Continuous-flow microfluidics has evolved very rapidly in the last twenty years. Biochemical applications can be performed in parallel and automatically on continuous-flow microfluidic chips, by which more precise results with higher throughput can be achieved. To date, most continuous-flow microfluidics are still designed manually, which is time-consuming, error-prone, and especially impractical for large-scale-integration. Design automation researchers started to develop design automation tools for continuous-flow microfluidics about ten years ago. Some of these are front-end tools, which scheduled biological operations from assay protocols and allocated these operations to microfluidic devices. Some others are back-end tools, which performed automatic device placement and micro-channel routing. As new on-chip components, e.g., sieve valves, and new architectures, e.g., a homogeneous valve-centered architecture, are continuously invented and adopted, design automation tools also evolve accordingly over time. This dissertation first provides a brief review of design automation research for continuous-flow microfluidics over the last decade. In this review, a detailed description is given for the microfluidic architecture and the general topics of front-end as well as back-end research in design automation field. Selected research results are then introduced according to their appearance over time, which also matches the general trend of research topics from front-end to back-end. The main body of the dissertation presents new research results that cover four topics in four chapters: temporary caching of fluids, sieve valve exploration, synthesis for reconfigurable chips, and the layout generation tool Columba.
«
Continuous-flow microfluidics has evolved very rapidly in the last twenty years. Biochemical applications can be performed in parallel and automatically on continuous-flow microfluidic chips, by which more precise results with higher throughput can be achieved. To date, most continuous-flow microfluidics are still designed manually, which is time-consuming, error-prone, and especially impractical for large-scale-integration. Design automation researchers started to develop design automation tool...
»