The impact of sensor placement on graph-neural-network-based leakage detection

TL;DR

This paper examines how sensor placement affects the effectiveness of graph neural networks in detecting leaks in water networks, proposing a new placement method based on PageRank centrality.

Contribution

It introduces a novel sensor placement strategy using PageRank centrality and evaluates its impact on GNN-based leakage detection performance.

Findings

Sensor placement significantly influences GNN detection accuracy.

PageRank-Centrality-based placement improves leak detection performance.

The proposed method outperforms existing placement strategies.

Abstract

Sensor placement for leakage detection in water distribution networks is an important and practical challenge for water utilities. Recent work has shown that graph neural networks can estimate and predict pressures and detect leaks, but their performance strongly depends on the available sensor measurements and configurations. In this paper, we investigate how sensor placement influences the performance of GNN-based leakage detection. We propose a novel PageRank-Centrality-based sensor placement method and demonstrate that it substantially impacts reconstruction, prediction, and leakage detection on the EPANET Net1.