Machine learning technologies for predicting and preventing production failures

This paper analyzes the use of machine learning technologies to predict and prevent production failures in the electrochromic field.It specifically focuses on the cycling process of lite ids.Various issues such as bad reading of windows controllers, high voltage explosions, and voltage gaps during cycling have been identified and resolved through the efforts of a subject matter expert, Viralsinh.By implementing changes to the cycling conditions, including increasing accuracy in tint drive, hold voltage, and number of cycles, the validation rate has increased significantly, resulting in a reduction of re-works and cost savings for the company.Automation is identified as a key area for improvement, with the implementation of an enterprise system leading to significant time-savings in quality checks and disposition decisions for lite ids as well.