PreSA: an intelligent blockchain-based platform for monitoring and predicting water quality for smart aquaculture

Water quality is an important factor for the survival of most living things, such as the production process of intensive aquaculture systems. In addition, it is crucial to protect fishes from any possible catastrophe caused by pollution. In this context, it is essential to monitor, control and predict water quality to have high-quality fish farming water. In this research, machine learning (ML) and blockchain technologies offer efficient and dependable solution for smart aquaculture providing greater control, management and security. In this paper, we proposed using ML and blockchain to develop an intelligent platform collecting and predicting water pollution using trophic index (TRIX). TRIX index is a metric for assessing the trophic state of an aquatic ecosystem and understanding ecological health. Autoregressive integrated moving average (ARIMA), random forest (RF) and K-nearest neighbor (KNN) models were used to predict TRIX and to help control centers for quick decision-making and real-time interventions. Different evaluation metrics have been used to identify the best ML model. Blockchain is used to secure data and ensure alerts traceability. The evaluation confirms that RF model provides better accuracy compared to other ML models. So, this study provides a secure water quality prediction system to early detect pollution and improve water quality in aquaculture environment.