by FRANCIS and KATHARINE DRAKE
LAST year our scheduled airlines suffered 11 major crashes in which 149 passengers and crew members were killed. Non-scheduled charter operators suffered 8 major crashes in which 79 passengers and crew members were killed. These crashes echoed over the nation, frightening the public out of its air-mindedness and out of non-scheduled and scheduled planes alike. It did not matter that on a mileage basis our regular airlines had achieved in 1946 the best safety statistics in their history; that anyone stepping into a scheduled plane enjoyed a 113,000 to 1 chance of stepping out again intact, at the other end; that out of 2,474,000 separate flights, only 8 resulted in passenger fatalities. The headlines, the photographs, were remembered.
The reasons why the accidents happened did not make headlines. They trickled out months later in official findings which the public rarely read. In the meantime the basic issues have been so obscured by smoke-screening and buck-passing that by this time the public is being hypnotized into the belief that super-instruments which activate beams at the approach of danger, or project warning pictures into infallible television screens, are the sure-fire means by which the existing accident rate can be diminished. At least two instruments have been publicized as the way to make flying foolproof— the radafr-operated Ground Controlled Approach (GCA) and its radio cousin, Instrument Landing System (ILS).
It can be categorically stated that neither of these valuable instruments, nor any other pushbutton panacea known to man, could have prevented at least 84 per cent of the flying fatalities which occurred last year on our scheduled airlines. The truth is that the overwhelming majority of these crashes were brought about by avoidable human error, frequently deriving from faulty policy on the part of airline management. Why have crashes caused by human error risen so sharply in the past two years? When so many of them are caused by factors that can be remedied, and remedied at once, it is surely sensible to explore the human situation first, before wandering off into the upper reaches of electronics.
The human error
The great majority of pilots flying our scheduled airlines are the most highly skilled, careful, and responsible flyers to be found anywhere. This goes double for the ground personnel — the mechanics, weathermen, radio operators, dispatchers, and so on — needed to keep a single plane aloft. These two groups have helped to build for us the safest air transport system in the world, notwithstanding headlines.
The remaining fraction, a relatively small percentage in the air and on the ground, do not measure up to the top standards of the majority. In this group we find a minute but critically important number of pilots who, despite a high degree of mechanical skill, are temperamentally unfitted for the sober responsibility of public carrier work. They are the men who cut corners with the regulations, radio back incorrect position reports, dismiss wiser pilots as “sissies,” believe in luck, take chances with the weather and who, ninety-nine times out of a hundred, get away with it. The term for them among conservative airmen is “cocky.” Said a veteran transport pilot recently: “Some lines have had captains flying for them that I wouldn’t have as a fifth assistant junior co-pilot.” It does not need many such men — a fraction of one per cent, is enough — to destroy the reputation that tens of thousands of first-class pilots and ground men have labored for years to perfect.
Another problem that has arisen is due to the employment of war pilots. These men, many of them magnificent flyers, have had a long, tough conditioning to exceptional hazards. Their whole indoctrination was to accomplish a mission, to take the plane through at all costs, including, if necessary, the cost of their own lives. In airlines this indoctrination has to be reversed: never to try to take the plane through if any avoidable risk is involved. A deliberate deconditioning program by the airlines would therefore seem to be a wise precaution. Here are some of the causes of accidents, all drawn from government sources: —
Pilot: Disregarded altitude regulations.
Disobeyed course instructions.
Headed unnecessarily into thunderstorms.
Flew contact under instrument conditions.
Mistook train lights for runway.
Failed to use proper maps.
Failed to check and cross-check position.
Forgot to lower landing gear.
Retracted landing gear instead of flaps.
Reading magazine in cockpit.
Fuel lines carelessly connected.
Forgot to refuel tanks.
Defective landing gear.
Incorrect weather forecast.
Assigned wrong runway.
Obviously, not all these errors stemmed from “cockiness,” but most of them were due to factors which could have been controlled. Let us take, for example, this recurrent business of “plane hits mountain,” accountable last year for 5 major crashes in which a total of 108 people were killed.
On January 31, 1946, eastbound Flight 14, with 18 passengers and a crew of 3, was flying along the wide pass that gouges the Rockies between Rock Springs and Laramie. The night was fair, visibility on the airway good, the plane on course, on time. In less than ten minutes, perfectly operating Flight 14 was dashing itself into the ugly face of Elk Mountain, and a towering pyre of flame became the national headline of the morning: PLANE HITS MOUNTAIN — 21 KILLED.
