<![CDATA[Aeroponics is a cultivation method that involves suspending plant roots in the air. Aeroponic cultivation poses several challenges, such as maintaining optimal environmental conditions. In this research, the author designed an internet of things (IoT)-based automated environmental control system by implementing machine learning classification methods. The aim was to optimize the control function of environmental conditioning in aeroponic systems. The system was designed using an Arduino Uno microcontroller connected serially to a Raspberry Pi and connected to the Adafruit Io web platform via the MQTT communication protocol over the internet. In the classification system testing, the initial dataset underwent oversampling using the SMOTE method. The evaluation results showed that the K-Nearest Neighbors algorithm achieved an accuracy of 96,47% in classifying environmental conditions, while the Random Forest algorithm achieved an accuracy of 100%. Both algorithms successfully classified the aeroponic environmental conditions. In terms of the monitoring system, the system was able to send monitoring data to a web-based dashboard with a delay ranging from 0.091 seconds to 0.0142 seconds. Keywords: Aeroponic, Internet of Things (IoT), classification algorithm, KNN, Random Forest]]>
Copyrights © 2023
