2025 Volume 29 Issue 1 Pages 85-92
Labor-saving in equipment maintenance and management is an urgent issue for regional railways. Rail bonds, which ensure electrical connections between rails, are installed in large numbers and require frequent inspections. Therefore, rationalizing their management based on actual conditions is essential for improving efficiency.However, due to the ledger management unit and missing data, it has been difficult to accurately assess anomaly frequencies using the general exponential hazard model. In this study, we propose a method for constructing an integrated dataset that combines inspection and ledger data from both the electrical system, which manages rail bonds, and the track maintenance system. By applying this approach to 2, 092 rail bonds on actual railway lines, we empirically demonstrate that cross-system data integration significantly improves the accuracy of anomaly frequency analysis, even when using the same exponential hazard model.
