Microfluidics is the study of the manipulation of fluids in channels of the size of a few tens of micrometres (Whitesides, 2006). Droplet microfluidics is a branch of microfluidics that deals with generating and manipulating droplets in a continuous-phase medium flowing through an engineered microchannel. Droplet microfluidics finds applications in chemical reactions, particle and biomolecule synthesis, cytometry, drug discovery, disease diagnostics etc. These practical applications require manipulation and control of unit operations involving the droplets, such as mixing, synchronization, and encapsulation in a lab-on-chip device.

Any experimental work involving a high throughput of droplets in a device requires video processing methodologies to obtain valuable data from the system. As the first step of this work, a generic program was developed to detect and track the motion of droplets in any geometry of microfluidic channels. The program's reliability was tested using videos from experiments on a square-shaped 2D millifluidic device made of Polydimethylsiloxane, wherein droplets form. various spatio-temporal patterns.

Unlike 2D geometries, a 1D microfluidic network consists of several interconnected flow channels(Arun Sankar EM, 2021). The flow of droplets through these networks is more complicated than that of just a single-phase flow. Droplets flowing through a network choose a branch whenever they reach a node interconnecting two or more channels. In this work, we tried to use a network model(Schindler and Ajdari, 2008) to experimentally demonstrate the decision made by drops in a loop device, which is the simplest form of a network. Experiments were conducted on a millifluidic loop device with aqueous droplets as the discrete phase and silicone oil as the continuous phase. We observed that the droplets make different decisions for the same droplet configuration and topology over time. A simulation was developed based on the existing model for the same topology and experiment conditions. The simulation was modified to replicate the consecutive decisions of droplets in the experiments. The experiment and simulation data were compared, and the errors were tabulated. The events leading to such errors were analysed, and methods to avoid such errors for practical applications requiring precise manipulation of droplet decisions were proposed.