Marvellous completed Predictive Maintenance: Industrial ML for Fault Detection

Built an end-to-end rolling-bearing fault detector from vibration data: engineered time-domain and envelope-spectrum features, localized the fault to the outer race via BPFO, trained a compact depth-capped decision tree, and evaluated it on held-out bearings with a confusion matrix, catch rate, and false-alarm rate. The candidate also justified an alarm threshold with a cost-based maintenance memo using their own measured results and a clear embedded-compute rationale.

• Leakage-Free Feature Engineering
• Physics-Based Fault Localization
• Honest Held-Out Evaluation
• Cost-Based Maintenance Decisioning