Implementasi Smart Buoy untuk Prediksi Kualitas Air Tambak Udang dengan Double Exponential Smoothing

Abstract

White shrimp farming is highly dependent on water quality stability, particularly temperature and pH parameters that directly affect shrimp growth and survival. In practice, water quality monitoring in many shrimp ponds is still conducted manually and periodically, causing environmental changes to be detected too late for timely intervention. Furthermore, most existing monitoring systems only provide current condition information without predictive capabilities that can support preventive decision-making. This study aims to design and implement a Smart Buoy based on the Internet of Things for real-time water quality monitoring and short-term early warning generation. The proposed system integrates an ESP32 microcontroller, temperature and pH sensors, LoRa communication, Firebase cloud services, and a mobile application as the user interface. The Double Exponential Smoothing method was employed to predict temperature and pH conditions 30 minutes ahead, while model parameters were determined using a walk-forward validation approach. The results demonstrate that the system successfully performs continuous data acquisition, transmission, storage, and visualization of water quality information. Forecasting evaluation yielded Mean Absolute Percentage Error values of 0.62% for temperature and 0.32% for pH. The system also successfully delivered automatic danger and early warning notifications when water quality conditions were detected or predicted to exceed predefined safety thresholds. This study contributes to the development of an IoT-based shrimp pond water quality monitoring and prediction system by integrating a Smart Buoy for more representative data acquisition, the Double Exponential Smoothing method for short-term forecasting, and a mobile application that supports real-time monitoring and faster, more preventive decision-making in shrimp pond management.