An AI-based smart water quality monitoring system using synthetic sensor networks for sustainable management
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
In addition, water quality monitoring is a fundamental aspect of sustainable management of water resources, but traditional monitoring systems have certain drawbacks, such as limited monitoring spatial coverage, delayed detection of anomalies, and a lack of incorporation of predictive intelligence and secure data transmission capabilities. To overcome such issues, this research aims to develop an intelligent water quality monitoring system based on synthetic sensor networks and artificial intelligence for the sustainable management of water resources. To this aim, this research proposes an intelligent water quality monitoring system based on synthetic sensor networks and artificial intelligence, which incorporates various aspects of multivariate environmental simulation, IoT-inspired sensor nodes, secure data communication via AES encryption and RSA-based digital signatures, blockchain-based data integrity logging, graph-based spatial anomaly estimation, Bayesian LSTM-based forecasting with uncertainty quantification, autoencoder-based reconstruction anomaly scoring, and neural decision fusion for risk assessment. The system also attempts to mimic real-world water monitoring situations under both normal and adverse conditions, such as flood, drought, and extreme rain. The system also attempts to consider sensor drift, incomplete data, energy changes, and node trust in the process. The experimental results on the synthetic multivariate water quality dataset show that the proposed framework can effectively capture the temporal, spatial, and anomalous behavior in distributed sensing systems. The proposed framework’s performance in predicting temperature levels for the water quality dataset was measured at an RMSE of 0.436 and an MAE of 0.343. The proposed framework’s performance in classifying anomalies for the water quality dataset was measured at an accuracy level of 95.6 percent. The proposed framework provides a reliable, secure, and intelligent solution for the next-generation smart water quality monitoring system. Keywords: Smart Water Quality Monitoring, Synthetic Sensor Networks, Bayesian LSTM, Graph-Based Anomaly Detection, Sustainable Water Management.
Fields of Study & Subjects
Water Quality Monitoring Technologies
Environmental Monitoring and Data Management
Water Quality Monitoring and Analysis
Publication details
- Journal / source
- Computer Science & IT Research Journal
- Volume
- 7
- Issue
- 3
- Pages
- 228–248
- Publication date
- 26 March 2026
- Language
- EN