The Disorient Express

Despite the best training and technology, why do pilots still die from not knowing which end is up?

AS08 Vertigo 631x300.jpg
At Wright-Patterson Air Force Base in Ohio, a study subject is wired for a spin in the Dynamic Environment Simulator, a centrifuge that excels in inducing spatial disorientation. DEPT OF DEFENSE

On June 26, 2007, while on a training exercise off the Oregon coast, Major Gregory D. Young of the Air National Guard flew his F-15A fighter into the Pacific Ocean. The $32 million aircraft was destroyed and the pilot killed. There was no distress call, no attempt to eject, and no apparent aircraft malfunction. Young, 34, had 2,300 hours of flight time, more than 750 hours of it in F-15s.

As investigators sifted through the wreckage—what little was left—colleagues, family, and friends were left to wonder: What caused Young to guide his airplane right into the ocean at more than 600 mph? The answer, revealed in an investigative report two months later, was both profoundly unsettling and all too familiar. Young, in the prosaic terminology of the report, “experienced unrecognized (Type 1) spatial disorientation (SD), which caused him to misperceive his attitude, altitude, and airspeed. As a result, [he] was clearly unaware of his position and impacted the water.”

In other words: Young never knew what hit him.

Despite training, experience, and technology, all based on knowledge of how flight affects human physiology, Young had no idea that he was racing downward.

Once called pilot vertigo or aviator’s vertigo, spatial disorientation is a persistent killer. Federal Aviation Administration statistics show that the condition is at least partly responsible for about 15 percent of general aviation accidents, most of which occur in clouds or at night, and 90 percent of which are fatal. According to a 2004 study, the average life expectancy of a non-instrument-rated pilot who flies into clouds or instrument conditions is 178 seconds.

A U.S. Air Force review of 633 crashes between 1980 and 1989 showed that spatial disorientation was a factor in 13 percent, resulting in 115 deaths. Among crashes of high-performance aircraft, the rate was higher: 25 to 30 percent. A U.S. Navy study found that in contrast to general aviation accidents, a majority of accidents in high-performance aircraft occurred in daylight and in visual flight conditions. The pilots were an average of 30 years old, with 10 years in the cockpit and 1,500 hours of pilot-in-command or instructor time, and in the prior three months they had flown an average of 25 times—all of which shows that no amount of expertise, training, or experience immunizes against spatial disorientation.

Humans maintain orientation and posture through a system of senses: vision; the vestibular system (the labyrinthine series of ducts and canals in the inner ear); muscle-sense or proprioception, sensors in muscles and joints that inform us of our body’s position (standing versus sitting, for example); and the sense of gravity, or what we perceive as up and down. The system has evolved over eons, and is well adapted for Earth. But it is easily fooled. When you’re sitting on a stopped train and the train on an adjacent track begins to move, you’ll think that you’re the one who is moving.

In the air, things get more complicated. Early aviators were confronted by an assault on their senses, or “disturbances of equilibrium,” as Orville Wright described it. Until World War I, most flights were made during the day and limited to short, straight-and-level hops. Few risked flying at night, and fewer still flew into clouds, or at least did and lived to tell about it.

Research in the 19th and early 20th centuries helped shed light on the vestibular system and how it maintains equilibrium. In 1906, Robert Barany devised a swivel chair to simulate the effects of spatial disorientation on pilots.

At an FAA-sponsored safety seminar in Rhode Island recently, program manager Jack Keenan offered me a seat in a Barany chair, a device not unlike a barber’s chair, with an ersatz control stick. As a group of other pilots stood around, he blindfolded me and began to spin the chair to the left, telling me to move the control stick in the direction of the spin. I dutifully moved the stick to the left.

As Keenan gently turned the chair, he said to the group, “Your body keeps you alive. We learn to recognize cues from our environment.” The problem, he added, “is that our bodies are meant to walk on Earth.” As he spoke, the chair seemed to quit spinning. I moved the stick to the neutral position. Keenan then rattled off a litany of phenomena ready to befall pilots: the leans, the graveyard spiral, the inversion illusion, the elevator illusion, false horizons (see “Vertigo: A Primer,” below). As he spoke, the chair then seemed to reverse direction, spinning to the right. I moved the stick to the right. Some in the group tittered. Keenan then pulled off the blindfold, and I saw that the chair had stopped. “Get up carefully,” warned Keenan, helping me to my feet. “You’re still spinning.” Three other volunteers followed my lead. In each case, Keenan first got the chair spinning. After a bit, he gently brought the chair to a stop. In each case, the volunteer moved the stick exactly as I did.

