- » Projects
|2003 - 2005
FloodNet: Pervasive Computing in the Environment: The utilisation of intelligent sensors linked to a hydraulic model to optimise data gathering and prediction in flood monitoring
the Geodata Institute project manages the FloodNet project as well as providing inputs to the Hydraulic modelling and GIS data management and visualisation.
The project primary goal is to demonstrate a methodology whereby a set of sensors monitoring the river and functional floodplain environment at a particular location are connected by wireless links to other nodes to provide an "intelligent" sensor network.
Data will be utilised in a predictive hydraulic model which will use real-time data to provide accurate flood level predictions on the floodplain up to 6 hours in advance (PhD student Jeff Neal).
This arrangement is intelligent in the sense that the behaviour of individual devices adapts dynamically to prevailing circumstances, both environmental (e.g. increased sample frequency with increasing likelihood of flooding) and infrastructural (e.g. synchronization of sensor hibernation to conserve battery life, adaptation of network routing around different failure modes of nodes and links).
The intelligence and adaptive sampling is an essential and innovative component of the project, since it permits local sensor "experience" of the flood event and optimises the sampling and reporting rate to suit the environmental conditions.
Data is streamed to a GIS graphics and visualisation output developed and maintained at the GeoData Institute