Lecture 8: AI for Failure & Fatigue Prediction

- Using time-series sensor data
- Features: peaks, cycles, RMS, stress/strain sequences
- Classical models vs ANN-based predictors
- Introduction to RUL (Remaining Useful Life) concepts
- Real-world challenges:
	- limited failure examples
	- imbalanced data
	- saftey requirements and false positives