Quantum Machine Learning and Railway Deterioration and Operations

This study investigates the potential of Quantum Support Vector Machines (QSVMs) for predicting track geometry failures, a critical challenge in railway safety. Modern track inspection methods generate massive datasets exhibiting characteristics of big data. Classical machine learning techniques may struggle to efficiently analyze this data. To address this challenge, the research team explores QSVMs, leveraging the unique properties of quantum mechanics for potentially faster and more efficient data processing. The research compares the performance of different QSVM circuit layouts against a classical SVM using track geometry data. The results show that circular and shifted alternating circular QSVM circuits outperform the classical SVM in predicting track failures, achieving test accuracy exceeding 65%. While other circuits perform comparably to the classical SVM, all QSVM circuits exhibited superior training accuracy. These findings suggest that QSVMs can be a valuable tool for track maintenance, potentially exceeding the capabilities of traditional machine learning techniques. However, selecting the optimal QSVM approach requires careful consideration of circuit layout, depth, and available computational resources. Future research should focus on optimizing QSVM algorithms and exploring their broader applicability in railway safety as quantum computing technology advances.

Language

• English

Project

• Status: Active
• Funding: $50000
• Sponsor Organizations:

  Office of the Assistant Secretary for Research and Technology

  University Transportation Centers Program
  Department of Transportation
  Washington, DC  United States  20590
• Managing Organizations:

  Howard University

  2400 6th Street, NW
  Washington, DC  United States  20059
• Performing Organizations:

  University of Maryland, College Park

  Department of Civil and Environmental Engineering
  College Park, MD  United States  20742
• Principal Investigators:

  Attoh-Okine, Nii

• Start Date: 20231001
• Expected Completion Date: 0
• Actual Completion Date: 0
• USDOT Program: University Transportation Centers Program

Subject/Index Terms

• TRT Terms: Deterioration; Failure; Geometry; Machine learning; Railroad tracks
• Subject Areas: Maintenance and Preservation; Planning and Forecasting; Railroads;

Filing Info

• Accession Number: 01931534
• Record Type: Research project
• Source Agency: Research and Education in Promoting Safety (REPS) University Transportation Center
• Files: UTC, RIP
• Created Date: Sep 20 2024 9:17PM