Pengembangan Aplikasi Pengukur Kualitas Air Berbasis IoT Menggunakan Metode Support Vector Machine (SVM)

Abstract

Manual monitoring of bore well water quality requires considerable time and is unable to provide real-time information, thereby posing potential risks to public health. This study develops Waterlytics, an Internet of Things (IoT)-based bore well water quality monitoring system that integrates Support Vector Machine (SVM) classification with a Radial Basis Function (RBF) kernel. The system utilizes a NodeMCU ESP8266 microcontroller equipped with a TDS Meter V1.0 sensor, Turbidity Sensor, and DS18B20 temperature sensor connected through an ADS1115 module. Sensor data are transmitted every 10 seconds to a Flask REST API server, stored in a MySQL database, and automatically classified into three water quality categories based on the Indonesian Ministry of Health Regulation No. 2 of 2023 standards. The SVM model parameters were optimized using Grid Search Cross-Validation with k-fold=5. Classification results are visualized through an interactive Flask-based web dashboard and delivered as automatic notifications via a Telegram Bot. The results indicate that the SVM model achieved an accuracy of 93.94% with optimal parameters of C=10 and gamma=10, exceeding the minimum target of 85%. Sensor validation against laboratory reference values from the Bali Provincial Environmental and Forestry Agency (DKLH) showed deviations below 2% for all measured parameters. Furthermore, both Black Box testing (25 scenarios) and White Box testing (19 logic cases) were successfully completed.