nce of automation has been a bottleneck for successful introduction of automation in Air Trac Control. Strategic conformal automation has been proven to increase automation acceptance, by creating a better match between automation and operator decision-making. In this paper strategic conformal automation for Air Trac Control is designed using machine learning techniques. Rather than having pre-dened control strategies, which do not always match with individual operator decision-making, the automation is based on the operator's decision-making. Results show that when operators demonstrate
their control strategies, machine learning techniques can identify these strategies and use them to learn similar control strategies. Apart from mimicking control strategies in iden tical trac scenarios is it possible to use machine learning to solve similar, yet dierent con icts by applying similar control strategies, without the need of human demonstrations for that particular conict scenario. Future research should be done to investigate whether strategic conformal automation indeed increases automation acceptance, as well to investigate how the approach taken in this study can be applied to real-life trac scenarios.
The different versions of the original document can be found in: