Recommendation

SUBJECT - POSITIONING AND USE OF LIGHTBARS ASSOCIATED

WITH GLOBAL POSITIONING SYSTEMS IN AERIAL AGRICULTURAL USE





INTRODUCTION



Background





The lightbar unit of a global positioning system (GPS) "flagger"
provides pilots with accurate tracking for successive swathe runs
by means of a series of lights, portraying the alignment of each
swathe run. This system replaced almost all ground marker
personnel, whose use under the occupational health and safety
legislation of the majority of States became considerably more
restricted.



In 1995, the Bureau of Air Safety Investigation was alerted to
possible safety implications resulting from difficulties
experienced by pilots in the use of GPSs then being introduced into
aerial agricultural operations.



A safety deficiency raised at that time identified a lack of
standards in the training and use of satellite systems employed in
aerial agricultural operations.



In May 1996, BASI issued Interim Recommendation IR960016, to the
Civil Aviation Safety Authority (CASA), recommending that CASA, in
conjunction with the aerial agricultural industry, formulate a
policy covering the use and training of GPS in aerial agricultural
operations.



In response to this recommendation, CASA in October 1996 advised
that the Aerial Agricultural Association of Australia (AAAA) had
offered to provide two or three experienced pilots to assist in
developing a standard for the use of the GPS in aerial agricultural
operations. BASI was also advised that CASA would consider taking
up this offer at a later date as the GPS Implementation Team (GIT)
progressed with its work. This matter was subsequently addressed by
the GIT. Their decision was that because it related to a
specialised area, it needed to be dealt with by CASA and the
relevant industry body. To date, BASI has not received notification
that this has been done.



CASA also stated in its response that the AAAA had advised that
"the introduction of GPS into agricultural work has been relatively
free of incidents and we would be interested to learn of BASI
reports, if to the contrary".



In May 1998 another deficiency was raised, as a result of several
occurrences, indicating that the lightbar unit used in conjunction
with the GPS "flagger" may have distracted pilots during spraying
operations. Subsequently, BASI reviewed the human factors issues
associated with the positioning and use of the lightbar unit.





SAFETY DEFICIENCY



The mounting position and colour of a lightbar used in conjunction
with a GPS for aerial agricultural operations may be distracting
and reduce a pilot's ability to see and avoid ground-based
obstacles and aircraft warning lights.





FACTUAL INFORMATION



GPS-related agricultural occurrences



The use of GPS with lightbar units has been implicated in a number
of recent accidents in agricultural operations.



In September 1997, an aircraft engaged in pasture spraying
operations struck a tree and crashed, fatally injuring the pilot
(Occurrence 9703150). A factor in the accident was considered to be
pilot distraction caused by the lightbar associated with the GPS
navigation system. The lightbar was positioned about one metre in
front of the pilot's eyes.



In May 1998, an aircraft struck powerlines during spraying
operations (Occurrence 9801676). The pilot reported he had been
watching the GPS marker system lightbar, positioned on the nose of
the aircraft, and thus had not seen the power poles in his
peripheral vision.



In October 1997, an aircraft engaged in spraying operations
suffered total engine failure during a spray run (Occurrence
9703511). The aircraft's fuel had been exhausted. The pilot
reported the effectiveness of the red fuel warning light on the
upper instrument panel had been reduced by the proximity of the red
lights in the lightbar of the GPS marker system mounted on the
aircraft coaming, ahead of the windshield.





Lightbar usage



Information from the lightbar was used primarily to accurately
align each swathe run and avoid repeated or missed spraying of an
area. Reliance on lightbar information has increased where lack of
ground reference features, especially in broad acreages, required
additional guidance. The lightbar has been useful when spraying
cultivated land where the tillage was not parallel to the swathe
runs. In these circumstances, the ground patterns presented very
powerful visual cues, making straight flight difficult without
reference to some form of guidance system.



Most spraying operations are carried out across the wind to avoid
flying through the mist of the previous swathe run. Although only
small drift angles are encountered in these crosswind conditions, a
nose-mounted lightbar is sufficiently displaced from a pilot's
forward view as to require diverting visual attention away from the
flightpath.





Positioning of the lightbar



Discussions with representatives of the aerial agricultural
industry produced varied opinion regarding the best position for
mounting a lightbar. Although by far the most common position was
near the windshield, this was not universally the case.



Lightbar units were first installed at the front of the engine
cowling, as far forward as possible, to place them as near to a
pilot's distant vision as possible. Through experience, most
operators favour a much closer location of the lightbar. Some units
have even been mounted inside the cockpit on the instrument coaming
to avoid becoming obscured through dust and bug-smeared
windshields.