The findings, published months later, were that the pilot had “deviated from the route and followed the most direct course.” Airways are not designed by caprice. The reason that the airway in question was plotted as an 88-mile dog-leg around this area, instead of as a 79-mile direct course, is that Elk Mountain looms dangerously across the short cut. The pilot, disregarding instructions, turned off the airway, headed into the clouds around Elk Mountain, and ended up 340 feet below the summit. It was a straight case of corner-cutting. Had the pilot got away with his short cut to Laramie, he would have saved less than four minutes from the prescribed flight plan.
This was the first of the 1946 mountain crashes. During the year, here were some other headlines: —
San Diego, Cal. Mar. 3
PLANE HITS MOUNTAIN — 25 KILLED
Lebec, Cal. Nov. 13
PLANE HITS MOUNTAIN — 11 KILLED
San Diego, Cal. Dec. SJt
PLANE HITS MOUNTAIN — 12 KILLED
Planes that crashed were found to have been flying anywhere from 1200 to 3200 feet below the altitudes required by Civil Air Regulations, which call for 1000 feet above the highest obstruction on the airway. In the case of mechanical failure, a pilot, of course, may be unable to maintain height, but no suggestion of such a failure was found by painstaking government experts. Another surmise often put forward is that a plane may have encountered a downdraft. In the case of the first San Diego crash, wrecked at 4870 feet while supposedly following a flight plan of 8000 feet, the Civil Aeronautics Board findings stated: “It is inconceivable that downdrafts could have reached the proportions necessary to cause a 3200-foot loss of altitude under the conditions that prevailed at the time of the flight.”
Cutting corners on a flight plan is a habit that is far more frequent than anyone cares to admit. Airliners formerly carried a flight analyzer, a device familiarly known as the Snooper. This instrument plots a continuous record of course, time, and height on a chart, which is checked on the completion of each trip. During the war the Snooper system — highly unpopular with some airline pilots — was dropped because of unavailability of parts required for maintenance, and it has still not been restored.
The fifth and worst of the mountain crashes, killing 39 persons, took place at Stephenville, Newfoundland, on October 3rd. On the night of the accident the pilot was taxiing out to Runway 300, generally used by the Air Transport Command (in which he had served), when the tower operator, a relatively inexperienced man, directed him to take off from Runway 07 instead. Considering the wind conditions and the high terrain immediately in the flight path, it was a direction no captain should have followed without a major revision of his take-off plan. The pilot neither revised his plan nor challenged the direction. Two minutes after he had taken off, under a 5000-foot ceiling, with 10-mile visibility and a perfectly functioning plane, he had piled up with a shattering explosion against a hill facing the runway, killing everybody aboard. What prompted his whole take-off procedure remains a complete mystery to his fellow pilots, the airline, and the CAB alike.
Altogether, “plane hits mountain” accounted for 108 fatalities, or 72 per cent of the deaths that took place last year on our scheduled domestic and international airlines. In none of these could government experts detect a shred of evidence indicating structural or power failure. Counting in the Cheshire, Connecticut, crash on January 18 (caused by fire resulting from a defect since remedied) brings to 84 per cent the fatalities that could not have been prevented by either GCA or ILS.
The crashes which these two devices might have prevented were 4 in number — all landing accidents. The worst occurred in Shannon, Ireland. On December 27 a planeload of passengers had boarded a shiny new four-engine Constellation in Paris, with every prospect of celebrating the New Year in Manhattan. Over Shannon, where they were to stop for refueling, the weather was bad, and during the pilot’s final approach, the tower was reporting a ceiling of 400 feet, with some clouds at 250 feet. Despite borderline conditions, the pilot decided to land rather than to make for his alternate refueling spot at Prestwick, Scotland. His power was still full on, his altimeter registering 500 feet when the ground — black, viscous Irish bog — reared out of nowhere and flung itself on top of him. PLANE FLIES INTO GROUND —13 KILLED, flashed the headlines. It was established that a careless maintenance stall had reversed the static selector valves on the altimeter tubes, with the result that the instrument was registering 500 feet when the plane was actually about to hit the ground. “Plane flies into ground” accounted for a total of 20 lives, or 13 per cent of the fatalities that took place on our scheduled airlines last year.
The truth is that GCA and ILS are valuable and long-overdue aids to only one department of flying: blind landings. The operating ranges of GCA and ILS are both restricted. Neither will operate in a bend, over mountain ranges or around corners, or under certain storm conditions, and GCA still is unreliable in heavy snow, just when it is most needed. It is significant that the Air Transport Command and the Naval Air Transport Service, both fully equipped with GCA, had 6 and 9 times, respectively, the fatal accident rate of our scheduled airlines last year.
