Online Detection of Water Contamination Under Concept Drift

2025-01-03

AI Summary

Existing water quality monitoring systems exhibit insufficient robustness under concept drift—such as sudden contamination events—particularly due to chlorine sensor drift in drinking water distribution networks, leading to false positives or missed detections. To address this, we propose an unsupervised online contamination detection and localization method. Our approach introduces a novel dual-threshold concept drift detection mechanism that explicitly models sensor bias as concept drift; designs a decentralized deployment architecture enabling node-level contamination source localization; and establishes an end-to-end framework integrating an LSTM-based variational autoencoder (LSTM-VAE) with the dual-threshold joint detection scheme. Extensive experiments on two real-world water distribution networks demonstrate that our method significantly outperforms baseline approaches, achieving high-accuracy contamination identification and precise spatial localization—even under persistent sensor drift conditions.

Technology Category

• Constraint Satisfaction and Optimization: Distributed CSP/Optimization
• Data Mining & Knowledge Management: Anomaly/Outlier Detection
• Search and Optimization: Distributed Search

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Water Distribution Networks (WDNs) are vital infrastructures, and contamination poses serious public health risks. Harmful substances can interact with disinfectants like chlorine, making chlorine monitoring essential for detecting contaminants. However, chlorine sensors often become unreliable and require frequent calibration. This study introduces the Dual-Threshold Anomaly and Drift Detection (AD&DD) method, an unsupervised approach combining a dual-threshold drift detection mechanism with an LSTM-based Variational Autoencoder(LSTM-VAE) for real-time contamination detection. Tested on two realistic WDNs, AD&DD effectively identifies anomalies with sensor offsets as concept drift, and outperforms other methods. A proposed decentralized architecture enables accurate contamination detection and localization by deploying AD&DD on selected nodes.

Problem

• Water Quality Monitoring
• Chlorine Content Stability
• Real-time Pollution Detection

Innovation

• AD&DD
• Real-time Monitoring
• Water Quality Pollution Detection