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

University  University of Nebraska-Lincoln (UNL)
Principal Investigators  Hamid Sharif, PhD., Electrical and Computer Engineering (PI)
Micheal Hempel PhD., Electrical and Computer Engineering (Co-PI)
PI Contact Information  PKI 200C
Office (402) 554-3628
Funding Source(s) and Amounts Provided (by each agency or organization)  UTCRS (USDOT UTC Program): $94,808
Union Pacific Railroad: $100,000
Advanced TEL Lab: $22,000
Total Project Cost  $216,808
Agency ID or Contract Number  DTRT13-G-UTC59
Start and End Dates October 2016 - June 2018
Brief Description of Research Project  With this project, we investigated the use of Big Data Analytics to make rail transportation safer, by preventing derailments due to equipment failure. Railroads typically schedule railcar maintenance on best-practice intervals, which may not include the plethora of information available from their maintenance logs, track data, sensors information, bills of lading, manufacturer history, etc. This project explored the use of this data to adapt maintenance scheduling to reduce cost and increase safety. We showed the great potential inherent in this approach.
Keywords big data, data fusion, failure prediction, maintenance schedule optimization, machine learning, derailment prevention 
Describe Implementation of Research Outcomes (or why not implemented) Place Any Photos Here See Report.
Impacts/Benefits of Implementation (actual, not anticipated) See Report.
Report http://www.utrgv.edu/railwaysafety/_files/documents/research/operations/utcrs_sharif_unifying-railcar-monitoring-sensor-data_final-report.pdf
Project Website http://www.utrgv.edu/railwaysafety/research/operations/rail-equipment-safety/index.htm