Water Consumption Pattern Analysis Using Biclustering: When, Why and How

Sensors deployed within water distribution systems collect consumption data that enable the application of analysis techniques to extract essential information. Time series clustering has been traditionally applied for modeling end-user profiles aid management. However, its effectiveness is limited by diversity and local nature patterns. In addition, existing cannot adequately handle changes in household composition, disruptive events (e.g., vacations), dynamics at different time scales. this context, biclustering approaches provide a natural alternative detect groups end-users with coherent during periods while addressing aforementioned limitations. This work discusses when, why how apply analysis, further proposes methodology end. To best our knowledge, first introducing analysis. Results on from real-world system—Quinta do Lago, Portugal—confirm potentialities proposed approach pattern discovery guarantees statistical significance robustness entities can rely strategic planning.