Monitoring and Prediction of Water Quality in Catfish Biofloc Ponds at SEIN Farm using IoT and Linear Regression

The catfish industry in Indonesia has shown sustained growth in both cultivation and consumption, making it a crucial sector that the government is actively working to enhance within the broader fishery industry. In order to realize the full potential of this catfish business, meticulous care and handling are imperative. Despite this, catfish farming often grapples with challenges stemming from the quality of pond water. By harnessing the biofloc technique, significant strides can be made in elevating water quality management. However, a manual component persists within this approach, potentially tempering the efficacy of the biofloc method when applied to catfish cultivation. Advancements in contemporary technology offer a beacon of hope, as the integration of digital innovations holds the key to propelling catfish biofloc ponds into a new era. This research endeavors to implement an Internet of Things (IoT)-based framework tailored to monitor and regulate water quality parameters in catfish biofloc ponds. Moreover, the system goes beyond mere monitoring, delving into the realm of predictive analytics through the utilization of linear regression techniques. This ambitious initiative aspires to optimize the efficiency of biofloc ponds, thereby ensuring the production of top-tier catfish specimens while ushering the entire catfish farming landscape into a modernized domain. The resultant predictive model is underscored by its remarkable performance, achieving impressive Mean Absolute Percentage Error (MAPE) values of 9% for dissolved oxygen (DO), 1% for pH, and 1% for temperature predictions