An enhanced method for water end-use disaggregation and classification based on 1-min resolution data
“Population growth, urbanization, and climate change are currently increasing water stress in large areas worldwide and undermining the availability of water resources . Efficient and effective management of water systems is therefore essential to face the challenges posed by water scarcity. In this context, the adoption of smart water meters allows the understanding of water-consumption profiles at a remarkably high level of spatiotemporal detail, supporting water utilities in water infrastructure management and providing helpful feedback to users . In addition, in-depth information about residential end uses of water can be obtained by disaggregating and classifying smart-meter data collected at the household level and with sufficiently fine temporal resolution. However, most of the developed disaggregation and classification methods (e.g. [3-5]) can automatically process only high-resolution data (i.e. 1-10 s), whereas methods relying on medium-resolution data (i.e. 1 min) have only been tested with synthetically generated end-use data  or with very limited end-use datasets collected in the field [7-8].
In this work, an enhanced version of the rule-based, automated methodology for end-use disaggregation and classification originally developed in  is presented and applied to real-world end-use water consumption data collected at 14 Italian and Dutch households. In greater detail, the method exploits household-level water consumption data at the 1-min resolution, which is in line with that of most commercial smart meters. Automated end-use disaggregation and classification are then performed considering one end-use category at a time, through the adoption of Euclidean-distance-based metrics evaluating the deviation between the characteristics of each water-use event detected at the household level and the average features of each end-use category.”
(Citation: Mazzoni, F., Blokker, E.J.M., et.al. – An enhanced method for water end-use disaggregation and classification based on 1-min resolution data – Mazzoni, F., Blokker, E.J.M., et.al. – An enhanced method for water end-use disaggregation and classification based on 1-min resolution data – 19th International Computing & Control for the Water Industry Conference, 4-7 September 2023), 4-7 September 2023)