A Review On Statistical Models For Anomaly Detection Using Metro Turnout Data

• Author(s):

  Ramavtar Singh Sikarwar | Dr. Sunil Sugandhi

• Keywords:

  Statistical Models, Metro Turnout, Imbalanced Datasets, Machine Learning, Classificaiton Error.

• Abstract:

  Metro Railway Systems Are Among The Most Important Modes Of Urban Transportation, Providing Safe, Reliable, And Efficient Mobility For Millions Of Passengers Every Day. One Of The Most Critical Components Of A Metro Railway Network Is The Turnout Or Switch System, Which Enables Trains To Change Tracks And Facilitates Smooth Traffic Management. Since Turnouts Are Subjected To Continuous Mechanical Stress And Environmental Variations, They Are Vulnerable To Wear, Degradation, And Unexpected Failures. Therefore, Timely Fault Detection In Metro Turnout Data Is Essential To Ensure Operational Safety, Minimize Service Disruptions, And Reduce Maintenance Costs.A Turnout System Consists Of Several Mechanical And Electrical Components, Including Switch Rails, Point Machines, Locking Mechanisms, Motors, And Sensors. During Operation, These Components Generate Various Forms Of Data Such As Current Signals, Voltage Measurements, Vibration Signatures, Temperature Readings, And Switching Times. The Analysis Of These Data Enables Maintenance Engineers To Assess The Health Condition Of The Turnout And Identify Abnormalities That May Indicate The Onset Of Faults. Effective Fault Detection Systems Can Provide Early Warnings And Prevent Catastrophic Failures. This Paper Presents A Review Of Existing Statistical Models In The Domain Of Research.

---

Other Details

• Paper id:

  IJSARTV12I6105663

• Published in:

  Volume: 12 Issue: 6 June 2026

• Publication Date:

  2026-06-10

Download Article