Fatigue-Indicator in Operational Settings: Vocal Changes

Fatigue is an important factor in aviation accidents and incidents. Since fatigue cannot always be prevented, it needs to be detected in real time so that countermeasures can be taken. This study researches whether vocal changes (in vocal intensity and fundamental frequency) can be used as a measure for fatigue in an operational aviation setting. Sixteen participants were measured two times. Before the first test moment, they were asked to sleep eight hours or more and before the second test moment six hours or less. During each test moment, they performed a PVT, filled in the KSS, and did two speech tasks. One task was aimed at free speech and one task was aimed at procedural speech. Pre-processing included segmentation of the speech into words and extracting fundamental frequency (f0) and intensity values. An overall mean of both variables was calculated for both free and procedural speech. Speech, PVT reaction time, PVT lapses and KSS scores were analyzed in SPSS using Paired Samples t-Tests and Wilcoxon Signed Ranks Tests. Participants slept significantly less during the night before the second test moment and scored significantly higher on the KSS. For the PVT, no differences in both reaction time and lapses were found. No significant differences in average f0 and intensity for both free and procedural speech were found either. The results did not show a significant relationship between fundamental frequency, intensity and fatigue. Further research is needed to examine if vocal changes can be used as a reliable fatigue measure.