Fatigue Detection: Can Fatigue Detection Devices Predict the Driving Performance of Sleep-Deprived Drivers?

Principal Investigator:

John Bloomfield, Research Associate, College of Design


Project Summary:

Fatigue is a subjective state, but is clearly related to sleep deprivation. This study examined the relationship between sleep deprivation and driving performance. The objective of this study was to determine whether or not there are correlations between sleep deprivation, driving performance, and various potential measures of impairment.

Each of twenty commercial motor vehicle (CMV) drivers participated in a single twenty-hour experimental session, during which they were continuously kept awake, but were allowed to ingest caffeine and use tobacco as they would in real-world conditions. Each participant drove in a fixed-base advanced driving simulator for approximately one hour on four occasions (at 9:00am, 3:00pm, 9:00pm, and 3:00am). The 59.5-mile test route was designed with overpasses, intersections, and changes in speed limits, in order to make the driving experience more like real-world driving. (Participants were driven to the University of Minnesota's General Clinical Research Center after the experiment so they could sleep before driving themselves home.)

The main result was that the steering performance of CMV drivers was impaired when they stayed awake for an extended period: There was a considerable increase in steering instability between the morning drive at 9:00am and the nighttime drive at 3:00pm, an increase likely to have been produced by sleep deprivation.

Other results were: (1) stopping behavior improved throughout the session-suggesting practice effects; (2) after the fourth drive, there was less reduction in the participants' pupil size, but since there was no difference in pupil size before the fourth drive, there was no evidence to suggest that pupil size reductions could be used to predict sleep deprivation; and (3) data from other visual performance tests.


Project Details:

  • Start date: 12/2001
  • Project Status: Completed
  • Research Area: Transportation Safety and Traffic Flow
  • Topics: Intelligent Vehicles, Safety