Identification of Expert Tower Controller Visual Scanning Patterns in Support of the Development of Automated Training Tools

Abstract

Researchers from the Federal Aviation Administration’s (FAA) Civil Aerospace Medical Institute and from the University of Oklahoma’s School of Industrial and Systems Engineering are studying the characteristics of expert tower controller visual scanning behavior in support of the FAA’s exploration of ways to enhance controller training. Training enhancements potentially include the use of advanced simulation tools (such as virtual reality systems) to teach controller trainees critical scanning skill(s). We collected eye-tracking data from controller subject matter experts while they controlled simulated air traffic scenarios in a high fidelity tower cab simulator. In this paper, we describe the methodology used to collect and analyze the data as well as summarize the results of the analyses. These results may inform the design of scanning training tools. Furthermore, we summarize lessons learned from our use of simulation and our methods of collecting performance measures that may be useful for those developing scanning training tools that will also use simulation. Our findings suggest that training tools should continue to train what is taught in today’s curriculum regarding scanning, to frequently scan “hot spots” such as both ends of an active runway, and to prioritize traffic at the airfield before traffic occurring farther out. Our findings also suggest that controllers could be taught to use different scanning patterns based on the ATC task they are currently carrying out and to use these patterns consistently.