Is Pilots' Visual Scanning Adequate To Avoid Mid-Air Collisions?
The article seeks to find out whether pilots engage in the prescribed scanning patterns set out by regulatory authorities. They analyse what pattern pilots actually use and how effective it is. Their results showed that the participation pilots did not use the prescribed method of visual scanning.
The table below shows the percentage of time the pilots gaze was directed outside at different workload phases set by the researchers. There was significant difference between all 12 pilots. When the pilots gaze was focused inside the flight deck attention was mainly focussed on the instrument panel (results not reported).
|Time Spent Gaze Directed Outside|
|Condition of Flight||Percentage of Time Looking Outside|
Note: see procedure section for flight phase description
The next table shows fixation rates of the pilots over the six AOIs. Again they highlighted significant differences between individual pilots gaze.
|Average Rate of Fixation on the AOIs|
|Area of Interest||Fixation in Minutes|
|Centre Front (OS)||41|
|Centre Right (OS)||10|
participants fixated the instrument panel far more frequently than the windscreens (during a VFR flight).
participants fixated the centre-front windscreen far more frequently than the other three windscreens.
Experimental research approach
12 Pilots were used in the current experiment.
All 12 pilots held a current FAA instrument rating and appropriate aircraft rating.
All pilots had 20/20 vision or corrected to the value.
On average each pilot had 1400 total pilot hours and 15 years flying experience
Experimental design, one-group posttest only design. All pilots were given a scripted, 45 minute familiarization flight and then all pilots flew the experimental scenario in which data was collected.
Independent variable: workload
Dependent variable: visual scanning pattern
A Line of Sight (LOS) system from ISCAN Inc. was used to collect eye scanning data. The system included:
- a headband fitted with a camera - determines head position
- a magnetic sensor - determines head orientation
- a computer - to perform the computations necessary to determine where the pilot was looking in the flight deck
- AST Hawk 201 FAA-approved flight-training device (single-engine piston aircraft)
Flight deck set-up with areas of interest (AOI)
- x4 AOIs were windscreens (17" CRTs) displaying the (simulated) outside world (showing terrain, sky, and traffic as programmed)
- x2 AOIs were the instrument and engine indicator panels
Firstly pilots undertook a 45 minute familiarisation flight to allow them to get used to their simulated environment. Pilots were then calibrated on the eye-tracking apparatus.
The participants then undertook the experimental 45 minute VFR cross-country flight. Once cruising altitude was reached, the participants were subjected to a series of experimental scenarios with varying workloads:
- low workload period - 3 minutes (LWL 1)
- high workload period (moderate turbulence in the vicinity of terrain) - 3 minutes (TURB)
- low workload period - 3 minutes (LWL 2)
- traffic period - 14 minutes (TRAFFIC). Aircraft appeared for different time periods at various crossing angles. The amount of aircraft appearing at any time varied with different altitudes to the simulated aircraft.
- final workload period - 3 minutes (LWL 3)
Eye movements were extracted from the raw data by the absolute deviation method.
Cluster sequences were then analyzed for areas of fixation.
Data collected in this study has the potential to provide valuable information to aviation professionals in understanding pilots scanning patterns at different workloads during a flight. However the study had several shortfalls that would prevent it being representative of the larger population (the piloting world):
- small sample size
- aircraft type used
- simulated flight conditions (unable to determine how real the simulated flight was)
- using one type of aircraft (small, general aviation type)
- researchers did not mention how the current sample of pilots were chosen.
For this research to have any meaning to a larger population the experiment would have to be carried out on a much larger scale.
The purpose of the researchers experiment was to show if pilots visual scanning techniques were adequate enough to avoid mid-air collisions, however the current experimental design could not conclusively answer the question.
Colvin, K., Dodhia, R., Dismukes, K. (2005). Is Pilots' Visual Scanning Adequate To Avoid Mid-Air Collisions?. In Proceedings of the 13th International Symposium on Aviation Psychology (pp.104-109).
Want to know more?
Applications Symposium, (pp. 71-78). New York: ACM Press. : A description of eye fixations and saccades.
- Mumaw, R. J, Sarter, N. B, & Wickens, C. D. (2001). Analysis of pilots’ monitoring and performance on an automated flight deck. Conference Proceedings of the 11th International Symposium on Aviation Psychology. Columbus, OH: The Ohio State University.
- Further reading on Pilots attention focus on a large aircraft automated flight deck.
Contributors to this page