BACKEND IMPLEMENTATION ON RED TILAPIA FEEDING AUTOMATION SYSTEM

• Main Author: Rayhan Ravianda, Muhammad
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/77860
• Institution: Institut Teknologi Bandung

Aquaculture is the largest sector in the fishery resource production industry, especially in Indonesia, but the process is still not running effectively. An automated feeding system using machine learning with a prediction approach for fish hunger conditions is one solution to reduce underfeeding and overfeeding as an effort to streamline aquaculture activities. The red tilapia feeding automation system using machine learning is an integrated fish feeding system to perform automatic feeding with a prediction approach for fish hunger conditions and provide reports on feeding and fish hunger that can be analyzed by users. The backend system is needed to support the running of the application through the server so that it does not burden the power on the application user's device. The fish feeder system with camera is integrated with the backend system to perform feeding and store fish feeding data. The backend system is connected to the machine learning model to get prediction results that are used to determine feeding decisions on the fish feeder system, as well as storing fish hunger data. System testing is carried out by testing the functionality scenarios of the application backend system and the feeder system as a whole and the system's load handling performance. The test results show that the architecture of the backend system and feeder system plays an important role in supporting the functionality of the application so that it can run optimally.