New trends in intelligent system design for embedded and measurement applications

Intelligent systems adopt soft-computing techniques (encompassing neural networks, fuzzy logic, genetic algorithms, and expert systems) to solve complex problems by mimicking human reasoning. On the other hand, conventional algorithmic approaches are extremely powerful and efficient in tackling applications for which a procedural solution can be easily defined. By themselves, each of these techniques may be the optimal solution for a subproblem, but not efficient enough to solve the problem as a whole. Composite systems, consisting of conventional and soft-computing components in cooperation, are now more than a promise to face complex application needs. In this article we present recent advances in the design of composite systems, with specific reference to embedded and measurement applications.