Effects of Visual Clutter on Flight Performance, Workload and Gaze Pattern

Keywords: visual clutter, primary flight display (PFD), performance, workload, eye tracker Background The aerospace industry has seen a rapid increase in the amount of information made available to the pilot on the primary... [ view full abstract ]

Keywords: visual clutter, primary flight display (PFD), performance, workload, eye tracker
Background
The aerospace industry has seen a rapid increase in the amount of information made available to the pilot on the primary flight display (PFD). An overabundance of information, or the presence of information irrelevant for the task at hand, increases the level of visual clutter on the PFD. Past research found a mixed effect of visual clutter on the pilot’s performance during flying task (Alexander, Wickens, & Hardy, 2005; Alexander et al., 2005; Kim et al., 2011; Ververs & Wickens, 1998). In a previous study (Doyon-Poulin, Ouellette, & Robert, 2012), we explained this mixed effect by the fact that past research changed the flight guidance function used on the PFD at the same time than the level of visual clutter – i.e. flight director compared to tunnel-in-the-sky symbology. This may have masked the effect due to visual clutter alone.

In this study, we created three experimental PFDs with different visual clutter levels (low, medium, high) and having the same flight guidance function – i.e. localizer and glideslope instruments. These experimental PFDs allowed this study to control for the influence of other variables that were not considered in past research, such as flight guidance functions.
Method
Setup: Twelve pilots flew an instrument approach in a fixed-based, side stick-controlled flight simulator. Scenario was in instrument meteorological condition (IMC) with a cloud ceiling at 400 feet. PFD was presented on a 30.5 x 22.9 cm screen located at a viewing distance of 86 cm from the pilot’s eyes. Pilots’ gaze was recorded using a FaceLAB 5.0 eye tracker (© Seeing Machines) with the cameras positioned under the PFD.

Independent variable: The independent variable was the PFD clutter level, with 3 values: low-clutter, medium-clutter and high-clutter. Figure 1 shows the three PFDs used in the study.

Dependent variable: Pilots completed the approach using each of the three PFDs, following a within-subject plan. The order of presentation of PFDs was balanced between pilots. For each display, pilots repeated the approach three times before using the next display.

After having completed all three repetitions for a display, pilots filled the clutter rating scale to measure the perceived level of clutter of the display (Kaber et al., 2008), the NASA-TLX mental workload scale (Hart & Staveland, 1988) and the visual qualities rating sheet (Lavie & Tractinsky, 2004) to evaluate the perceived display aesthetics.
Results
We analyzed results using a within-subject ANOVA with three levels of clutter. We present results for the final leg starting at the Final Approach Fix (FAF) and ending when the aircraft crossed the cloud ceiling at an altitude of 400 feet.

We found significant effects of visual clutter on all three subjective measures. There was a strongly significant effect of PFD clutter level on subjective clutter ratings (F2,22 = 42.39, p < .001, partial-η2 = .79), on workload score (F2,22 = 8.67, p < .005, partial-η2 = .44, see Figure 2) and on the visual quality ratings (F2,22 = 22.60, p < .001, partial-η2 = .67).

For flight technical performance, we found a significant effect of visual clutter on the mean localizer deviation during the final leg (F2,22 = 3.70, p < .05, partial-η2 = .25, see Figure 3), with the medium-clutter PFD showing a lower deviation (0.22 dot) compared to the low-medium PFD (0.30 dot) and high-clutter PFD (0.28 dot). Surprisingly, there was no significant effect of visual clutter on all the other flight performance parameters, i.e., glideslope deviation, side stick activity and vertical speed control (all ps > .10) and they will not be discussed.

Results from the eye tracker revealed that mean fixation time was 15% longer for the medium-clutter PFD (555 ms) compared to the low- and high-clutter PFDs (476 ms and 483 ms, respectively). Length of a saccade, i.e. distance between two fixations, was 15% longer for the high-clutter PFD (169 pixel) compared to the low- and medium-clutter clutter PFDs (147 and 146 pixel, respectively)
Discussion
This study focused on the effects of visual clutter on pilot performance, using three experimental PFDs designed to control with confidence the display clutter level. We measured three groups of dependent variables to test for the effects of visual clutter: subjective perception, flight technical performance, and gaze pattern.

We found strongly significant effect of visual clutter on all three subjective measures. Pilots perceived the medium-clutter PFD has demanding the lowest level of workload and being visual more pleasing than the low- and high-clutter PFDs. While the effect of visual clutter on workload was clear, its effect on flight performance was weak. Our results showed that the localizer deviation was the only variable to be significantly different, with the medium-clutter PFD having the best precision. We found no effect of visual clutter on the other variables (i.e., glideslope deviation, control of the vertical speed and side stick activity). Pilots’ experience, with over 4000 flight hours on average, and the higher workload levels for the low- and high-clutter PFDs suggest that pilots compensated the effects of visual clutter by investing more attention to the task to achieve a similar level of performance.
Conclusion
In this study, we changed the clutter level of three PFDs with the same flight guidance function. We found that pilots using medium-clutter PFD had lower workload score and better lateral control of the aircraft during a simulated approach than with low- and high-clutter PFDs. We also found that pilots perceived the medium-clutter PFD as visually more pleasant. Results from the eye tracker showed that pilots looked for a longer time during each fixation with the medium-clutter PFD than with the other displays. These results show the relevancy to optimize visual clutter of complex aerospace displays, as it impacts subjective and objective performance of experienced pilots.
References
Alexander, A. L., Wickens, C. D., & Hardy, T. J. (2005). Synthetic Vision Systems: The Effects of Guidance Symbology, Display Size, and Field of View. Human Factors, 47(4), 693‑707.
Doyon-Poulin, P., Ouellette, B., & Robert, J.-M. (2012). Review of visual clutter and its effects on pilot performance: A new look at past research. Dans Digital Avionics Systems Conference (DASC), 2012 IEEE/AIAA 31st (p. 2D1‑1 ‑2D1‑11). doi:10.1109/DASC.2012.6382290
Hart, S. G., & Staveland, L. E. (1988). Development of NASA-TLX (Task Load Index): Results of empirical and theoretical research. Dans P. A. Hancock & N. Meshkati (Éd.), Human mental workload (p. 239‑250). Amsterdam, Netherlands.
Kaber, D. B., Alexander, A. L., Stelzer, E. M., Kim, S.-H., Kaufmann, K., & Hsiang, S. (2008). Perceived Clutter in Advanced Cockpit Displays: Measurement and Modeling with Experienced Pilots. Aviation, Space, and Environmental Medicine, 79, 1007‑1018. doi:10.3357/ASEM.2319.2008
Kim, S.-H., Prinzel, L. J., Kaber, D. B., Alexander, A. L., Stelzer, E. M., Kaufmann, K., & Veil, T. (2011). Multidimensional measure of display clutter and pilot performance for advanced head-up display. Aviation, Space, and Environmental Medicine, 82(11), 1013‑1022. doi:10.3357/ASEM.3017.2011
Lavie, T., & Tractinsky, N. (2004). Assessing dimensions of perceived visual aesthetics of web sites. International Journal of Human-Computer Studies, 60(3), 269‑298. doi:10.1016/j.ijhcs.2003.09.002
Ververs, P. M., & Wickens, C. D. (1998). Head-Up Displays: Effect of Clutter, Display Intensity, and Display Location on Pilot Performance. The International Journal of Aviation Psychology, 8(4), 377‑403. doi:10.1207/s15327108ijap0804_4