Stochastic Methods in Microsystems Engineering

Introduction

Microsystems Engineering is a growing field, that can be described as the ensemble of techniques, technologies, and practices involved in the realization of structures and devices with dimensions of the order of micrometers (Banks. D., 2006). Microsystems Engineering draws its roots from the microelectronics industry, but as the field evolved, the technological framework has been adapted and modified, in order to allow the development of new innovative devices, with unprecedented functionalities.

Since the arising of the micro-engineering domain, thousands of researchers have spent a lot of time developing new components design or exploring basic functionalities, such as piezoresistive, capacitive and vibration sensing schemes, in the case of micro-sensor devices or electrostatic driving in the case of micro-actuators. Microsystem devices, also known as MEMS (“Micro Electro Mechanical Systems”) devices that have been built to date are highly specialized to a particular application, and are realized in custom technologies that may not be widely available. The progress in microsystem development has been limited by the absence of standard commercial components. Most of today’s devices are application specific, which requires that unique components must be developed for each application with their own specific requirements. In the present, it does not exist a standardized cost-efficient technology for these devices (Gianchandani Y. B., 2008, Fair R. B., 2007).

Standardization of fabrication technologies implies definitions of standards on manufacturing tolerances, especially for the new polymer micro-manufacturing techniques, that have not yet been fully established, while most of the data is ‘in-house’ determined (Qin Y., 2010). Design for manufacturability and fault tolerance are expected to emerge as compulsory design considerations (Chakrabarty K. & Zeng J., 2014). Due to the multiple energy domains employed and mixed technology used for developing MEMS devices, they exhibit unique failure mechanisms and defects as compared to microelectronic systems.

In the recent years, fault tolerance analyses and design for manufacturability methodologies for MEMS have become a major concern in the scientific community as well as in the industry. However, from this perspective, the research has focused mainly on the development of experimental techniques for reliability testing of MEMS devices. The fabrication tolerances lead to variations of the geometrical parameters of the fabricated device, which leads to variations of the performance parameters. These variations in performance need to be quantified and controlled. Most of the techniques developed up to date for monitoring the effects of the deviations of geometrical parameters on the behaviour of microsystems are based on experimental methods (Chakrabarty K., 2014, Frangi A., 2008). The experimental testing of the device is not always convenient, since many applications like bio-medicine, pharmaceutics or environmental monitoring require the development of sterile or disposable devices and physical testing would imply the device contamination. In other applications like biophysics or biochemistry, functionalised surfaces are needed and the experimental testing would imply the damaging of these surfaces. Another approach has been to integrate testing mechanisms in the device (like optical sensors, capacitance measurement electrodes, impedance sensors, etc). This approach brings along two main disadvantages: the design complexity and implicitly the cost of the device are increased and on the other hand, there are needed supplementary fabrication steps, which are not always compatible with the materials used for manufacturing the MEMS device.