Predictive Maintenance Scheduling for Student-Operated Lathes in Engineering Education Workshops: A Vibration-Feature Degradation Model Linking Bearing-Wear Thresholds to Unplanned Downtime Reduction

Student-operated lathes in engineering education workshops fail unpredictably, and maintenance is typically reactive. The model treats the root-mean-square of the vibration velocity in the high-frequency band as a monotone degradation signal, defines a warning threshold from the ISO 10816 vibration severity zones, and derives the expected number of unplanned failures per lathe as a function of the inspection interval. A design and power analysis was conducted for the BACI estimator under a declared variance structure (between-cluster SD = 6.0, residual SD = 4.5, intraclass correlation = 0.64), with 2000 replications per design cell across 8, 16, 24 and 32 clusters and effect sizes of 3, 5 and 8 percentage points. The analysis indicates that 54 clusters are required to detect a 5-percentage-point reduction in unplanned downtime at 80% power. The binding constraint on predictive maintenance in educational workshops is not the degradation model but the cluster-level design of the evaluation, and that workshop networks should be sized accordingly before scheduling is deployed.