AI-Driven Anomaly Detection in Component Burn-In & Screening

In high-reliability sectors (like space) electronic components undergo rigorous environmental stress screening (ESS), including Burn-In testing (operating components at elevated temperatures, e.g., 125°C for extended periods). Traditional screening relies on static parametric pass/fail limits. However, 'latent defects'—components that pass the absolute limits but exhibit subtle, anomalous drift over time—often escape into final payloads, leading to catastrophic field failures.

Development of a predictive machine learning model that analyzes time-series parametric data (e.g., standby current Iddq, leakage currents, or propagation delays measured at intervals like 0h, 24h, 96h, and 168h to detect anomalous components.

Module A: The outlier detection system Static limits catch obvious failures. Participants need to develop a 'Dynamic' outlier detection system. If a lot has an average leakage current of 10µA, a part showing 45 µA is a massive anomaly, even if the absolute datasheet maximum limit is 50 µA. Module B: Time-Series Drift Predictor Build a predictive regression model that takes Value_0h and Value_24h as inputs and forecasts Value_168h. If the predicted 168h drift rate exceeds a calculated safety slope, the system flags the component for early rejection. Evaluation Metrics: • Anomaly Detection Score: a False Negative (missing a defective part) is catastrophic, penalizing teams that let bad parts escape. • Drift Prediction Accuracy : The mean absolute error between the predicted Value_168h and the actual hidden ground-truth values. • Explainability : Can the model justify its classification to a QA inspector, or is it a complete black box?