Customer quality control requirements are tightening at a rate that makes it difficult to keep up using standard quality practices. Traditional methods of producing to specification (certificate of analysis) and control limits no longer guarantee material performance. To address these challenges, EMD Electronics combines a data-driven approach with a physics-based expertise. This unique approach pairs machine learning with engineering to identify critical process and quality parameters to control for. In addition, predictive models and multivariate control charts are used to vet material batches before they are shipped to the customer. As a result, the time for task force resolution has been reduced from months to weeks and material quality has improved 125% at the customer´s application process. The resultant reduced supply risk and increased customer confidence have enabled greater collaboration and faster process change approvals. Learn how EMD Electronics has employed data analytics to achieve these results, and their vision for how data science will revolutionize quality control in the semiconductor industry.
