Predicting power problems: Contract condition-based maintenance at BNSF's new Alliance, Neb., Reliability Center is improving locomotive efficiency and utilization

This article describes preliminary findings three years into a 12-year arrangement between Burlington Northern Santa Fe (BNSF) Railway and Alstom Transport for maintenance on 474 freight locomotives at BNSF shops, many of which are outfitted with "Reliability Centers." The article relates how the railroad’s newest center is using a new predictive diagnostic known as condition-based maintenance (CBM). The centers are intended to identify locomotives that are overly prone to breakdowns and convert them into reliable machines. The threshold for qualifying for special attention is fewer than 20 to 30 days mean time between failures or a high defect rate measured per day. Rather than spend a set amount of time on each machine, the staff at the center are allowed to devote extra time on the weaker performing locomotives. The CBM system enables maintenance workers to predict part replacement or overhaul, reducing failures on the road and increasing overall reliability. The CBM processes at the new center are a little different because they use stress wave analysis to predict mechanical failures, integrate all CBM data into a single software database, and use the Core Competency Network to streamline diagnostics. The article describes the diagnostic tools, the strategy for evaluating recommendations, and how the technique can detect shortcomings in mechanical parts that are otherwise difficult to observe and measure. It also describes a technique called "fragmented overhaul," which eliminates the need for periodic block rebuilds. To date, experience supports the approach.


  • English

Media Info

  • Media Type: Print
  • Features: Figures;
  • Pagination: p 23
  • Serial:

Subject/Index Terms

Filing Info

  • Accession Number: 01010978
  • Record Type: Publication
  • Source Agency: UC Berkeley Transportation Library
  • Files: BTRIS, TRIS
  • Created Date: Nov 29 2005 2:36PM