The record shows many cases where regular airline pilots became “unaccountably lost" while flying in average weather conditions over familiar routes. It may well be asked how much of this is due to the failure of airline managements to furnish standard airway maps for their pilots, and to insist on their use. It is a well-known fact that most of our airlines use privately purchased charts and landing diagrams alone. Many pilots admit to flying solely by radio charts, which do not show contours or landmarks, and others navigate by radio alone, in spite of the chance of static confusing receptivity or of beacons going out.
It has been found that cockpit discipline has become lax in other ways. Pilots are known to impair their landing vision by flying with the cockpit lights full on, to absorb themselves in a book or magazine while the automatic flies the plane, to obscure the dials by resting their feet on the instrument panel, to radio down purely guesswork position reports that are as much as 50 miles out. Until such men are systematically weeded out, they will continue to imperil the safety record of the vast majority of transport pilots, to say nothing of the lives of unsuspecting passengers.
Skill, temperament, character
Hiring the man in the cockpit, the man to whose care will be consigned human lives and the reputation of an airline, is the responsibility of management. It is a responsibility nobody need envy. In the early days the main qualification for a good pilot was called “flying sense,” which was supposed to reside in the seat of the pants, and which was a combination of skill and instinct, supplemented by glimpses from an open cockpit. Today a first-class transport pilot combines so many complex qualities that only a very exceptional man can fully qualify for a command. He must have the necessary skill, the necessary temperament, and, above all, the necessary character.
The skill required to obtain mastery over a modern airliner is prodigious in itself, the ability to interpret correct ly a ceaseless stream of technical information flowing from the 66-instrument panel and from the earphones. Recent tendencies in airplane engineering make the job increasingly harder. The concentration required during an instrument landing alone is immense. Hands and feet flying the plane, ears listening to Approach Control, the pilot must keep his eyes glued to six instruments at once: Artificial Horizon, Rate-ofDescent indicator, Turn indicator, Altimeter, Directional and Gyro compasses. With some corner of his mind, he has still to pick the right moment to issue a string of orders concerning flaps, throttles, and landing gear, and through it all keep a sixth sense feeling for an invisible runway rushing towards him at two miles a minute.
The temperament of a pilot is of critical importance during an emergency. It is on his ability to think coolly and independently when mechanical aids fail, when the whole world is blanked out in snow and static and confusion, and on his ability to evaluate instrument readings correctly and come up with all the right answers, that the fate of the plane depends. Such emergencies do not happen often, perhaps a few times a year, but when they do, the margin of allowable error is sometimes razor-fine.
Take the scene in the cockpit when they are plowing through the soup — a scene that may well have existed in those last grim minutes over San Diego or Shannon or Lebec. Jet blackness everywhere. Behind is a black curtain to overpower any stray gleams. In front is the black windshield with little bubbles of fog or icy mush shimmying down the rims. Beyond that, nothing — no nose, no propellers, no wings — nothing. The radio crackles and snaps like a ton of peanuts in a million cellophane bags. If a crisis comes up, it will fail. (That is why the pilot who flew from New York to Washington to Baltimore to Philadelphia to New York, hunting for help, had finally to crash-land on Jones Beach.) Go on? Turn back? Try somewhere else? How are the runways on the nearest alternate, how high are the hills, how good are the lights? How is the fuel reserve? Has the wind strengthened? That mush is hardening on the windscreen — going to ice up? Rushing through the air at 300 miles per hour, with thirty frightened passengers in the cabin behind, and death the certain penalty for a mistake in judgment, it takes a very exceptional man to remain unruffled and keep on making all the right moves.
Nor arc skill and temperament sufficient in themselves. Our good transport pilots are men of a certain uniform character —— men in whom a sense of responsibility and a remorseless conscience are wedded to a healthy respect for rules, even when they may privately believe a specific rule unnecessary. Some miracle instrument of the future will doubtless keep tabs on the cockpit through a television screen, but for the present there is no monitor there save the pilot’s own conscience.
