1 Introduction

SoCs that will embed sensors and actuators, regardless of their energy domain, are still more a concept than a reality. Nevertheless this concept can soon lead to very powerful microsystems. Just imagine a chip having MEMS parts, interacting with chemical/optical/biological micro entities. Examples of these may include tiny devices that can sense and react intelligently with respect to the environment, being highly autonomous by harvesting surrounding energy, and communicating with nearby microsystems or a central computer, via radiant or wire links.

Adding new types of components to the microelectronics design flow demands new Computer-Aided Design (CAD) tools, not to mention that the validation of the overall device can prove to be a very difficult task. Fabrication of such devices may soon be made possible in semiconductor foundries by adding new material layers in their standard CMOS processes. But new types of functionality, based on new material layers, and the interaction with the environment as well, can dramatically reduce the reliability and the safety levels. Self-test and fault tolerance will be compulsory. The way in which MEMS parts embedded in the new generation of systems will be tested is not yet known and research in this direction had already started several years ago (Kolpekwar et al., 1997). As in microelectronics, the development of cost-effective tests for large systems embedding MEMS may require test stimuli targeting actual faults and defects, developing fault lists and fault models for...