The evolution and widespread of MEMS (Micro Electro Mechanical Systems) devices over the last decades created a new field of study and leveraged the creation of whole new applications and businesses at a worldwide level. The improvements on the manufacturing precision and reliability, allowed some of these devices to be used in mission-critical equipment, such as in airbag systems. The relative low price made it available to home consumer markets, resulting in a paradigm-shift in many application areas.

The test and characterization of each individual manufactured MEMS structure is of major importance, both from the economic and quality standpoints. The dissimilarity of MEMS devices applications and physical principles creates an obstacle, since no single system can fit all the devices singularities, as opposite to what occurred in the Integrated Circuits (IC) industry. Also, new studies have sprung over the recent years that improve our understanding of the underlying fundamental physical phenomena and characteristics of the microsystems.

The Design for Test (DFT) paradigm is becoming an increasing trend and is placing more pressure on the designers to overcome the test costs and reduce the time-tomarket by applying embedded mechanisms fully dedicated for testing purposes. This thesis attempts to contribute to the maturity of MEMS testing and control systems development by proposing a flexible digital architecture for real-time control and analysis of structure movements. By the development of an hardware assisted processing platform with basic actuation and sensing mechanisms, realtime and determinism are guaranteed by design to ensure a flexible and adaptable system.

The proposed architecture is evaluated on two applications: First, a fast characterization platform is developed over the base system to enable a time-wise full-wafer characterization, where the speed and accuracy are optimized and conclusions about manufacturing issues are evaluated. The second application is based on a multi-order sigma-delta closed loop operation, where configurability and feedback delay are critical to allow a high-level algorithm to operate and perform an automated parameter optimization.

In both applications, the major issues are identified and a change in the base platform is performed to fit to each applications goals. In each case, several structures were tested, the results were evaluated and some improvements over the current state of-the-art methodologies and timings were achieved.