Novel MEMS tactile sensors for In-Vivo tissue characterization measurements.

Abstract

This dissertation focuses on developing a tactile sensor system for the measurement of contact force and elasticity. By using the developed sensors, the elasticity of various objects (e.g. tissue) can be measured by simply touching the targeted object with the sensor. In its most basic form, each developed sensor consists of a pair of contact elements that have different values of stiffness. The relative deformations of the two sensing components during contact can be used to calculate the elasticity parameters of Young's modulus or shear modulus. Analytical formulations that validate the proposed sensing technique are presented followed by sensor fabrication and experimental evaluation. Several prototypes of tactile sensors are fabricated through various MEMS processes. The first generation of prototype sensor is built with a surface micromachining process. Silicon nitride is selected as the structural material. With this fabrication process, the sensor can be fabricated down to a size of 1mm x 1mm x 500µm. A second generation of prototypes is developed through a polymer MEMS process. This type of sensor has a favorable flexible structure, which enables the sensor to be integrated on end-effectors for robotic or biomedical applications. Further, the sensing principle can be extended to measure shear force and shear elasticity. This extended ability with the polymer sensor is obtained by a design that includes a quad-electrode structure in each sensing cell. Finally, an easily fabricated tactile sensor of larger size is developed and attached on a touch probe. This prototype of sensor can provide reliable elasticity measurement in a handheld operating mode. Along with the advancement of tactile sensors, an estimation algorithm for the hand-held device, which employs a recursive least square method with adaptive forgetting factors, has also been developed. Experimental results show that this sensor can differentiate between a variety of rubber specimens. Further, the sensors have been characterized by fresh porcine tissue specimens and show the potential to provide reliable in-vivo measurement of tissue elasticity. The dissertation starts with introducing the background and motivation for this effort. This is followed by the conceptual design and mathematical models of the sensing principle. Different generations of prototype sensors are then discussed in the following chapters. Characterization results with tissue specimens obtained with these sensors will then be presented. Finally, the dissertation concludes with a discussion and summary of the work.