The British Civil Aviation Authority (CAA) commissioned NLR to assess whether it could formulate a means of accurately and reliably measuring and predicting fatigue in individual pilots. In close cooperation with the Netherlands Institute for Neuroscience (Nederlands Instituut voor Neurowetenschappen – NIN), initial testing was concluded in 2014 for a fit-to-fly test that rapidly ascertains the degree of fatigue and alertness.
CAA sought answers to the following three questions. First, can pilot fatigue be measured in such a way as to establish a link with flight performance? Second, can the consequences of this loss in performance be converted into quantitative recommendations in the area of safety? And finally, can such recommendations be translated into operational safety management? Current research focuses largely on the first question.
In an extensive series of ambulatory, laboratory and simulator tests, 32 pilots were monitored and measured over a nine-day period. With the aid of mobile registration equipment attached to the body, a wide array of fatigue-related variables was monitored and charted over the first eight days. On the ninth day, the pilots completed numerous tests in the laboratory, culminating in a flight test in NLR’s GRACE simulator.
The study assessed which variables were most relevant in registering fatigue. This included variables that contribute to fatigue, such as sleep and stress, as well as variables that are indicators of fatigue. These were measured on the basis of an electroencephalogram (EEG), for example the relative contraction of the pupils and responses to particular actions. The outcome of these variables was then compared with flight performance in the simulator. The aim was to establish the relationship between the degree of pilot fatigue, on one hand, and the resultant flight performance, on the other.
Analysis of these tests will be used to define a standardised set of variables that can be used to conclusively ascertain fatigue: the fit-to-fly test. Further studies are needed to establish whether it would be feasible to implement such a test in practice and, if so, in what way.