Driver fatigue is recognized as a major causal factor in accidents involving long-haul commercial drivers. The development of an on-board driver performance/fatigue monitoring system could potentially assist drivers in identifying the onset of fatigue. Such a system would monitor drivers on their ability to perform this complex, visually, intensive manual-control tracking task for which, in addition to keeping the vehicle within the confines of the roadway, drivers must continuously avoid collisions with other vehicles, pedestrians, and other impediments that may lie in their path of travel.
The objective of the research described herein was to determine whether self-assessment of fatigue and temporal vehicle separation were reliable and valid indicators of driver fatigue in an operational setting. Such a determination will serve to aid future researchers in determining whether these metrics accurately reflect the onset of driver fatigue. The results from this study could be applied toward developing guidelines for increasing driver awareness of fatigue and max contribute to the development of driver monitoring devices and warning systems based on self-assessment of fatigue.
Fatigue Measurement: An Overview
Although the concept of fatigue is fairly easily understood, the scientific community has yet to reach a universally agreed-upon definition. Because fatigue cannot be measured directly, it is often defined operationally in terms of its observable symptoms. Driver fatigue has been measured in a variety of ways corresponding to four categories of observable variables: performance, physiological, opinion (subjective), and behavioral or cognitive measures (Hardee, Dingus, & Wierwille, 1986; Wierwille & Muto, 1981).
Measures of direct vehicle control offer the most intuitive and least intrusive methods for detecting the onset of fatigue. A major limiting factor of performance-based measures, however; is that a decline in performance capacity may in fact occur prior to changes in driver performance (Dinges & Graeber, as cited in Williamson, Feyer, Friswell, & Leslie, 1994). This phenomenon can be attributed to driver skill and the ability of more experienced drivers to compensate during a routine driving task, despite their diminished capacity. Despite this shortcoming, performance measures (velocity maintenance, steering measures, and unplanned lane deviations) representing physical manifestations of driver performance provide insight into the operational effects of fatigue.
Dinges and Graeber (as cited in Williamson et al., 1994) indicated that a serious loss of physiological arousal and slowed sensorimotor function may occur despite a driver's ability to compensate while suffering from a diminished capacity attributable to fatigue. Although driving performance measures may be relatively insensitive to a moderately diminished mental capacity, diminished capacity occurring as a result of fatigue is likely to be detectable by a set of fundamental physiological measures. Because physiological measures provide a direct objective measurement of the fatiguing system, they are the measures most likely to provide a direct measure of fatigue. This is especially true of measures such as electroencephalography (EEG), which provides insight into cerebral arousal. Notable shortcomings of physiological measures are they are not easily measured in an unobtrusive manner and are not easily deployable without the constant vigilance of a trained experimenter. Common physiological measures sensitive to the onset and detection of fatigue include adrenaline/noradrenaline production, cortocosteroid production, brain electrical activity, eyelid closure, eye position/eye movement, heart rate, and gross body movement.
Despite the variety of objective measures available, fatigue remains essentially a subjective experience. Several methods of self-report have been developed to assess the subjective experience of fatigue (Hoddes, Zarcone, Smythe, Phillips, & Dement, 1973; Mackie & Miller, 1978; Williamson et al. …