Towards a global environmental sensor network

Large numbers of sensors are being deployed around the world to monitor the natural environment. By their nature these tend to record data for one specific application. However sophisticated higher level applications need access to data from these disparate systems and global monitoring certainly necessitates global data gathering. The reality of these sensor networks is that they range from home-grown to established commercial systems and their data is in an equality diverse variety of formats. The task of harvesting data thus relies on mapping any existing defacto data standards to a common set. This problem is analogous to the many which are being tackled by semantic web and GRID technologies. This paper will present an overview of how these modern approaches can be implemented and allow intelligent data harvesting of environmental data. It will draw on experiences from the Envisense centre for pervasive computing in the environment in the University of Southampton UK.