| Recent advances in molecular biology such as cloning demonstrate that increasingly complex micromanipulation strategies for manipulating individual biological cells are required. In this thesis, an autonomous microrobotic system capable of manipulating individual biological cells using real-time vision is developed, and the micromanipulation of deformable biological cells is investigated. To obtain force feedback during microrobotic cell manipulation, a MEMS (Micro ElectroMechanical Systems) multi-axis cellular force sensor is developed and integrated into a microrobotic cell manipulation system. Aided by the cellular force sensing system, quantitative relationships between applied forces and biomembrane deformations have been established on mouse oocyte and embryo Zona Pellucida (ZP), which quantitates ZP hardening in post fertilization. To characterize the mechanical properties of biomembranes, a biomembrane point-load mechanical model is constructed. This model provides information on mouse ZP material properties for mouse ICSI (IntraCytoplasmic Sperm Injection) studies, enables a microrobotic system to predict biomembrane geometry changes, and provides a foundation to explore the use of vision tracking of biomembrane deformation combined with biomembrane material properties to create a vision-based biomembrane force sensing system. To increase the microforce sensor dynamic range, sensor stiffness is modulated using a force balancing technique made possible by integrated electrostatic microactuators. Subsequently, electrostatic microactuator dynamic issues are investigated. Electrostatic microactuators have had a fundamental limitation in that the allowable travel range is always limited to one-third of the total gap between comb fingers. Travel beyond this allowable range results in “pull-in” instability. An active control algorithm is proposed to extend electrostatic microactuator motion range. The advantages of this algorithm to existing approaches are that it does not add further complexity to the microfabrication processes, does not require additional hardware, and extends the motion range to the furthest possible. A visual servoing system is developed for positioning electrostatic microactuators. Visually servoing electrostatic microactuators for micromanipulation has the advantage of requiring minimum system calibration. Despite the nonlinear characteristics exhibited by the electrostatic microactuator, the proposed visual servoing framework allows the device to be positioned with nanometer precision along two axes. |
