstrated 100% accuracy in matching physical soil data with
the rule base across all tested scenarios. The system proves to be a practical and
affordable decision-support solution for farmers in determining suitable vegetable crops
based on real soil conditions.