No instruments will ever lessen the need in transport captains for these last two qualifications, but there are things that could be done, even today, to make the pilot’s lot mechanically a better one. Many of the existing runways are not thick enough to take the weight of big carriers, and the Civil Aeronautics Administration has stated that over 150 of our 250 designated airports are deficient in length and clearness of approach. Most of the runways were built over fifteen years ago, before the days of four-engined planes, and have not been improved since 1941. Among other things that would be of invaluable help to pilots are Very High Frequency radios, high-intensity approach and runway lights, and a slashing simplification of Rules. So thick and complicated is the CAR volume today, so crammed with confusing fine-print regulations and local corollaries, that there are times when a pilot literally makes an approach “with the book in one hand and the airplane in the other.” This move for simplification of rules and instruments is long overdue.
What management can do
The chief factor in the fight for safety is necessarily management. Management frames the policies, inspires morale, buys the planes, hires the pilots, sets the schedules, controls the advertising. Frustrated through the war years, management flung itself on the first opportunity for expansion with a vehemence which has all but wrecked the industry. It plunged ahead with insufficient planes, crews, runways, airport facilities, traffic controls, navigational aids. It slashed 10 per cent off fares, raised passenger mileage from 4 billion in 1945 to 7.2 billion in 1946, and, although starved for equipment with which to handle existing business, filed applications for scores of thousands of miles of new routes. In this headlong course management was abetted by the government, which encouraged the moves and matched the fare cut with a slash in airmail payments. As a result of all this shortsightedness, some of the airlines are now in financial trouble.
To achieve their remarkable expansion in traffic, airlines had to increase time in the air for every plane, decrease time on the ground for maintenance and overhaul, and round up large numbers of new air and ground crews to handle the flights. During this process general efficiency sank to around 70 per cent of normal, and half the employees had seen less than a year’s service in the airlines.
One after another, the elements of too hasty expansion have boomeranged on the airlines. It is becoming increasingly clear that all the whoopde-do in the world will not induce the public to accept planes as routine transportation unless the accident rate can be brought down to a reasonable comparison with ships, trains, and buses.
Here is the comparison for 1946: —
|Estimated Deaths per 100 Million Passenger||Miles|
|Scheduled air transport||1.6|
|(Domestic only: 1.2) Non-scheduled charter planes (Government estimate)||25.0 to 35.0|
Thus, it was 5 times more dangerous to fly on a regular airliner than it was to take the train or bus, and between 110 to 150 times more dangerous to fly on a non-scheduled charter plane — a difference for which the scheduled airlines seldom get the credit, and which underlines the necessity for equal regulation of the charter planes.
The first step towards better operating policies must be a general overhaul of management, a step already initiated by some of the airlines. Time is required for objective thinking and planning, for a more realistic approach to schedules, publicity, fares, and airmail payments, for the setting up of pooled maintenance, traffic, passenger and baggage facilities, and, above all, for devising a sounder system of selection and ‘payment of airline pilots.
It is probable that the greatest single step to advance safety would be the setting up by the airlines themselves of a national board of senior transport pilots to examine the qualifications of every airline captain appointed within the last five years. Certain minimum standards are already set by the CAB, but a board of transport-wise airmen is needed to evaluate the extremely high standards of skill, temperament, and character required of an airline pilot.
Such a board could be created immediately, without government aid, through the Air Transport Association. Simultaneously airlines should discontinue the pernicious system of making captains’ salaries dependent on their hours in the air. For no justifiable reason, their small base pay is augmented by flying pay, which means that every lime a pilot’s weather sense advises him to cancel a trip, he must automatically kiss a good many dollars good-bye. This practice, coupled with the airlines’ infatuation with impractical schedules, has imposed considerable pressure on the judgment of pilots.
Specific steps necessary to achieve improved safety would seem to imply a two-part program, based on some such structure as the following: —
For Immediate Implementation
1. Overhauling of airline managements.
2. Screening of all pilots by an independent air-
3. Radical simplification of Civil Air Regulations.
4. Installation of flight analyzers in all transport
5. Insistence on standard airway maps in all cock-
For Implementation As Soon As Possible
1. Very High Frequency radio.
2. High-intensity lighting for approaches and run-
3. Longer, wider, heavier runways.
4. GCA, ILS, and landing aids.
5. Automatic devices, wherever possible, to lessen
the pilot’s need for manual operation.
6. Commensurate financial support by government.
The responsibilities of airline management are no greater than the brilliant opportunities that lie immediately before it. New equipment — huge multi-engined planes carrying large reserves of fuel, with pressurized cabins enabling flights to be conducted high above bad weather, with cockpits equipped to utilize the latest electronic aids now being installed — is opening up vistas that are at once splendid and inspiring. However, on old and new equipment alike, a conservative handling of the human factors must result in degrees of safety and regularity that will be above all reasonable criticism.