The inner ear is designed to detect motion, or rather, acceleration. Thus, when the chair began to turn, I sensed it. However, once the turn rate was constant, the fluid in my inner ear returned to equilibrium, and without the benefit of visual cues, I could not tell the difference between turning and sitting still. So when the chair stopped turning, I sensed that as a turn in the opposite direction.

Vestibular illusions fall into two categories: somatogyral, for spinning illusions (“somato” is Greek for “body”), and somatogravic, for acceleration illusions. The Barany chair demonstrates a basic somatogyral illusion. An airplane in a stable, level turn will feel the same as an airplane in straight-and-level flight. If the airplane is returned to straight-and-level flight, or if the bank is decreased, a pilot’s natural reaction would be to make a correction that would steepen the actual turn. If at the same time the pilot’s head were tilted—if he were reading a map or picking up a pencil—the deception to the vestibular system would be compounded along a third axis, meaning that when the airplane returned to straight-and-level flight or the pilot lifted his head, he would sense not only a turn in the opposite direction but a feeling of pitching up or down.

Somatogravic illusions refer to situations in which an airplane that begins accelerating will feel the same as one climbing, and an airplane decelerating will feel the same as one descending. Because we live on the surface of the Earth, where the force of gravity pulling us toward the ground is more or less constant, or 1 G, our vestibular system cannot distinguish the difference between pitch and acceleration. Today’s full-motion simulators take advantage of this fact to create the illusion of flight. For example, inside the simulator pod, as the pilot moves the throttles forward for takeoff and sees and feels the “airplane” accelerating down the runway, the pod itself begins to tilt up. The motions created by the simulators are so realistic that students have become airsick in them.

In 1917, Elmer Sperry invented the gyroscopic turn indicator, based on a similar device he had invented for ships. The indicator joined Sperry’s gyroscopic compass to make up what would later be the core of the panel for instrument flight. But as late as 1928, the idea of flying solely by reference to instruments—“flying blind”—remained as foreign as travel to other planets. Pilots were convinced that their most valuable tools  were skill and instinct.

In 1926, Army Air Corps Captain William Ocker, who had been experimenting with Sperry’s turn indicator, took a medical exam that included a spin in a Barany chair to test his vestibular system. Experiencing the same spinning illusion, he had the revelation, writes William Langewiesche in Inside the Sky (Pantheon, 1998), “that instinct is worse than useless in the clouds, that it can induce deadly spirals, and that as a result having gyroscopes is not enough, that pilots must learn against all contradictory sensations the difficult discipline of an absolute belief in their instruments.” Ocker, with the zeal of a fundamentalist minister, began preaching the necessity of developing procedures and instructional programs in instrument flight. He was unable, however, to convert his superiors. Twice the Army had Ocker hospitalized to test his sanity. (In 1932, a vindicated Ocker coauthored the first treatise on instrument flying, Blind Flying in Theory and Practice.)

In 1927, a group of scientists and pilots that included Sperry and Army Air Corps Lieutenant James Doolittle built an artificial horizon, a gyroscopic device that gives the pilot a graphic representation of the airplane’s attitude in relation to the horizon. Doolittle used it in 1929 to make the first flight and landing solely by reference to the aircraft’s instruments, proving the feasibility of instrument-only flight.

Making trustworthy instruments was one thing, but making pilots trust them was another. At first, pilots reported the instruments seemed to work only in clear weather, that in clouds the devices went haywire, indicating turns the pilot was certain the airplane was not making. The instruments worked just fine; the pilots had to be taught to resist the instinct to fly “by the seat of their pants”—that is, by sensation alone.

Today, primary flight training for all pilots requires instruction in flight based on instruments and recovery from unusual attitudes, in which the flight instructor has the student close his eyes while the aircraft goes through a disorienting series of turns, climbs, and descents, then has the student return the airplane to straight-and-level flight. Military aviators, in addition to being subjected to periodic proficiency reviews, are required to attend, every five years, refresher courses in human physiology that include a section on spatial disorientation.

