Current approaches to flood monitoring (e.g. in river valleys) involve statically deploying depth and ultrasoundbased flow sensors across flood-prone areas, and feeding the collected data off-site (e.g. using GSM) to grid-based computational models which predict flooding trends. We believe that there is considerable scope for improvement in such scenarios. In particular, we are investigating selectively shifting the execution of prediction models to the sensor network itself, which thus acts as a 'mini-grid'. Computations organised in this way can be used not only to provide more timely flood warnings, but also to help to dynamically adapt the wireless sensor network (WSN) and thus optimise it for current or predicted environmental conditions. For example, the network can employ a low power, low throughput organisation in quiescent conditions, and switch to a high power, resilient, high throughput organisation when flooding is imminent. To achieve this vision we have developed a sensor network framework based on an appropriate combination of software and hardware.
ACM Copyright Notice: © ACM, 2006. This is the author's version of the work. It is posted here by permission of ACM for your personal use. Not for redistribution. The definitive version was published in the Proceedings of the 4th international Conference on Embedded Networked Sensor Systems (SenSys '06), held in Boulder, Colorado, USA, October 31 - November 03, 2006).complete documentdoi:http://doi.org/1182807.1182869 (publisher's link)