Unplanned industrial equipment failures cause significant financial losses andsafety hazards across aviation, energy, and manufacturing sectors. Predictive Maintenance (PdM) addresses this by estimating the Remaining Useful Life (RUL) — the numberof operational cycles before a machine requires intervention — enabling proactive, costoptimal scheduling. This paper presents a complete end-to-end machine learning pipelinefor RUL prediction on the NASA C-MAPSS turbofan engine degradation dataset 17. Wecompare four models — Linear Regression, Random Forest, Gradient Boosting, and a twolayer Long Short-Term Memory (LSTM) network implemented in pure NumPy underrigorous Group K-Fold cross-validation to eliminate engine-level data leakage. Our featureengineering pipeline transforms 21 raw sensor channels into 37 temporal features includingrolling statistics, a composite health index, and degradation rate. Random Forestachieves the best performance: RMSE = 9. 95 cycles, R2 = 0. 9398, MAE = 5. 77, and NASA PHM asymmetric score = 1. 9. We further contribute bootstrap predictionintervals achieving 90. 0% empirical coverage at the 95% nominal level, and a maintenancescheduling module projecting 430, 000 in cost savings on a 20-engine fleet.
Rohan Mahmudul Hasan Rohan (Fri,) studied this question.