Measurement and Estimation of Sleep in Railroad Worker Employees

Fatigue risk management systems provide a means to plan for and manage fatigue in round-the-clock operations such as railroading. Biomathematical fatigue models predict opportunities for sleep associated with a work schedule. The accuracy of the predictions depends, in part, upon the accuracy of the sleep estimation. The purpose of this study was to validate the sleep estimation methods used in the Sleep, Activity, Fatigue, and Task Effectiveness (SAFTE) model as implemented in the Fatigue Avoidance Scheduling Tool (FAST). The AutoSleep algorithm incorporated in FAST estimates sleep. The results of predictions from FAST were compared with actual sleep data as recorded by four groups of railroad workers in daily logs over a 2-week period. AutoSleep underestimated sleep for all groups of day railroad workers; however, for night workers, it overestimated sleep for night dispatchers but underestimated sleep for night train and engine service employees. Overall agreement ranged from 92 percent for signalmen to 79 percent for night dispatchers. FAST also provides a measure of effectiveness for each half hour of a work period. Although the efficiency predictions based on AutoSleep estimates of sleep differed from those based on the logbook records, the two estimates did not differ substantially. These findings validate the AutoSleep algorithm as a reasonable method to estimate sleep based on work histories when applying a biomathematical fatigue model such as SAFTE.

• Record URL:
  https://railroads.dot.gov/elibrary/measurement-and-estimation-sleep-railroad-worker-employees
• 
• Record URL:
  https://rosap.ntl.bts.gov/view/dot/61305
• Corporate Authors:

  Institutes for Behavior Resources

  2104 Maryland Avenue
  Baltimore, MD  United States  21218

  QinetiQ North America, Technology Solutions Group

  350 Second Avenue
  Waltham, MA  United States  02451-1196

  Federal Railroad Administration

  Office of Research and Development, 1200 New Jersey Avenue, SE
  Washington, DC  United States  20590
• Authors:
  • Hursh, Steven
  • Gertler, Judith
  • Raslear, Thomas
• Publication Date: 2011-2

Language

• English

Media Info

• Media Type: Digital/other
• Features: Figures; References; Tables;
• Pagination: 4p
• Serial:
  • Research Results
  • Publisher: Federal Railroad Administration
• Publication flags:

  Open Access (libre)

Subject/Index Terms

• TRT Terms: Algorithms; Employees; Estimation theory; Fatigue (Physiological condition); Mathematical prediction; Measurement; Railroad safety; Railroads; Sleep; Validation
• Uncontrolled Terms: Fatigue models
• Subject Areas: Railroads; Safety and Human Factors; Society; I83: Accidents and the Human Factor;

Filing Info

• Accession Number: 01475706
• Record Type: Publication
• Files: NTL, TRIS
• Created Date: Mar 15 2013 9:47AM