Remaining useful life prediction for turbofan engines (NASA C-MAPSS)

Predict how many operating cycles an engine has left from multivariate sensor data, using the public NASA C-MAPSS dataset, and compare classical and deep-learning models.

Scope

Explore and clean the C-MAPSS sub-datasets; study how sensors degrade over time.
Build baselines (linear regression, random forest) and a sequence model (LSTM or 1D-CNN).
Evaluate with RMSE and the asymmetric scoring function used in the literature.

Milestones

Literature review and data exploration.
Feature engineering and baselines.
Deep-learning model and tuning.
Comparison, explainability (which sensors matter), and a simple dashboard.

Extension

Apply the same pipeline to vibration data from a lab motor or pump.