Devraj Sangharajka completed Predictive Maintenance: Industrial ML for Fault Detection

Built an end-to-end rolling-bearing fault detector using vibration features, a compact decision tree, and held-out evaluation. The work identified kurtosis as the key impulsive fault signal, located the defect at the outer race via BPFO dominance, and defended a 0.38 alarm threshold with concrete catch-rate and false-alarm trade-offs for maintenance use.

• Leakage-Free Feature Engineering
• Honest Held-Out Evaluation
• Cost-Based Operational Decision-Making