Factors influencing the usability of collision alerting systems in gliding

The dissertation at hand identifies and analyzes how well glider pilots use low-cost collision alerting systems. While being generally recognized as a commendable tool for helping glider pilots see and avoid other traffic, these systems have been cited as possible contributing factors in several accidents. In literature, no in-depth research on how glider pilots may interpret or misinterpret their indications was found. At the beginning of this dissertation, a market study of human-machine interfaces for low-cost collision alerting systems is presented. During the study, different human-machine interfaces are taxonomized. The low-complexity and radar-style displays were found to be popular display formats. Also it was discovered that a perspective presentation of traffic has been evaluated for military applications, but not for a gliding context. Thus, a perspective presentation of traffic is proposed. The prototype of a perspective display format for gliding is developed by relying on a user-centered design process. Then, the design features of the low-complexity, radar-style and perspective displays are compared. This results in several hypotheses comparing the usability of the three display formats being postulated. In order to experimentally evaluate these hypotheses, 137 glider pilots partake in a laboratory experiment. They are presented with traffic information on one of the three display formats installed in a flight simulator. The participants then indicate where they suspect the traffic to be located in the outside world while being exposed to different flight conditions. Performance and subjective satisfaction measurements are recorded during the experiment. Inferential statistics are used to evaluate the experimental data. The perspective display format results in the most precise estimates of where traffic is located. Generally, errors in estimating traffic position increase as the participants’ ownship deviates from straight and level flight. Reaction time does not vary notably between display formats or different flight conditions. Subjective learnability and usability ratings favor the perspective display format over the two other formats analyzed. Overall, the perspective display format exhibits optimized usability in all dimensions when compared to the other two formats. During the usability analysis, circumstantial evidence arises which suggests that not all participants might interpret the data shown on their display similarly. A probable cause for this may be different knowledge deficits which are experienced between participants. These deficits may result in participants mentally modeling the collision alerting system incorrectly, thus leading to incorrect coordinate systems for interpreting the traffic information. A method for identifying these mental models is developed. The ensuing analysis reveals that most participants using low-complexity or radar-style display formats incorrectly interpret traffic information in an ownship-fixed fashion. Contrary, most participants working with the perspective display format perform at least some of the required rotations of their personal coordinate systems. The concept of different mental models based on different personal coordinate systems shows potential as an analysis tool for future display designs. From these findings multiple recommendations are deduced. They are directed at different stakeholders in the gliding community, including glider pilots, aircraft owners and operators, regulatory authorities, glider manufacturers, flight schools, competition rule makers and organizers, as well as designers of collision alerting systems and associated human-machine interfaces. Closing this dissertation, the potential for future human factors research in the gliding community is highlighted.