Water Quality Monitoring Analysis Using Machine Learning and Internet of Things (IoT) for Catfish Farming

Abstract

Water quality is a crucial factor in catfish aquaculture, as it directly affects fish growth, health, and survival rate. This study aims to develop and analyze a water quality monitoring system based on the Internet of Things (IoT), integrated with a Support Vector Machine (SVM) algorithm for real-time classification of water conditions. The system is designed using an ESP NodeMCU microcontroller equipped with a TDS sensor to measure total dissolved solids, a DS18B20 sensor for water temperature, and an OLED display for local visualization. Collected data is transmitted to a server via an internet connection and analyzed using a quantitative approach based on machine learning, employing the SVM algorithm to classify water quality into “good” or “poor” categories. Test results show that the system can continuously monitor water conditions with a classification accuracy of 96.67% using simulated data and 100% using real data. These findings demonstrate that the developed system can serve as an effective decision-support tool for fish farmers, enabling quick and accurate responses to maintain optimal water quality, increase farming efficiency, and reduce fish mortality risk.

Keywords: water quality, catfish aquaculture, Internet of Things, machine learning, Support Vector Machine (SVM)