Melanie Senn

Title: Melanie Senn and Manufacturing Defect Detection

Introduction

Melanie Senn is an inventor associated with Sunnyvale, California, and Cepheid Corporation. Her patent work addresses the use of machine learning to identify product defects during manufacturing.

Latest Patents

Senn is named on one patent, 1 patent: “Transfer Learning Methods and Models Facilitating Defect Detection.” The invention describes methods and systems for training models to detect manufacturing defects by combining supervised transfer learning with supervised and unsupervised learning. It uses expert-labeled data for broader tasks, such as product classification, and transfers that learning to more specific tasks, including identifying product features and detecting anomalies. Further expert labeling and training can refine the model, while experimentally induced defects can help make training more efficient. The patent describes sample cartridges as one possible product, with the model using cartridge and manufacturing-process data to identify faulty units.

Career Highlights

Senn’s listed patent connects her work at Cepheid Corporation with machine-learning approaches to manufacturing quality control. The invention focuses on applying learned patterns and anomaly detection to help identify defective products.

Collaborations

Aristotelis Fotkatzikis is listed as a coworker associated with Senn.

Conclusion

Melanie Senn’s patent describes a transfer-learning approach to automated defect detection, with potential application to sample-cartridge manufacturing and quality assurance.