Model-based Microfluidic Device Design for Refreshable Braille and Tactile Graphic

A low-cost and reliable technology to support electronic braille displays would significantly increase access to digital media for blind braille readers. Even better, a full-page dense array of refreshable braille-sized dots could be used to display both braille and tactile graphics---a feat not achievable by existing technologies. The work in this dissertation is two-pronged. First, a perceptual study was carried out that motivates and guides the design of a full-page display. Second, a model-based design approach is adopted to develop microfluidic technology to overcome critical challenges in creating a full-page display.;A full-page display is made up of thousands of densely-packed dots, each requiring a dedicated actuator. Some design solutions have been proposed which reduce the number of actuators by using a small array that travels along with the reading finger (think: braille cell on a computer mouse), eliminating sliding contact as a consequence. Based on this consideration, a perceptual study was carried out that informs the design of a full-page display. The study quantifies the reduction in braille character recognition imposed by display designs that reduce sliding contact.;Microfluidic technology is employed to fabricate tiny channels and chambers in soft silicone that address and raise braille-sized bubble features on a surface. The challenge associated with controlling many individual fluidic features is met by integrating digital fluidic logic networks with the actuators. Each dot is addressed by the combination of a fluid actuator and a basic fluidic memory unit. A memory unit stores high or low pressure as a binary signal that determines whether its corresponding actuator is up or down. Crucially, memory units can be cascaded so the output of one controls the input of another. Pressure-encoded 1s and 0s can be shifted along cascaded memory units, thereby reducing the required number of external valves. To ensure that the design is sufficiently scalable and manufacturable to support the realization of a large dense array of pins at braille spacing, two performance criteria must be attained. The first is ensuring the memory units are cascadable. The second is controlling memory units at speeds suitable for refreshable braille. The work in this dissertation addresses these two performance criteria by developing models of digital fluidic circuits with particular attention to design parameters that affect circuit cascadability and speed.