|
Article Excerpt
INTRODUCTION
Spatial disorientation (SD) during flight (Benson, 1978a, 1978b) is a cause of critical aircraft mishaps with aircraft and pilot losses. It can be characterized as depending on "an erroneous sense of one's position and motion relative to the plane of the earth's surface" (Gillingham, 1992, p. 297). It has also been described as a situation "where the aviator fails to sense correctly the position, motion or attitude of his aircraft or of himself within the fixed coordinate system provided by the surface of the earth and the gravitational vertical" (Benson, 1978b, p. 405). Information about spatial orientation (SO) is primarily provided by vision, proprioception, and the equilibrium sense (the otolith organs and the semicircular canals). In visual meteorological conditions (VMC), vision is mostly correct when body orientation can be related to the externally fixed world frame of reference. The equilibrium sense and proprioception, however, assume that the G force is vertically directed, which is often incorrect in the flight situation (i.e., gravity or g vs. gravity-inertial or G). Thus, especially when visibility is bad, the equilibrium sense and proprioception may dominate perceived SO and generate numerous illusions causing fatal pilot actions (Gillingham & Previc, 1993). These illusions are most often overcome if visibility of the ground and horizon is good.
The documentation of the U.S. Air Force Class A mishap rates reveal that the SD mishap rate per 100,000 flying hrs has remained relatively constant over the past 30 years (1972-2000). Over the same period, the total Class A mishap rate has been reduced significantly (Heinle & Ercoline, 2002). Thus the need for more effective SO visual supports has been obvious to the aviation domain for some time, and the significance of ecological interfaces has been emphasized from various perspectives (e.g., Flach, 1999; Malcolm, 1984; Previc & Ercoline, 1999). However, displays for central vision providing information about SO have to be attended to and cognitively interpreted, and they do not elicit the natural mode of ambient spatial perception. Procedures of checking traditional flight instruments such as horizontal gyro, altimeter, and head-up display symbology rely on directed attention that uses up resources in competition with other attention-demanding tasks. Consequently, when pilots fly under stress in instrument meteorological conditions (IMC), maintaining SO exclusively by means of these visual interfaces, the risk for SD accidents increases despite intense training, experience, and hammered-in instructions to fly by the instruments. The bottom line is that the pilot's perceptual processing is not in contact with the crucial factors that effectively contribute to the overcoming of spatial illusions.
The optokinetic cervical reflex (OKCR; Patterson, Cacioppo, Gallimore, Hinman, & Nalepka, 1997) is an indication of how little the flight instruments of today resonate with, or trigger sensory reflexes of, the SO mechanism. The OKCR expresses itself as a lateral tilt of the head during aircraft rolls in VMC, whereas in IMC the head maintains alignment with the vertical axis of the aircraft. The eliciting factor seems to be the perception of the true horizon, whereas horizon symbology on an attitude indicator or head-mounted display (HMD) fails to elicit it (e.g., Liggett, 2002). The OKCR is thus related to shifts in frames of reference between IMC and VMC (Johnson & Roscoe, 1972). In VMC, the spatial frame of reference lies outside the aircraft, whereas it is situated inside the aircraft in IMC, and transitioning between the reference frames of the outside world and the aircraft-framed instruments may lead to orientation conflicts causing SD (Liggett, 2002; Patterson et al., 1997).
Vision can be functionally dichotomized into a focal and an ambient subsystem (Creem & Profitt, 2001; Leibowitz & Post, 1982; Schneider, 1969). The ambient system processes the entire visual field, whereas the focal system gets its information from parts of the central visual field. Among other things, the ambient system is associated with SO, whereas the focal system is more specialized in perceiving objects and external events. The entire ambient visual field contributes to the perception of SO and is typically not dependent on attention resources. Previous research has shown that information for SO is primarily provided by peripheral or ambient vision and is largely independent of focal vision (Johansson, 1977; Johansson & Borjesson, 1989; Leibowitz, 1988; Leibowitz & Post, 1982). In addition, central scotoma (e.g., macula degeneration) does not degrade the ability to maintain SO, whereas the reverse phenomenon of damage to the peripheral retina severely deteriorates SO (e.g., Kolb & Whishaw, 1990).
A visual aid that resonates with the mechanism normally underlying SO has the potential of reducing or counteracting SD significantly. To optimize the support of pilot SO, it is therefore necessary to go beyond central visual field presentations and better utilize the capability of the ambient visual system. Accordingly, engineering psychologists have argued that conventional aircraft navigation instruments are relatively ineffective because they are restricted to foveal vision and thus ignore the usefulness of the visual periphery (Wickens & Hollands, 2000). An exception is the Malcolm Horizon (Malcolm, 1984). This device projects a gyro-stabilized "thin line of laser light representing the real horizon onto an aircraft's instrument panel" (Gawron & Knotts, 1984, p. 539). However, this line representation of the horizon projected into a relatively minor part of the wide-angle visual field at noninfinity on the instrument panel probably made its effectiveness not entirely convincing. An effective visual interface has to be in better resonance with the natural mode of perceiving SO to reduce the risk for SD mishaps, and the aim should be to reestablish the dominance of vision and counteract the erroneous influence of proprioception and the equilibrium sense.
It has long been recognized that vision can be regarded as "an autonomous kinaesthetic sense" (Lishman & Lee, 1973, p. 294). An expanding visual flow elicits perception of forward ego motion (e.g., Andersen & Braunstein, 1985), and a rotational flow elicits perceived observer rotation (e.g., Dichgans & Brandt, 1978). This phenomenon is often referred to as vection (Sauvan & Bonnet, 1995; Telford & Frost, 1993) and constitutes one of three more narrowly defined topics within visually induced self-motion. The two other topics are the perception of direction of self-motion (i.e., "heading"; Warren, Morris, & Kalish, 1988) and postural control (Berthoz, Lacour, Soechting, & Vidal, 1979; Fli.ickiger & Baumberger, 1988; Stoffregen, 1986). In essence, optic flow (Gibson, 1950; Lee, 1980) is an effective source of spatial information that has been shown to dominate information input even from other senses (Lee & Lishman, 1975; Lishman & Lee, 1975).
Von Hofsten and Rosander (1997) proposed a wide-field display consisting of an artificial horizon corresponding to the real one and enhanced with a flight-adapted visual flow of a synthetic ground beneath. An important and critical part of such a display is to get it firmly anchored in the external world by compensating for movement of pilot and aircraft (e.g., requiring high-resolution and high-update-rate head-tracking and image control). To produce an optimal design one needs to explore a number of parameters, and some of the basic ones have to do with how much of the visual field should be covered by the display. For several reasons it is important not to use the pilot's central visual field for this kind of presentation but to leave it open for other varieties of information crucial for the pilot's tasks. Because foveal vision is of minor importance when it comes to perceived SO, this part of the visual field can be left out. An...
|
