Abhishek Balindra Karki completed Predictive Maintenance: Industrial ML for Fault Detection

Built a bearing fault detector end to end: engineered vibration features, used envelope-spectrum energy at the supplied defect frequencies to localize an outer-race fault, trained a compact depth-capped decision tree on training bearings only, and evaluated it on a held-out set with catch-rate and false-alarm metrics. Chose an operating threshold using explicit downtime-versus-inspection cost reasoning and wrote a maintenance memo grounded in the project’s own numbers.

• Leakage-Free Feature Engineering
• Honest Model Evaluation
• Cost-Based Decision Making
• Physics-Informed Fault Localization