Quantifying monitoring awareness of air traffic controllers using eye-tracking signal analysis

• Main Author: Neo, Jin
• Other Authors: Lye Sun Woh
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2020
• Subjects: Engineering::Aeronautical engineering
• Online Access: https://hdl.handle.net/10356/141072
• Institution: Nanyang Technological University

Global air traffic is projected to grow rapidly with the continually rising demand for air travel. The growth in air traffic volume has put an increasing amount of pressure on the air traffic controllers, who are responsible for maintaining the safety and efficiency of the global airspace. New automation aids and procedures introduced over the years to help the controllers cope with the rising air traffic were found to have counter-productive effects on the performance of the controllers. The study of human factors associated with operating complex air traffic control systems such as situational awareness, has long been recognised as a necessary matter. Present approaches in assessing the situational awareness of the controllers were found to have its limitations. Physiological measures such as eye-tracking have received much less attention in comparison but are gaining popularity in recent times. In this study, a remote eye-tracker was utilised to capture the eye fixations of test participants for one-hour sessions of air traffic monitoring. The experimental data collected were used to construct suitable eye-tracking metrics and monitoring parameters, to quantify the participants’ monitoring awareness and performance. Based on statistical analysis, noticeable differences were found between the expert (licensed) and non-expert (non-licensed) participants. The study proposes a novel methodology using eye-tracking as a potential performance screening tool for air traffic controllers. Monitoring parameters that could statistically distinguish the two groups were used to develop a binary classification model. The model could be employed to predict an individual’s class: expert or non-expert, based on the monitoring parameters. Air traffic controllers are expected to fall under the expert class, and any deviations from this standard could be captured. Support Vector Machines (SVM) classification models were tested in this present study, and desirable prediction performances were achieved.