AquaSCALE- Exploring Resilience of Community Water Systems Poster.pdf

AquaSCALE is a cyber-physical-human middleware for gathering, analyzing and adapting operations of increasingly failure-prone community water services. Today, detection of anomalous events in civil infrastructures (e.g. water pipe breaks and leaks) is time consuming and often takes hours or days. AquaSCALE leverages dynamic data from multiple information sources including IoT sensing data, geophysical data, human input, and simulation and modeling engines to accurately and quickly identify vulnerable spots in water networks. Such sensor-simulation-data integration platforms can assist in design time tasks (e.g. optimize IoT device placement), enable fault detection (e.g isolation of leaky pipes), trigger runtime adaptation mechanisms (e.g. control of valves) and improve resilience of the overall system and predict/reduce cascading impacts (e.g. flood).
AquaSCALE bridges the infrastructure/application gap by transforming input sensor data streams collected at lower base layer to semantic streams that capture application-level entities at higher service layer. This allows different classes of users to address different concerns. Given knowledge of the network structure, robust simulation methods using a commercial grade hydraulic simulator EPANET enhanced with the support for IoT sensors and failure modeling are used to generate profiles of anomalous events. These profiles are then trained by diverse plug-and-play machine learning strategies to rapidly isolate anomalies, and explore adaptation of network flows to mitigate impacts. To evaluate AquaSCALE, we developed two applications - leak event detection and flood event prediction. Our results indicate that phased mechanisms deployed in AquaSCALE can accurately locate a failure events at fine levels of granularity (individual pipeline level). Furthermore, the proposed two-phase approach with offline training and dynamic data integration reduces detection time by orders of magnitude.