Proposed MMADC Project Plan for FY09

In the first and second year of this NRA, we developed and tested a multidimensional measure of display clutter as well as models of perceived clutter based on visual display properties and information elements. These studies focused on HUD SV/EV technologies in simulations of the advanced commercial aircraft cockpit. The second year study successfully demonstrated sensitivity and specificity of the new clutter measure and models for assessing pilot experiences with advanced HUDs in standard ILS landing approaches.

In the third and final year of the NRA, we seek to validate our measurement approach and predictive models for implementation across domains that may exploit NextGen technologies. A step-wise validation approach involves first assessing the multidimensional measure of clutter and models of perceived clutter using similar technology (SV/EV displays) but in a different domain; for example, application of such technologies in space systems like the Altair cockpit. Although the current Altair display concepts are head-down (HDD) and include color, the information content of the displays and formats of presentation are similar to those used in the HUD SV/EV displays, which we have investigated. The color features of the Altair displays are more discriminable than the hues of “green” in a HUD SV/EV displays, but can be assessed using the same subscales of our new clutter measure. A final step in such a validation process would involve assessing our measure and models on different technologies (e.g., NextGen ATC displays) within the same domain (i.e., aviation). (This could be a future research collaboration between NASA, NC State and Aptima.) By initially examining application of the new measure and models to similar technology in the space domain, we can structure/stage lander pilot tasks with the SV/EV displays in a manner similar to the aircraft pilot landing/approach tasks, as studied in the second year project. This would allow us to make focused assessment of the sensitivity and reliability of our measure to environment and display changes.

The scientific justification for implementing the measure validation in the space domain, while continuing to focus on SV/EV technology, is two-fold:

(1)Since the PFD concept we explored in the IFD simulator in the second year of the project is similar to the PFD concept in the Altair cockpit design, the visual properties are expected to be analogous between the two domains. Any differences in these properties can be quantified using display analysis software applications we have developed. We can then formulate specific hypotheses on how these differences will impact perceptions of clutter (captured through our measure) and model predictions of clutter. We can also make comparison of the relationships of perceived clutter with performance and workload results across domains.

(2)Although the PFD content may be comparable across cockpit designs (IFD and Altair), there may be differences in how the same types of PFD information are used for aviation and space domains. The same information elements may have different relevance to a pilot given the specific task context inherent to the given domain. We can specify differences in display information relevance across phases of flight in the two domains and make predictions about how they will impact performance, workload and clutter ratings.

The above points represent specific opportunities for assessment of clutter measure and model sensitivity. It is important to emphasize that the displays we want to investigate through the Altair cockpit are concept displays. The exact designs have yet to be prototyped based on test requirements. The configuration, information features and format of the concept displays can be manipulated to be more of less similar to the IFD displays we studied in Year 2. For example, we focused on pilot use of the HUD with PFD content and use of a HDD PFD and navigation display (ND). It is possible to present only a PFD and ND in the Altair simulation instead of the quad-display presentation (see the attached PowerPoint slides). In general, this configuration would be more representative of the IFD display setup.

With respect to the details of the validation approach, we seek to demonstrate that our multidimensional measure of clutter achieves a range of psychometric criteria, including validity, sensitivity, diagnosticity, selectivity and reliability (Wickens, 1992). Validity will be assessed in terms of the significance of correlations of clutter scores with quantitative measures of information density and occlusion across display conditions. Sensitivity will be assessed based changes in pilot perceptions of clutter for various SV/EV display conditions. It can also be assessed in terms of differences in clutter scores for pilots with different levels of experience. Diagnosticity will be evaluated in terms of whether the clutter score discriminates among causes of variation in pilot perceptions of clutter (i.e., do different design factors have different influence on scores across task and environment conditions). Selectivity will be assessed in terms of whether specific subscales of the multidimensional measure of clutter are more or less sensitive to display visual property and information feature changes. Regarding reliability, we will assess consistency in clutter scores across display conditions for pilots with a given level of experience. We believe that this represents a thorough and sound validation approach for the new measure of clutter for the final year of the NRA project.

References:

Wickens, C. D. (1992). Engineering Psychology and Human Performance (2nd edn.) New York: HarperCollins.