Gearbox Pinion Tooth-Break Severity Data for Predictive Maintenance

An experimental dataset covering combinations of load, speed, and failure severity for spur gearboxes. The data, published by Antonio Pérez-Torres on figshare, is used to validate a dynamic principal component analysis methodology for early failure detection. It contains vibration signals characterized using condition indicators in the time and frequency domains.

Use Cases

• Training anomaly detection models based on vibration signal condition indicators.
• Benchmarking multivariate statistical process control methods for gearbox diagnostics.
• Studying the relationship between operational parameters (load, speed) and progressive gear degradation.
• Developing early-stage failure detection algorithms based on Hotelling’s T² statistic and squared prediction error.

Strengths

• Data is empirically validated through experimental combinations of load, speed, and failure severity.
• The methodology demonstrated high sensitivity to progressive degradation and accurate early-stage detection.
• Available under a permissive CC-BY-4.0 license for reuse.

Limitations

• The dataset is very small at 5.5 KB, indicating limited scope.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.

Provenance

Source

Antonio Pérez-Torres via figshare.

Collection Method

Experimental dataset from vibration signal analysis of spur gearboxes.

Freshness

Last updated 2026-05-18 17:46:27; freshness should be verified.