Real-time information from the 'real world' becomes essential to sustain effective decision making and control of vital processes and infrastructure in our lives. As such sensor networks will play an increasingly dominant role in our every day life.

Ubiquitous sensory networks capable to adopt to the ever increasing complexity of our environment, will soon become an indispensable part of the society. This vision is known as the sensor revolution.

Several technological developments offer unprecedented opportunities in the areas of sensing the physical world, processing and linking of information and processes, and taking - use - of information. The time between sensing - data acquisition - and the demand for information based on the current "measured" values in the physical world, moves in big steps over to real-time services.

It is conceivable that in the near future information is requested that was not expected at that moment, but nevertheless can be provided by the self-adjusting sensor network.

The dynamic environment in which these sensors and sensor networks are located is complex and unpredictable as the system is in the 'real world'. Also, the sensor systems have to be available for multiple users and adaptive to automatically adapt to the demands of users and because of that optimally meet the information needs.

Adaptive, Scalable & Multiuse

Effectively interpret observations for decision imposes requirements on the process of collecting, processing and selected by the system. The requirements are:

Time-critical; vital for the usefulness of information depends on when this is 'delivered'.

Multi-use; multi-purpose users and a "must" in large-scale sensor systems. At the same time, and changing situations requires organisation and self-optimization to conflicting requests and resource management to settle.

Scalability; miniaturization and cost reduction of sensors force up the number of sensors and processing elements in the network and requires self-steering solutions. Especially in case of failure.

Demand-driven; as the future situation is still unknown, the information is still developing, the system real-time adaptive behaviour data is always optimal to process information for the new user demand.