Multi-Sensor Edge Computing Framework for Precision Shrimp Aquaculture with Real-Time Water Quality and Biomass Monitoring

Abstract

As global demand for shrimp rises and the importance of resource management becomes more critical, precision shrimp aquaculture has become essential for sustainable seafood production. Traditional shrimp production is known to have poor water quality assessment, delayed water quality responses, high operation costs, and low productivity due to pond instability. Thus, there is a need for real-time intelligent monitoring systems to enhance productivity and environmental sustainability. This research introduces a multi-sensor edge computing model to monitor water quality and biomass parameters in shrimp farms in real time, supporting precision shrimp farming. The system combines pH, dissolved oxygen, temperature, salinity, turbidity and ammonia sensors with edge enabled processing units providing low-latency monitoring and intelligent analytics. Water-quality prediction programs are based on Long Short-Term Memory (LSTM) models and underwater image analysis to predict biomass is based on MobileNetV2 models. The experimental results show that the monitoring accuracy is high, the inference time is reduced and the energy efficiency is improved when compared to traditional cloud-based systems. The proposed framework was able to predict the water quality with an accuracy of more than 95% and biomass estimation with an accuracy of more than 93%. It offers a scalable, energy-saving and smart solution for future shrimp precision farming.