Muret-l'Herm, France- The Service d'exploitation de la formation Aéronautique (SEFA) announced the purchase of two WOMBAT-CS systems from Montreal-based ÆRO INNOVATION, INC. The 45-year old institution, owned by the French Government, hires 665 people and trains over 1000 pilots every year. SEFA operates eight centers located primarily in southern France with headquarters in Muret-l'Herm, near Toulouse, training is conducted in Melun, Saint-Yan, Grenoble, Saint-Auban, Montpelier, Carcassonne, Castelnaudary and Biscarrosse, with a centralized major maintenance facility newly located in Castelnaudary.
SEFA provides airline pilots to organizations such as Air France, Air Inter, Air Algérie, Air Gabon, Viet-Nam Airlines, Air Kampuchea, DGAC, the French Air Force, and manufacturers such as ATR, Airbus Industries, and Aérospatiale. Psychologist and Selection Specialist Dr. Sammy Szpic acted as the liaison officer between the SEFA and ÆRO INNOVATION. The decision to choose the $54,500 (USD) WOMBAT-CS systems was based on the recognition of the urgency to assess Situational Awareness of modern pilots flying complex, automated airplanes. Dr. Szpic explained: "I have been involved in selection for many years. WOMBAT-CS is clearly well beyond any other method we know. WOMBAT's situational awareness assessment is essential to determine, in advance, whether an applicant will not only succeed in our training program but also will have success during his career." Dr. Szpic's long-term planning is a demonstration of how serious SEFA is in facing the next century. SEFA is also investing heavily in computer-based training systems, multi-crew training, and modern CRM training: the DuoWOMBAT-CS from ÆRO INNOVATION.
In the late evening of the first day of 1978, an Air India B-747 took off from Santa Cruz Airport in Bombay, bound for Dubai, United Arab Emirates. As the airplane climbed out over the Arabian Sea and the surface lights passed out of view, Captain Kukar gently rolled into a 14-degree bank to the right with aileron inputs varying from 9 to 16 degrees during the interval 62 to 70 seconds after liftoff. Captain Kukar relaxed his aileron pressure over the next 13 seconds, and the plane gradually returned to wings-level. Then it continued to roll, slowly, into a 9-degree left bank. At this point Captain Kukar made an abrupt left-aileron input, momentarily reversed his input, then went back to hard-over left. From an altitude of 1,460 feet the big Boeing rolled into a 108-degree left bank, 35 degrees nose down, and crashed into the sea at an airspeed of 330 knots just 101 seconds after liftoff, killing all 210 aboard. Captain Kukar held hard-left aileron and rudder to the end.The partially recovered wreckage revealed no evidence of fire, explosion, or any mechanical or electrical failure, and the initial theory of sabotage was convincingly ruled out. However, the cockpit voice recorder preserved the following barely intelligible words:
At 87 seconds after liftoff, with a 32-degree left bank and rolling left, Captain Kukar exclaimed, "Arey yar, my instrument!" Two seconds later, with the bank now at 47 degrees, First Officer Virmani replied, "My ... mine's also (toppled?) [not clearly intelligible], " and Captain Kukar simultaneously said, "Check your instrument." At 95 seconds, First Officer Virmani repeated [more clearly], "Mine has also toppled." One second later, just 5 seconds before impact, Flight Engineer Fario interjected, "No but, go by this Captain." At 99, Captain Kukar: "Just check the instrument. Yar!" At 101, First Officer Virmani: "Check what?" (Sound of impact.)
For whatever reason, Captain Kukar evidently was confused by the attitude indication of his flight director instrument, and Flight Engineer Fario diagnosed the cause of the Captain's confusion and tried to redirect his attention to another instrument, possibly the turn indicator. One theory advanced during litigation over this accident is that the Captain's attitude indicator stuck just as the airplane reached its maximum right bank of 14 degrees. Then, when the Captain's initial left aileron input failed to restore a wings-level indication, he applied full left aileron until the plane dove into the sea. I do not dismiss this possibility, but there is another explanation with a far higher probability than the sudden failure of an instrument just at the moment outside visual reference was lost.
Every year in the United States, about 100 people die in general aviation accidents when outside visual reference is suddenly lost and the pilot makes a reversed response to the "artificial horizon" and rolls the airplane into a high-speed spiral dive, mis-named "the graveyard spin." An experiment at the University of Illinois showed that of 20 private pilots without instrument flight training who were suddenly deprived of outside visual reference, all lost directional control in an average of three minutes. In trying to maintain altitude, they only tightened their diving turns. Making such bank-control reversals while using a conventional attitude indicator is primarily a general aviation problem. But even professional test pilots occasionally perceive the moving horizon line on such instruments as if it represented the wings of the airplane instead of the fixed reference against which the plane is moving.
In fact, the frame of reference for display presentation has been a subject of controversy from the invention of gyroscopic flight instruments, including turn indicators, directional gyros, and attitude gyros. The controversy, of course, is whether the aircraft symbol or the horizon bar should move with reference to the fixed display coordinates. The issue has not been resolved satisfactorily after more than half a century despite the evidence favoring the moving-airplane presentation from experiments conducted by universities, aircraft manufacturers, the Naval Research Lab, and others over a period of more 40 years. The best of the experiments, a series conducted at the University of Illinois during the early 1970s (reported in detail in the August 1975 issue of Human Factors), provide the most comprehensive and objective tests to date and supply a promising solution to the problem. Bur first a bit more on the problem itself.