Rogers Shaw, a director at the FAA Civil Aeromedical Institute in Oklahoma City, admits that training exercises such as unusual-attitude recovery are limited by the fact that the student knows and expects to have to make a correction to return the airplane to straight-and-level flight. Spatial disorientation is so insidious, and the sensations it creates so compelling, that unless you suspect you have a problem, you would never know there is one. Unlike other airborne emergencies—an engine quitting, loss of electrical power, smoke in the cockpit—there’s no principal event to indicate anything is wrong. If the pilot does realize something is not quite right, he may react too late, or in a way that aggravates the situation. Or, as in the case of Major Young, the pilot may not react at all.

The crash of John F. Kennedy Jr. on the night of July 16, 1999, off the island of Martha’s Vineyard,  which killed him, his wife, and her sister, brought public attention to the consequences of spatial disorientation. The investigation of an air crash, says Richard Bunker of the Massachusetts Aeronautical Commission, who investigated the Kennedy crash for the state, is a process of elimination. You start with the airplane. After eliminating structural or mechanical problems, you look at external factors, such as weather.

Then the investigation turns to the pilot. You examine his or her training and experience, medical history, personal life, and possible extenuating factors. Eventually, Bunker says, the evidence and the circumstances point to “well, maybe we’re looking at spatial disorientation.”

Kennedy did not have an instrument rating. He was flying at night over water with visibility as low as three miles in haze, meaning there were few lights and no visual horizon for reference. About 10 miles from Martha’s Vineyard, he deviated from course and made a number of maneuvers suggesting he was lost or disoriented. The final radar track showed the airplane in a tightening right-hand turn—called a graveyard spiral—that reached a descent rate exceeding 4,700 feet per minute before the airplane hit the water.

In the case of Major Young, it was all over in less than a minute.

Young, call sign Grumpy One,  flew the lead aircraft in a formation of two F-15s in a combat exercise against two F/A-18s over the Pacific Ocean, about 50 miles west of Cape Arch, Oregon. The visibility was 10 miles or greater, with the horizon discernible in all directions.

While Young’s wingman, Lieutenant Colonel Paul Fitzgerald, call sign Grumpy Two, engaged the two F/A-18s, Cowboy One and Two, Young began a climbing right turn that peaked at 18,800 feet, then began descending in the direction of his wingman and the other two aircraft.

As he did so, Cowboy Two, having maneuvered into position behind Grumpy Two, radioed over a common frequency monitored by all the pilots that he had “killed” one of the two F-15s.

By now, Young’s descent rate had nearly doubled, to 30,000 feet per minute, and he was nearing 5,000 feet—a floor set for the exercise to allow for a margin of safety; at that altitude, Young should have broken off the engagement.

Eight seconds later, Young’s airplane hit the water. Young’s wingman told investigators that all he saw was “a big white splash that reminded me of Niagara Falls.”

Young’s remains were recovered along with some of the wreckage, the pieces of which were no larger than “a small trash can,” in the words of the accident report. With the airplane almost completely destroyed, analysis of the engine and airframe was limited to a review of maintenance records and interviews with ground personnel. These things, along with the fact that Young had never indicated a problem and the airplane had performed as expected, strongly suggested that the problem was not mechanical. (Coincidentally, a few months after Young’s crash, a Missouri Air National Guard F-15C broke apart in flight, setting off a fleet-wide grounding of F-15s to investigate failing longerons.)

With the airplane’s flight data recorder also destroyed, investigators were limited to reconstructing the flight path using radar tracking data, videotapes of the other airplanes’ head-up displays (which project critical flight information on a transparent display above the instrument panel), and data from their flight recorders, in addition to the testimony of the other pilots. Investigators determined that Young’s airplane hit the water at an angle of 24 degrees at a speed of 630 mph.

The report says that Young’s helmet showed he was sitting head-up, indicating he was likely conscious at the time of impact. Analysis further suggested Young was looking up and slightly to the right, not at the ocean in front of him, at his head-up display, or at his instruments. His G-suit was not fully inflated, indicating that he was not pulling significant Gs to arrest his descent.

Increasingly, the evidence pointed to spatial disorientation.

As Young went from climb to descent in his final maneuver, he would have been susceptible to a somatogravic illusion making his dive angle seem much shallower than it actually was. He may, in fact, have thought he was inverted. The fact that his rate of descent increased significantly in the final seconds indicates that Young “may have even believed he was climbing in the final moments, although he was actually still descending,” the investigators’ report said.

