Due to far-reaching computerisation, smart metering, the use of (online) sensor systems in drinking water production companies and distribution networks, and automation, the quantity of ‘water’ data will grow inexorably into the future. The urgent need to distil information from these data flows more effectively was expressed by companies participating in a workshop in 2013. Scientific developments in the area of big data (data analytics and knowledge discovery techniques) offer the potential for more effectively distilling information with a view to improved operational (utility) management.
The idea is not only to measure the information in a water network with sensors, but also to analyse it with big data techniques. The translation of much of the data into useful information makes proactive operations possible, as opposed to the current reactive approach. In this TKI project, real-time intelligence is developed for the detection of anomalies in a data stream caused by a leakage or other disturbance.
KWR draws on its knowledge to develop and test algorithms using data-driven techniques. These are capable, on the basis of sensor signals in the Vitens Innovation Playground (VIP), of distinguishing between an anomaly and a regular pattern. As the end-user, Vitens supplies the data from its VIP. The ICT consultancy, HydroLogic, develops a demo to visualise the results on a platform. PWN and WLN further strengthen the consortium through their advisory input.
A data-driven tool which, on the basis of sensor signals from the distribution network, detects prospective anomalies that can lead to failures.