Presentation

· Contributors · Organizations · Search Program · Flagged · Happening Now

Determination of Occupational Fatigue Detection Thresholds Using Naturalistic Fatigue Datasets

DescriptionFatigue has contributed to multiple vehicular crashes and disasters in the oil and gas industry. Existing studies have differed with fatigue thresholds when using the same metrics in laboratory settings. There is a need to create accurate and reliable thresholds for occupational settings as fatigue-related incidents may be fatal. Participants were recruited from two offshore ships where they provide subjective, performance, and physiological measures daily for two-weeks. Performance and subjective measures were combined using K-means clustering to form three groups of fatigue. The difference between these groups, with respect to physiological, sleep and other subjective measures, were determined using a Kruskal-Wallis’ test. The combination with the highest number of significant differences was chosen as the best fatigue clustering. A combination of mental exhaustion ratings and lapses from the Psychomotor Vigilance Test (PVT) provided the most significant differences between the fatigue groups. Preliminary validation using physiological, sleep and other subjective measures demonstrated that the fatigue groups differed in terms of physiological and other subjective measures. The fatigue thresholds created with subjective and performance measures were able to capture physiological fatigue changes indicating that these thresholds may be used to identify fatigue for early fatigue mitigation.

Contributors