More About Aircraft Attitude and Steering Displays

MORE ABOUT AIRCRAFT ATTITUDE
AND STEERING DISPLAYS

This discussion of the problem of displaying aircraft attitude and guidance information is adapted from a chapter by Stanley N. Roscoe in a forthcoming book titled "Human Performance in General Aviation," edited by David O'hare and published by Ashgate in Aldershot, England.

ERGONOMIC COCKPITS

From the earliest blind flying experiments by Lt. Jimmie Doolittle with the original gyro-stabilized Sperry horizon there was controversy over whether the airplane symbol or the horizon bar should move relative to the fixed display coordinates (Roscoe, Johnson, & Williges, 1980). Those with "common sense" argued that the "artificial" horizon bar should move to maintain "congruency" with the real horizon (Poppen, 1936), and that's the way Sperry built it. But Doolittle and other early-day instrument fliers had trouble seeing it that way and remembering that the fixed "little airplane" was what they were supposed to control and not the longer, moving horizon bar.

All pilots do learn to control flight attitude by reference to the artificial horizon, and by the time they qualify for an instrument rating they have long since learned the correct responses to the display's indications. But they still see the horizon bar as the part of the display that moves and not the little airplane symbol, and in the perceptual confusion of a sudden, unanticipated entry into an unusual attitude, there is a strong tendency to control the part of the display that is moving, not the part that is fixed. They naturally expect the moving part to move in the same direction as the movement of the controls, as in steering a car or a bicycle.

The time has past when it would be reasonable to consider reversing the control/display relationships in flight attitude indicators. In fact, the reversed relationship is not the best. A much easier change results in an even better display, one that would be trivial to implement in modern planes with "glass cockpits" and more difficult but not unreasonable with electromechanical instruments. A flight path predictor can be added to the conventional moving horizon display by allowing the airplane symbol to move in immediate response to control inputs and in the same, expected direction. It sounds simple, but the effect is magical.

Conventional moving horizon display with the addition of a flight path predictor and an angles-to-turn-through command symbol for landing-approach or en route guidance.

An attitude/director display with a flight path predictor (as shown above) was tested extensively in both airplanes and flight simulators at the University of Illinois starting in the 1970s (Roscoe, Johnson, & Williges, 1980). In addition to responding immediately to control inputs, changes in flight path direction were indicated by having the airplane symbol move laterally from the display center parallel to the horizon in proportion to the rate of turn. This creates a superior flight-director presentation in conjunction with a steering command symbol. Later a flight path predictor was integrated in the computer-animated visual system of a primary training simulator, resulting in immediate improvement in landing performance, as shown below (adapted from Lintern, Roscoe, & Sivier, 1990).

Initial performances of independent groups of pilot trainees learning to land a flight simulator. The unaugmented pictorial display was a computer-animated view of an airport scene that was dynamically responsive to the changing attitude and flight path of the simulated airplane. The principles applied to the augmented displays included direction-of-motion compatibility with flight path prediction and a simplified steering command indication.

The experiment dealt with the performances of independent groups of ab initio pilot trainees learning to land a flight simulator. An unaugmented pictorial display that served as a control condition presented a simplified view of an airport scene that was responsive to the changing attitude and flight path of the simulated airplane. An augmented pictorial display included, in addition, an integrated pictorial presentation of flight path prediction and guidance symbology with direction-of-motion compatibility and optimum scaling. An augmented symbolic display embodied the same principles except that they were incorporated in a symbolic format with no pictorial background scene.

The experiments at Illinois have covered a wide range of operationally realistic contact and instrument flight tasks. Not only do beginning students learn to land airplanes and fly by instruments more quickly, but also their terminal performance is far more precise. The latter is also true for experienced instrument pilots who have no trouble taking advantage of the flight path predictor without having to unlearn their overlearned responses to the moving horizon display. And because pilots of highly automated planes do little hand flying these days, they will welcome the assistance of a flight path predictor when they do have to take control.

The perceptual-motor problems encountered by pilots during the evolution of aircraft instrumentation have received serious experimental attention during the half century since World War II. The beauty of the research is that we now have several established display design principles that have application far beyond the immediate settings of the individual experiments. The cumulative benefit of the application of the various principles is not merely the additive sum of their individual benefits-it is more akin to their mathematical product, as suggested by the findings from the experiment described above.

The display principles that make integrated pictorial aircraft attitude and flight path displays so effective are also applicable to downward-looking map-type navigation displays with integrated traffic and terrain information and integrated altitude, vertical flight path, thrust, and speed displays. Unfortunately, these principles have not been applied consistently in modern jet airplanes with "glass cockpits" or in general aviation planes. If they were, huge reductions in training requirements would follow, with greater safety and productivity and a sudden competitive advantage for the first to adopt this scientifically based approach.

References

Lintern, G., Roscoe, S. N., & Sivier, J. E. (1990). Display principles, control dynamics, and environmental factors in pilot training and transfer. Human Factors, 32, 299-317.

Poppen, J. R., (1936). Equilibratory finctions in instrument flying. Journal of Aviation Medicine, 6, 148-160.

Roscoe, S. N., Johnson, S. L., & Williges, R. C. (1980). Display motion relationships. In S. N. Roscoe (Ed.), Aviation psychology (pp. 68-81). Ames: Iowa State University Press.