Storage Device Quality Control and Supply Chain Management Using Deep Machine Learning Models

Explore a 50-minute conference talk from SNIA Storage Developer Conference 2024 that delves into innovative approaches for storage device quality control and supply chain management using machine learning techniques. Learn how storage sensor data analysis can predict component failures before they impact system operations, with detailed insights into building predictive maintenance models that balance detection rates, false positives, and lead times. Discover how HPE addressed supply chain challenges during the Covid-19 pandemic by developing dual prediction models - one for immediate failure detection and another for long-term supply chain management with extended lead times. Examine real-world deployment examples showing 70-80% detection accuracy for maintenance predictions and 90-95% accuracy for quarterly quantity forecasting. Master the implementation of machine learning feature selection processes for predicting failures in various electronic components including hard drives, solid state drives, and voltage regulators, while understanding the distinct approaches needed for immediate maintenance versus long-term supply chain planning.