Eye-tracking parameters as a predictor of human performance in the detection of automation failures

The increasing amount of automation in aviation systems requires that the operators monitor those systems appropriately. "Operators monitoring appropriately" (OMA) have been defined as those who monitor in a way that enables them to detect automation failures and resume control if automation fails. Identifying OMA reliably is a current objective for the selection of future aviation personnel. Eyetracking data have been utilised to provide real-time measurements of visual and cognitive information processing. This raised the question of which eye-tracking parameters are important for differentiating between high performance and poor performance among operators. Previous studies had revealed time-sensitive eyetracking parameters that help identify OMA who are prepared to resume control. This study dealt with finding eye-tracking parameters that help identify OMA who are able to detect automation failures. An experiment was conducted with 33 candidates for the DFS (Deutsche Flugsicherung GmbH). A simulation tool called “MonT” (Monitoring Test) was developed, which required test subjects to monitor an automatic process and register automation failures while eye movements were recorded. Results have revealed suitable eye tracking parameters that help differentiate between the participants' performance level in detecting failures. In the long term, MonT will be further developed with the aim of meeting the criteria for future selection tests.