BACKGROUND
Mental workload assessment is a recurrent issue in air traffic control (ATC). Studies of ATC have used either objective aspects, i.e., numbers and distribution of aircraft, or subjective factors, such as self-imposed performance and stress levels, with mixed results. This is partly due to the difficulty in bringing together comparable data pertaining to both air traffic, with its ever-changing distribution, and judgement or quickly fluctuating psychophysiological variables.
METHODS
We propose a method of mental load estimation devised to take into account both objective traffic variables and the additional load imposed by subjective effects, including the seriousness of conflicts and the time-pressure for their resolution. First, we developed a traffic load index (TLI) to identify time boundaries during which additional load may occur. Then we quantified the additional load according to the air traffic situation.
RESULTS
TLI was developed from analysis of 25 h of recordings of radar control sessions involving 25 professional air traffic controllers at a major airport. Results were then compared with a simple objective index (number of aircraft) and subjective workload ratings (NASA-TLX test). The whole population (intersubjects analysis) showed a better correlation between the TLI and the self-rated workload than for the number of aircraft alone. Among the controllers who rated more than one level of workload through the TLX-test, 77.8% showed better correlation between TLI and TLX than between N and TLX (intrasubjects analysis).
CONCLUSION
Workload estimation should integrate both objective task variables and subjective evaluations associated with them.