A pilot's perception of banked and wings-level flight comes through two sensory organs, the eye and the inner ear. Scientists refer to the information they provide as visual (eye) and vestibular or labyrinthine (inner ear). When one set of sensations contradicts the other, the pilot must choose which to trust, and vertigo may result. A serious problem in flight arises when an aircraft accelerates about its roll axis below the pilot's vestibular sensory threshold. If the pilot's attention is diverted during this time, when he shifts his attention back to the attitude indicator, he will find the display portraying an unexpected attitude. That will result in a conflict between his vestibular sensations, which tell him he is flying straight and level, and his visual sensations, which tell him he is in a banked attitude. Should he initiate a sharp control movement to correct the undesired attitude shown on the display, he will feel as though he is rotating from a wings-level attitude into a bank, when just the opposite is the case. In other words, his eyes will tell him positively that he is correcting an undesirable situation, while his inner ear (vestibular sensations) will tell him positively that he is moving into one. In this situation, the vestibular sensations can be so compelling that the pilot will reverse his visual attitude reference for the brief remainder of his life. This is my theory of what happened in Bombay. Because of these sensory phenomena, it is important that artificial horizon instruments be designed to display the clearest possible relation between the airplane and horizon symbols and their real world counterparts. The late John Poppen, a naval flight surgeon, defended instruments which use a moving horizon display, with the declaration that the correct motion relations would be an exact analog of what the pilot sees though the windscreen in contact flight. Poppen thought of the attitude display as a porthole through which the pilot views a symbolic analog of the horizon. However, Poppen overlooked a psychological phenomenon: While it is true that pilots can learn to see the moving horizon line as the stable element about which the fixed airplane symbol moves -and in fact, do so most the time- it is also true that under conditions of stress and confusion this perceptual relationship can be reversed. For example, an F-86D scope-camera of an air-to-air attack by, of all people, Chuck McDaniel, our chief test pilot at Hughes Aircraft, clearly recorded such a horizon-control reversal. Fortunately, it happened at a high altitude on a clear day. But McDaniel will never forget an experience he could not explain until he also saw the scope movie. And that, he could scarcely believe. Years of research have shown that in the absence of specific training to the contrary, a human's natural or stereotypic response under a wide variety of experimental conditions is to expect a display element to move in the same direction as the control input. (Evidently because we make the control input we identify ourselves with the element that moves.) So when a sudden input results in display motion in the direction opposite to that we expect, the identity of the moving element may become momentarily ambiguous or actually reversed. That is why I believe that to assure correct roll and pitch responses in all circumstances, it is necessary that the airplane symbol be seen as the moving figure against the fixed background of the external world.
Such a dual-movement display is known as a &laqno;frequency-separated» presentation. Although it sounds confusing, it's extremely clear in use. Changes in aircraft attitude, which are relatively slow (low-frequency responses), are displayed as they always have been by a moving horizon indication. The airplane symbol rotates in direct response to -and in the same direction as- the relatively fast aileron control inputs (high-frequency responses). Thus the display's more rapid motion is directionally compatible with the plane's actual motion. Experiments show that new and experienced pilots easily learn to use the frequency-separated presentation. Moreover, simulator and flight experiments involving a wide variety of operationally realistic flight tasks have demonstrated large improvements in flight control and a strong defense against bank-control reversals. I believe that the frequency of design-induced bank-reversal errors is much higher than is generally recognized. In addition to losses in lives and equipment, the consequences of sticking with the moving-horizon display include the need for more pilot training than would otherwise be required. And who knows, in this age of fierce product liability litigation, we may be only a step away from a court decision in which a manufacturer will be found responsible for design-induced pilot errors because ways of guarding against them are scientifically established and yet the manufacturer chose not to implement a change in display.
Read About John F. Kennedy's Probable Design-Induced Spiral of July 1999 Read More About Horizon Control Reversal Read More About Graveyard Spirals
In a White House ceremony, President Bill Clinton awarded psychologist Roger Shepard the National Medal of Science "for his creative theoretical and experimental work elucidating how the human mind perceives and represents the physical world and why the human mind has evolved to represent objects as it does." In 1968 the field of cognitive psychology was dominated by theories of artificial intelligence based on the assumption that all thinking involved the manipulation of discrete mental symbols. But Shepard was convinced that some thought processes are nonsymbolic, that they are more like continuous simulations of external events. Then he hit upon a great idea. Using solid block figures, he embarked on a series of experiments on "mental rotation," later reported in a bombshell paper in Science (Shepard & Metzler, 1971). Shepard and his graduate students showed observers pairs of pictures of objects in different spatial orientations. Sometimes the objects were the same and sometimes not, and they measured the time it took to decide. The greater the difference in the orientation of the two objects, the greater the decision time. It became apparent that the observers were making comparisons by mentally rotating one of the two objects into nearly the same orientation as the other. The time differences even provided an indication of the rate of mental rotation. There was wide variation in the mental rotation speeds of individuals, and the differences were shown to be directly related to how people perceive and interpret complex visual scenes in ev-eryday life or in operating complex systems-as well as in the laboratory. Stimulated by these findings and the advanced experimental methods Shepard introduced during the 1970s and 1980s, mental imagery tasks have received more attention by psychologists than any other kind in recent years. One of the bonus tasks in the WOMBAT test is an adaptation of Shepard's idea. Two three-dimensional figures are displayed side-by-side. Either figure can be selected and rotated about all three axes using the two tracking controls. The task is to determine as quickly as possible whether the figures are identical, mirror images, or otherwise different and calls for a self-assessment of confidence that affects the worth of the answer. This task differs from Shepard's in that the figures can be rotated manually, but there are still wide differences in decision times depending on how closely individuals align the figures versus their facility in mental rotation. Reference Shepard, R. N., & Metzler, J. (1971). Mental rotation of three-dimensional objects. Science, 171, 701-703.