Open Collections

AlSawalhi, Abdulla

University of British Columbia

Digital microfluidic devices provide precise control over droplets, enabling a wide range of applications. These applications span research fields such as DNA testing, cell research, and agri-food quality control. Although the applications vary, they all rely on conventional (manual) digital microfluidic operation, which introduces several challenges. Manual droplet control requires significant time and effort, limiting experimental throughput and scalability, particularly in complex multi-droplet experiments. Furthermore, the necessity of human intervention increases the likelihood of errors, thereby compromising the reliability of results. These limitations present significant barriers to the broader adoption of digital microfluidics in rapidly advancing applications and different environments. To address these challenges, this thesis proposes an algorithmic framework to achieve automation in digital microfluidic devices. The algorithmic framework makes use of object detection and A* pathfinding to automatically route and actuate droplets in digital microfluidic experiments. Additionally, the algorithmic framework is designed to be highly adaptable, allowing for seamless integration across various digital microfluidic device configurations. A graphical user interface is employed to allow for input control commands while also serving as a means of communicating information back to the user. Eight case-specific experimental scenarios are presented to demonstrate the functionality of the algorithmic framework in practical settings. This also serves to showcase the robustness of the framework in handling specific scenarios encountered in digital microfluidic experiments. To further enhance automation, a closed-loop mechanism is introduced to address unexpected droplet halting. This mechanism can verify droplet positions in real time, dynamically updating electrode states and recalculating optimal routes when necessary. Additionally, a sample volume estimator is developed to quantify droplet volumes using a calibration curve, providing a practical tool for ensuring accurate sample handling in digital microfluidic experiments. The proposed algorithmic framework is designed to significantly minimize manual effort while enhancing efficiency and experimental throughput via parallelization. By expanding the capabilities of existing digital microfluidic devices, the algorithmic framework facilitates the execution of complex experiments, which can drive progress in digital microfluidic applications across various research domains.

Text

2025-04-09

Attribution-NonCommercial-NoDerivatives 4.0 International

http://hdl.handle.net/2429/90618

Electrical Engineering

University of British Columbia

UBCO

http://creativecommons.org/licenses/by-nc-nd/4.0/