Lack of standard lightbar displays



Presentation of information and lightbar displays, varies
considerably between the GPS manufacturers and the different
models. Of particular concern is the use of certain colours to
light the displays. One of the accidents quoted above demonstrated
how easily the criteria for the design and certification of warning
systems can be negated by the installation of ancillary equipment,
such as a lightbar, which may use inappropriate colours in its
display.



Further investigation into the human performance aspects of the
use of such systems, especially the positioning of the lightbar,
was considered necessary. Expert opinion was sought and it was
considered most likely that both attentional and visual aspects
were influences in determining where best to locate a
lightbar.





ANALYSIS



Human factors aspects



Two main human factor issues arise in relation to aerial work
using GPS systems with a lightbar:



- Visual perception; and

- Attention and cognitive processing.



To some extent these two factors interact, as indicated
below.





Visual perception



The positioning of the lightbar will determine its focal distance
when viewed by the pilot and hence the degree of accommodative
shift required for the pilot to change focus between the display
and an outside object. The time taken for this readjustment of
focus may well be significant during low-level flight, particularly
in high performance (e.g. turbine) aircraft. Placing the lightbar
inside the cockpit, e.g. on the instrument coaming, will increase
the accommodative shift required. This effect will be more marked
with older pilots as speed of accommodation decreases with age.
Placing the lightbar further from the pilot's position, e.g. as far
forward as possible on the nose of the aircraft, will reduce but
not eliminate the need for accommodative shift.





Attention and cognitive processing



It is possible that because of the task importance and visual
salience of the lightbar, it will disproportionately gain the
pilot's attention, a condition called "cognitive capture". To the
extent that cognitive capture diverts the pilot from closely
attending to outside visual cues, situational awareness may be
significantly compromised. This effect will be accentuated during
demanding or stressful operations.



Positioning the lightbar as far as possible from the pilot will
approximate the use of a head-up display (HUD), and previous work
in this area may be relevant. There is evidence that pilots can
become fixated on the symbology of HUD displays. Under visual
fixation, pilots are less likely to process other instrument
information or outside visual cues. In effect, this will reduce the
pilot's field of regard (i.e. the total field of view available to
the pilot when looking at a visual scene over a period of
time).



Research has shown that limiting a pilot's field of regard is
particularly detrimental to judgement for curved flight paths. It
has been reported that 30% of pilots experience disorientation when
using HUD displays. Between 1980 and 1985, the US Air Force lost 73
HUD-equipped aircraft as a result of pilots becoming disorientated
or misorientated. There is evidence that HUDs may improve detection
of anticipated objects but hinder detection of the truly
unexpected. For example, when a large runway obstacle (i.e. an
aircraft) was introduced into the simulated landing scene, almost
all pilots ignored it. It is likely that cognitive capture was a
significant factor in this result.





Interaction of visual and cognitive factors



It is possible that the need for the pilot to change focus may
have positive as well as negative effects. When a pilot switches
gaze from a panel-mounted instrument display to the outside world,
there are a number of cues to indicate that a switch of attention
is taking place, e.g. looking up, changing accommodation, and
changing convergence. These can be strong cues to remind the pilot
to switch attention. There is evidence from work on HUD systems
that when the visual cues associated with cognitive switching from
an instrument panel to the outside world are missing, cognitive
capture is more likely to occur.



Placing the light bar in the pilot's peripheral vision may have
advantages. A number of displays have been designed for peripheral
vision. For example, the Malcolm horizon projects a narrow line of
laser light across the instrument panel, and hence does not require
eye fixation and attention during instrument scan. It has been
shown that information in the periphery tends to be neglected at
times of stress or high workload. This may act as a self-limiting
mechanism to reduce the salience of the lightbar at times of high
workload.





Use of other sensory perceptors for track guidance



An alternative means of presenting track-guidance information
could be explored. Using senses other than vision may be possible.
For example, providing track guidance by means of an auditory
signal to the pilot through a stereo headphone set would allow the
pilot's entire visual concentration to be centred on external cues.
External visual cues are still the most preferred means of
establishing and maintaining a flightpath.



Such a system could also be enhanced to generate aural warnings of
proximity to powerlines.





CONCLUSION



Practical experience and related research suggests that a pilot
may become preoccupied with monitoring information from a
nose-mounted lightbar unit and yet become relatively myopic to the
more distant outside information while focussed on a near lightbar
installation.



The optimum location of the lightbar would appear to be in the
pilot's peripheral vision. Additionally, more attention should be
paid to the proximity of lightbar units to other aircraft warning
lights and annunciator panel lights, especially with those units
that have similar colours to aircraft warning and system status
lights.



A curved approach path, such as the completion of the procedure
turn onto a swathe run, appears to be a critical point at which
distraction from other visual cues such as a lightbar, should be
avoided. The provision of this information by other than visual
cues should be researched.