Leakage Detection in Waterpipes Networks using Acoustic Sensors and Identifying Codes

In future smart cities a grand challenge will be to sensorize large urban infrastructures at a feasible cost. In this paper we tackle the case of efficient leakage detection in water distribution systems. Deploying leakage detectors can cut operational costs for water utility providers. But, the cost for deploying them with sufficient granularity poses an high entrance barrier due to the scale of such infrastructures. We propose an algorithmic framework to efficiently deploy sensors in order to perform leakage/fault localization over large scale lattice-shaped networks. The novelty of our solution, combining covering sets and identifying codes is that it initially covers the network with low resolution, and thus fewer sensors. The set of sensors can then be extended in a way to progressively improve the resolution by which leakages are located. The proposed solution is validated through extensive numerical experiments.