Prediction of Oxygen Consumption in Milkfish Larvae using Single Exponential Smoothing Method in Edge Device

Abstract

Oxygen consumption in larvae is one of the physiological indicators that describe the metabolic status of larvae related to energy use. This study aims to develop an Internet of Things (IoT)-based dissolved oxygen consumption prediction system for milkfish larval production. The system is designed to monitor important parameters such as dissolved oxygen (DO) in real-time. This study was conducted using milkfish larvae as samples obtained from spawning milkfish mothers. Measurement of dissolved oxygen parameters was carried out using a sensor connected to an Arduino microcontroller, and the read data was transmitted at three-minute intervals to a Raspberry Pi as the data processing center in the system. Based on the data that has been collected by the Raspberry Pi, it is then processed to produce prediction data. Data analysis and forecasting model development uses the Single Exponential Smoothing method to predict future dissolved oxygen levels. Implementation of the system in an aquarium shows that the system can monitor and record water quality parameter data in real-time. Actuators such as water pumps can be activated based on dissolved oxygen predictions to maintain optimal conditions for milkfish larvae. Experimental results show that our IoT system can monitor and predict dissolved oxygen levels for milkfish larvae production.

Access to Document

Other files and links

• Link to publication in Scopus

Fingerprint