Optimizing Fish Health in Agricultural Farming Using Digital Twin and IoT-Enabled Online Learning Algorithms

Abstract

Integrating digital twin technology and IoT-enabled systems has revolutionized fish farming by offering real-time monitoring and adaptive control over key parameters crucial for optimal fish health and productivity. This research aims to develop an online learning algorithm integrated with digital twin technology to optimize fish production within a modular aquaculture system continuously. The proposed system analyzes real-time data inputs, including water temperature, pH levels, ammonia concentration, oxygen levels, and feeding routines. It updates control policies and adapts to emerging data trends. The online learning algorithm's capacity to incrementally learn from incoming data helps maintain water quality and supports the health and growth of the fish. Results from extended simulations indicate significant improvements in water conditions and more efficient feeding practices, ultimately leading to enhanced fish growth rates and increased farm productivity. After 1,000 iterations, the health status of fish production was graphically represented as 82.5%. This approach provides a robust framework for automated and real-time decision-making, thereby assisting fish farmers in sustaining healthy and productive aquaculture systems.