Unifying Railcar Monitoring Sensor Data, Maintenance Records, and Railcar Usage Information through Big Data Processing for Optimizing Railcar Maintenance and Safety

This proposed research effort addresses a timely and urgent need in transportation safety: preventing costly and devastating derailments through optimized equipment maintenance using Big Data Analytics. Safety continues to be of primary concern within the North American railroad industry, highlighted by efforts in freight train Wireless Sensor Network monitoring and Positive Train Control (PTC). Despite these efforts, statistics by the Federal Railroad Administration (FRA) Office of Safety Analysis show that from 2010 through 2015 over 1,000 derailments occurred directly linked to rolling stock equipment failure, causing over $240 million in losses. Current methods for equipment maintenance rely on fixed schedules, which either are too frequent and result in unnecessary operational expenses, or are not frequent enough and result in high equipment failure rates. Despite producing detailed records for all maintenance efforts, incidents, etc., this data remains largely under utilized in the optimization of operational processes such as maintenance scheduling, supplier quality ranking, parts optimization based on past component failures, etc. Optimization such as this is made possible through Big Data Analytics. However, the particular nature of the railroad application, combined with the myriad different report formats poses significant challenges to current data analytics approaches. The project team proposes to address the various research challenges that currently prevent Big Data Analytics, including data acquisition from handwritten records or incomplete reports, data normalization for proper significance assignment, forecasting of component failures and for optimized maintenance scheduling, multi-variate hyperdimensional clustering for trend and causality analyses to analyze supplier reliability, the impact of cargo types and routes traveled on failure probabilities, and so much more. The project will research all required methodologies and demonstrate Big Data Analytics capabilities using synthetic or real-world data provided by Union Pacific. The project manger believes that this approach is vital in further enhancing railroad operational safety and prevent derailments and the resulting significant monetary and environmental damages.

Language

  • English

Project

  • Status: Active
  • Funding: $216,808
  • Contract Numbers:

    DTRT13-G-UTC59

  • Sponsor Organizations:

    Nebraska Department of Roads

    1500 Highway 2
    Lincoln, Nebraska  United States  68509-4759

    Advanced Transportation Technology Institute

    1617B Wilcox Boulevard
    Chattanooga, TN  United States  37406

    Research and Innovative Technology Administration

    University Transportation Centers Program
    1200 New Jersey Avenue, SE
    Washington, DC  United States  20590
  • Managing Organizations:

    University of Nebraska, Lincoln

    College of Engineering and Technology
    Lincoln, NE  United States  68503
  • Performing Organizations:

    University of Nebraska, Lincoln

    College of Engineering and Technology
    Lincoln, NE  United States  68503
  • Principal Investigators:

    Sharif, Hamid

    Hempel, Michael

  • Start Date: 20161001
  • Expected Completion Date: 20160630
  • Actual Completion Date: 0
  • USDOT Program: University Transportation Centers
  • USDOT Program: University Transportation Centers Program

Subject/Index Terms

Filing Info

  • Accession Number: 01616226
  • Record Type: Research project
  • Source Agency: University Transportation Center for Railway Safety
  • Contract Numbers: DTRT13-G-UTC59
  • Files: UTC, RiP
  • Created Date: Nov 11 2016 12:00PM