FinGuard: An AI-Integrated Mobile Application for Predictive Health Monitoring in Ornamental Fish Keeping

Abstract

Fishkeeping has evolved into a commercially and recreationally successful venture in the Philippines, but fishkeepers still struggle in maintaining steady aquatic conditions as well as identifying any signs of fish stress or diseases. Existing fish monitoring solutions are fragmented, as some measure water quality parameters while others observe fish behavior without integrating them to provide predictive health analysis. In this paper, FinGuard – an AI-enabled fish predictive health monitoring mobile application is proposed. The system utilizes IoT sensors to monitor the water quality of ornamental fish aquariums in real-time by collecting data about water parameters, including temperature, pH, dissolved oxygen, and turbidity. FinGuard is able to analyze fish behaviors via a Computer Vision module, which analyzes video recordings and provides predictions of health issues based on the analyzed information and collected environmental conditions. The system also has a logbook for aquariums, an analytics dashboard showing the current status of all aquariums, and a recommendations module providing relevant suggestions about actions necessary to improve fish health. The mobile application is being built using Agile software development lifecycle with the use of Flutter framework for mobile app development, Python and Flask for backend, and TensorFlow and OpenCV libraries for machine learning and computer vision purposes, respectively. Finally, Firebase will be used as a cloud database.