| dc.description.abstract | Pilots operate in information dense environments that require continuous instrument scanning, communication, navigation, weather monitoring, traffic awareness, and external visual assessment. These demands become more challenging during mountainous operations, where terrain proximity, rapidly changing hazards, or competing sources of information increase workload and reduce decision time. Since human working memory capacity cannot increase to accommodate high workload flight conditions, flight deck interfaces should reduce unnecessary information processing demand and support timely situation assessment to improve pilot performance.
This dissertation develops and evaluates a context-specific interface framework that can support different aviation operations by separating stable display architecture from operation-specific information content. The framework first establishes a fixed-zone interface architecture from human factors principles, aviation display guidance, military display standards, and visual search and decluttering literature. I used mountainous operations to validate the generic interface, as operations at high altitude and around mountainous terrain often result in Controlled Flight Into Terrain (CFIT) accidents. I used complex network analysis and information from the NTSB accident database to identify dominant information categories associated with mountainous operations/CFIT accidents. Decision making, altitude control, aircraft control, and environmental monitoring were central factors in the accident network. These factors guided the population of the generic interface placeholder with CFIT-relevant information, which included terrain relative cues, aircraft performance information, route context, and alerts. The resulting tailored prototype was implemented in a flight simulation environment (experimental group) and evaluated against a baseline cockpit condition (control group) in two human subjects experiments.
The evaluation used subjective, physiological, operational, and visual attention measures including NASA-TLX, SART, heart rate variability, pupil dilation, flight path deviation, dwell time, and fixation measures. The tailored interface group reported improvements in perceived performance and situation understanding, along with reduced frustration. HRV results showed lower task lnRMSSD in the tailored interface group, suggesting task related engagement. Flight path deviation suggested tighter path control in the tailored interface group, although the flight path differences did not reach statistical significance. Visual attention showed that both groups allocated more attention to attitude information, while the tailored interface groups distributed less attention to some conventional instrument tape and also attended the interface.
Overall, this dissertation contributes to a human factors-based, accident-informed method for augmenting modular flight deck interfaces. This framework can adapt to different operational contexts by having a fixed zoned layout while using context specific risk analysis to determine which information should receive display priority. The CFIT case study demonstrates one application of the framework and provides initial evidence for its usefulness in supporting pilot situation assessment during high-risk operations. | en_US |
