On the design of an intelligent sensor network for flash flood monitoring, diagnosis and management in urban areas

We propose an intelligent sensor system based on a new sensing methodology, relying also on 3D map reconstruction techniques, for computing with high precision, in real-time and without human intervention the parameters needed for stream-flow computation: water levels, morphology of the streams of all potentially flooded areas by each controlled stream. The collected data will be continuously transmitted, through a communication infrastructure, to software agents designed to compute the stream-flow and to quantify the spatial distribution of flood risk for each controlled watershed. The computed risks, together with other data coming from other sources (barometric sensors, camera operators of public organizations, emergency agencies, private citizens), will be analyzed by a diagnostic decision system implementing a risk-alert scheduling strategy. This system will be able to diagnose the health state of the controlled environment and to define specialized alarm levels for each potentially interested area that will be used to alert all citizens (ubiquity) locally present (alerting locality).