This article was originally printed in the Jan/Feb 2010 issue of Torch magazine. It is reproduced here with the kind permission of the United States Air Education and Training Command.
We flew. We saw. We avoided.
Only when flying, that’s not always the case. As far back as you can remember, the “Mk-1 eyeball” has not let you down. The concept of scanning is probably first identified as “important to remember” during infancy when you initially run your “nugget” into non-moving objects strewn around your play space (like chairs, tables, walls, etc.). You learn to keep your eyes moving as you “fly” through the terrestrial environment, and confidence grows in those orbs that bracket your nose.
Yet, no matter how fundamental the concept of keeping your eyes open and scanning may seem, in the flying environment, there are dozens of events each year that record how close two aircraft come to a failure to maintain appropriate spacing ... or even crash midair.
On Nov. 28, 2007, for example, two T-6A Texan II training aircraft collided over Mississippi. The instructor and student pilots in the first aircraft simply didn’t see the second plane, also manned by an instructor and student, even though it was flying straight and level in the traffic pattern. All four pilots ejected, suffering only minor injuries. But both aircraft were lost... $10 million down the drain.
On Feb. 20, 2008, two F-15 Eagles collided off the coast of Florida when both pilots failed to clear their flight paths. One pilot was killed; the other ejected. The two destroyed aircraft cost $83.3 million combined.
Applying see-and-avoid while walking through a crowded street is reasonably easy considering the ground speed, closure rate and relative ease of maneuvering around objects. Applying principles of vision in flight is physically and mentally more challenging by far.
Early records of commercial aviation are strewn with examples of the failed scans of the eyeballs. In the period 1958 to 1988, there were some 40 airliner accidents in the United States that involved midair collisions, killing a total of 908 people according to the National Transportation Safety Board (NTSB). The number of midairs has gone down only slightly in subsequent decades. In the period 2005 to 2009, the NTSB reports there were 33 midairs and 53 associated fatalities in civil aviation.
Typically, midair collisions occur with a mix of high-speed traffic operating instrument flight rules and low-speed traffic operating visual flight rules with one aircraft climbing or descending. Stunning in this analysis is the fact that midair events typically take place in daylight, under good visibility, with the low-speed aircraft within 30 degrees relative bearing of the other aircraft’s flight path.
The most common cause cited in midair mishap reports to date is “failure of the pilots to see and avoid each other.”
It’s hard to deny that the conditions for midair are ripe in a high density environment packed with inexperienced pilots earning their wings; skills of scanning and accurately perceiving the threats are still being developed in the young aviator. Two aircraft occupying the same space when there is a well experienced aviator scanning the horizon is harder to explain.
Visual science can help unbind some of the dilemma.
The smallest image that can be perceived at the fovea (central region of the retina) ranges from 0.5 to 1.0 minutes of arc. Probability of detection for targets which exceed 1.4 minutes (0.023 degree) of arc, visual angle, is going to depend on the size of the object, the anticipation of the object, and visibility factors (weather and conspicuity). In other words, as long as your eyes are open and the squishy gray blob between your ears is in the “ON” position, there is no physical reason to miss a target of reasonable size entering your space.
If only we were restricted to single, well marked, pre-briefed objects in the airspace!
Objects in our visual world are not packaged with a convenient code according to shape, size or color, so the brain (visual cortex) has to decipher and analyze every snapshot that arrives at the door. Now tie in the reaction time to the physical recognition of a midair target. According to measurements taken in military studies, the time necessary for a pilot to recognize a potential midair target and take evasive action is 12.5 seconds. Nice figure in a perfect world. Possibly a more relevant factor to consider is where the object appears on your retina.
Unfortunately, there is a precipitous drop in visual acuity as the target moves out of central vision to peripheral vision.
The scanning technique is central to the lecture on vision during initial physiological training and becomes a practiced and conscious skill developed as you break lock from the instrument panel long enough to assemble outside visual references. But here’s the “big whammy” you may have overlooked: The peripheral eye is highly sensitive to moving objects and less sensitive to non-moving objects. Unlike central vision, if an object is projected on the peripheral viewing area with little or no movement, the receptors register “no factor” and a potential airspace conflict (e.g. target) is missed and may actually become a factor.
Despite relative motion of the aircraft, the object will only enlarge and not move across the retina. This was the case that involved a T-37 and a crop duster. Neither the student nor instructor in the Tweet saw the large yellow air tractor soaring to their right on a hazy midday in January 2005, and the aircraft collided, killing the civilian pilot. Size does matter, but so does perceived movement across the visual field.
How closely do bold, black letters set against a white background compare to complex targets against a low contrast background? In practice, not so much. But using this standardized measure, vision scientists have come to the consensus that the minimum visual angle to ensure somewhat reasonable accuracy and probability in detecting another aircraft is about 12 minutes (0.2 degrees) of arc.
One more ripple in this estimation is important to note. Most mathematical models account for closure rate, target aircraft size, meteorological visual range, but do not attempt to model physiological or mental processes underlying pilot performance. Individual situational awareness influences the speed and accuracy of time critical actions, and situational awareness can be bolstered or blasted by technology.
Work completed at the Massachusetts Institute of Technology that led to Traffic Alert and Collision Avoidance System advisory systems in the mid-1980s implied that the presence of TCAS increased traffic search effectiveness by a factor of eight (e.g. one second of search with TCAS was as effective as eight seconds of search without TCAS). Collision avoidance technology available in modern aircraft has been measured to speed acquisition time by up to 15 seconds and improve the probability of visual acquisition. Some estimates claim probability approaches 100 percent by the 10 seconds prior to collision mark.
Audio warnings and visual displays are well known to markedly improve visual-search effectiveness and TCAS computations for optimal avoidance manoeuvres complete the trifecta of see-and-avoid strategies. The sobering reality is that see-and-avoid has significant limiting factors when relegated to a purely visual task.
Nevertheless, the “Mk-1 eyeball” is as much a primary line of defense for the modern aviator as it was for the pioneers of flight. Combining the scan and cranium swivel technique with technology is a learned skill that must be practiced. Instructors and students deep in the weeds on knocking out syllabus objectives while perfecting flight techniques and skills-based learning need to remember to keep all available systems switched in the “ON” position when entering the pattern or manoeuvring through a military operations area.
It’s worth a second look. Check-6!