Extreme Ultraviolet Lithography is an essential step in producing all modern electronic devices and circuits such as the CPU/GPU in your computer and cellphone.
Lithography helps define complex circuit patterns by using optics to focus light onto a photosensitive silicon substrate.
Plasma droplets are vaporized during the lithography process in order to produce the high power EUV light source.
Our project focuses on using images of the plasma droplets to predict if the EUV light source performance will be satisfactory.
Using computer vision, deep learning, and optimization techniques, we can examine droplet images and metrics to evaluate and predict droplet performance.