In addition to primary flight data (attitude, airspeed, altitude, heading), the head-up displays in military cockpits provide the pilot a continuous view of what is directly in front of the aircraft. Displays also project flight information on the helmet visor so the pilot’s head is free to move. Three-dimensional “highway in the sky” displays give a pilot’s-eye view of the terrain and project a path to follow. Today’s pilots can maintain a level of situational awareness that their predecessors never dreamed of.

But when it comes to countering spatial disorientation, the new displays create their own problems, says Bill Ercoline, a scientist at California-based Wyle Laboratories who provides human factors research for the Air Force Research Laboratory at Brooks City-Base in Texas. Studies of unusual attitude recovery using head-up displays found that HUDs can actually interfere with recovery. The field of view is narrow, the manufacturers use symbols that are not universal, and the nature of the displays is not intuitive; compounding all that, there’s simply too much information to process. “It’s like drinking through a fire hose—it’s just difficult to get the right gulp,” Ercoline says. With so many more systems to manage and monitor, pilots end up devoting less time to actually flying.

The Air Force commissioned a team, led by NASA and the Air Force Research Laboratory at Ohio’s Wright-Patterson Air Force Base, to develop an autopilot that engages when the pilot is unconscious or unaware that he is about to hit the ground. The Automatic Ground Collision Avoidance System— Auto-GCAS—evaluates a variety of factors (aircraft weight and performance, navigational information, terrain and elevation data) to constantly calculate the aircraft’s position, time before impact, and maneuver required to prevent an impact. When the system determines that the airplane is within 1.5 seconds of the point of no return and the pilot still has taken no action, it will take control and perform an automatic rescue maneuver. The system, developed and tested over the past two decades, is now ready for use with F-16 and F-22 fighters. The Department of Defense hopes the system will virtually eliminate “controlled flight into terrain” crashes due to spatial disorientation or G-induced loss of consciousness.

While Auto-GCAS will certainly help, says William Albery, a senior scientist at Wright-Patterson, it won’t completely eliminate spatial disorientation.  Susceptibility to vertigo will continue, he says, as long as there are human pilots on airplanes, and even pilots not in airplanes—in several incidents, pilots who remotely fly aircraft have lost control due to vertigo. The only way to completely eliminate the problem, he says, is to develop fully automated aircraft.


Vertigo: A Primer

Spatial disorientation is classified into three types.

-- Unrecognized spatial disorientation (Type I) refers to situations in which the pilot fails to perceive a change from the desired orientation.

-- Recognized spatial disorientation (Type II) occurs when the pilot realizes there is a conflict between the flight instrument readings and what his body senses is the spatial orientation.

-- Incapacitating spatial disorientation (Type III) refers to situations in which the physical symptoms accompanying disorientation — visual impairment, muscle spasms, nausea, or panic — are severe enough to incapacitate the pilot.

Among the illusions pilots may experience:

The Leans  A somatogyral illusion in which, after a prolonged, gentle turn followed by a sudden return to level flight, a pilot will sense a turn (bank) in the opposite direction. A pilot experiencing the leans may lean in the direction of the original turn in an attempt to regain the perception of the correct vertical posture.

The Coriolis Illusion  A somatogyral illusion in which, while the aircraft is turning, a pilot tilts his head — say, to read a map. When the head is tilted out of the plane of rotation, the pilot will experience a sensation of rolling. Depending on the nature of the turn, the pilot may also experience a sensation that the aircraft is pitching, yawing, or both.

The Graveyard Spiral  Unaware the airplane is banked but sensing the nose drop and a loss in altitude, a pilot may pull back on the yoke to try to regain altitude or slow the rate of descent. The increase in back pressure on the control yoke usually results in a tighter turn and a drop of the nose, causing a further loss of altitude. The sequence may continue until the airplane stalls, breaks apart, or hits the ground.

The Inversion Illusion  A somatogravic illusion in which, after a sustained climb in a high-performance aircraft, the pilot levels the aircraft, creating a lighter “seat bottom” sensation while the acceleration maintains the seat-back pressure. The sensation is that of the aircraft continuing to increase in pitch. Soon the pilot perceives the aircraft is inverted.

Get the latest stories in your inbox every weekday.