One of the most important tasks in modern automation systems is to provide secure and efficient supervision of the plant. In fact, most of the code in industrial control applications, like nuclear power stations, airplanes, cars, etc. deal with supervision, how to handle extreme cases, and how to diagnose certain faults.
The area requires an active combination of advanced signal processing techniques, logics and finite state automata, and data base methodology. It is thus very well suited for the ISIS constellation.
The goal of the ISIS activities will be to develop methodology to show how these combinations can be implemented most efficiently. There will be close connections to applications, e.g. for supervision of functions in car engine control, and navigation systems.
Initially, the project will concentrate on how to integrate rather basic signal processing techniques (residual computations) with logic schemes, so as to better isolate a main fault that has occured (and which can give rise to an outburst of many elementary alarms).