Accelerometer Based in-situ monitoring of HDC Capsule Polishing

HDC polishing is an important aspect of the fabrica on process to reach the final thickness and surface finish of the capsule design.Challenges in polishing include both uniform defects such as surface roughness, sphericity, and final mass as well as isolated defects such as surface pits.The process of polishing shells currently requires large amounts of operator interven on to ensure that material removal rates are within normal levels and that the surface quality remains sa sfactory.The addi on of accelerometer sensors coupled with metrology data provides a unique opportunity to u lize machine learning techniques to determine acous c signals that can provide real-me process monitoring to enhance polishing speed and yield.This talk will give an overview of both experimental and analy cal techniques as well as preview results from domain adapta on machine learning models.