Menneske-maskinproblemstillinger i forhold til flysikkerhet
Report presented to the Association of Aviation Psychologists,
5 October, 1995.
GPS is a relatively inexpensive, high technology navigation system based on a constellation of 24 US Department of Defense satellites orbiting the earth at a very high altitude. With the advent of small, portable, low priced receivers delivering easy and accurate solutions to navigation problems, I expect to see the rapid and widespread use of GPS throughout all levels of the aviation industry. GPS is now approved for VFR (Visual Flight Rules) use in New Zealand (CAA, 1995a), and as a supplemental navigation aid when flying under IFR (Instrument Flight Rules) (CAA, 1995b). It will soon be approved for arrivals and primary means navigation in Australia (Airservices Australia, 1995).
Pilots are required to train for and to hold an instrument rating
to fly by reference to navigation instruments. However, GPS is available
for use without formal training, and for use by non-instrument rated pilots
in visual flight. Traditionally aviation technology has been developed
from advanced military and civil research, has then been incorporated into
airliners, and finally into light aircraft. This has generally followed
comprehensive testing, the development of regulations, and training procedures.
GPS is an example of flight instrument implementation that reverses this
trend. GPS has been rapidly adopted for GA use ahead of regulatory provisions
and training requirements due to the pace of commercial technological advancement,
availability, and affordability.
Human factor overveininger ga støtet
til denne undersøkelsen
The human factors considerations that prompted this study included
the ergonomics of the receiver controls and display design, and the effects
that GPS use might have on pilots' behaviour. Anecdotal evidence suggested
that GPS had been used quite extensively in an "un-approved" manner and
that its use may have altered some pilots' flying behaviour. Changes may
have occurred to the navigation strategies used, and to the decision making
and judgement processes that follow from these. Possible negative consequences
associated with using GPS inappropriately have been contemplated, and the
need to identify potential hazards has been discussed by O'Hare and St.
George (1994).
At present GPS is essentially a non-standardised "add-on" navigation system in GA aircraft. Introducing GPS into aviation in a piecemeal way could mean that human factors considerations are overlooked, or relegated until events occur that force a re-think for design, approved procedures, or training. A survey of pilots using GPS in New Zealand was carried out as an exploratory study to identify human factor problems with GPS design and operation that may have an impact upon flight safety.
Undersøkelsen tok opp seks spørsmål angående bruk av GPS
1: Er designet til GPS i samsvar med etablerte human factor prinsipper?
2: Hva er GA piloters holdninger til GPS?
3: Har pilotens holdninger forandret seg ved bruk av GPS?
4: Oppmuntrer GPS til atferd som kan virke negativ inn på flysikkerheten?
5: Antyder de problemene som GPS brukere rapporterer hvordan design kan forbedres for å øke flysikkerheten?
6:Antyder de rapporterte problemer
ved GPS bruk krav til trening for å bedre flysikkerheten?
METODE
Svar
De som svarte: 172 New Zealand piloter som
var GPS brukere. Fra PPL til ATPL. Bare tre kvinner. Alder fra 21
to 71 years (Mean = 42 , SD = 12). Brukerne hadde gjennomsnittelig
265 timers erfaring med GPS (3 til 4,000 timer).
Mean betyr gjennomsnitt og SD
= Standard Deviation som er et mål på spredning. Innen pluss-minus
et SD har en 68% av verdiene i en normalfordeling, innen +/- 2 SD har en
96%. Et lavt standardavvik betyr stor enighet mellom de som svarer.
Framgangsmåte:
Item Overall rating(1 to 7) (Extremely poor to extremely good) Mean SD N Unit size and shape 5.9 1.0 172 Power sources 5.6 1.4 170 Antenna connection 5.3 1.6 166 security Flying in turbulence 5.2 1.4 170 Entering information 5.0 1.4 172
Strømkilder:
Problemer som ble nevnt inkluderte at batteriet
ble uttømt eller at forbindelsen til flyets strømforsyning
var dårlig. Advarsel om dårlig
batteri kom like før det sluknet.
Signal reception
Mottagerproblemer fra satellitt hadde oftest
sammenheng med antennemontering.
Best var det med eksterne antenner.
GPS kontroll design
All vurderingene var enten "god" eller
"meget god".
Table 2: GPS controls ratings
Item Overall rating (1 to 7) (Extremely poor to extremely good) Mean SD n Key size 5.5 1.1 165 Key spacing 5.5 1.2 165 Key coding/symbols 5.4 1.2 165
De fleste vurderingene var i kategorien "god"
eller" meget god". At displayene noen ganger, under
visse lysforhold, var vanskelig å lese
skyldes at de var små. Displayene burde forbedres.
Item Overall rating (1 to 7) (Extremely poor to extremely good) Mean SD n Display character 5.4 1.1 172 legibility Reading the display in 5.2 1.3 172 flight Function/mode symbols 5.0 1.2 170 Warning indications 4.8 1.5 164
Item Overall rating (1 to 7) Hvor nyttig er GPS mht følgende: (Extremely poor to extremely good) Mean SD n Groundspeed 6.7 0.6 170 Present position 6.7 0.7 170 Distance to go 6.7 0.7 170 Go-to direct 6.6 0.8 163 CDI-bar 6.2 1.2 153 Simulator 5.2 1.6 46 Alarms 4.5 1.6 96
Tabell 5 viser at brukerne hadde meget stor
tillit til GPS informasjonen, og at de følte seg mer usikre uten
GPS
Brukerne var mindre sikre på mer komplekse
funksjoner som å lage ruter, og at disse funksjoner ikke er så
intuitive.
Item Overall rating (1 to 7)
How confident are you: - ("Not at all" to "extremely")
Mean SD n
With GPS information 6.2 0.8 172
Navigating with GPS 6.3 0.8 172
Navigating without GPS 5.6 1.2 172
Using basic GPS functions 6.3 0.9 170
Using all the GPS 4.6 1.8 170
functions
55% rapporterte en mengde problemer med å entre data, mens 45% rapporterte aldri. De fleste oppdaget sine feil ved kryssjekking. Feilene var av mange typer, men kan deles i de som hadde sammenheng med å finne de rette verdier fra kart etc. og de som hadde med fysisk betjening å gjøre.
Noen incidents har forekommet pga input feil.
Feillesning:
20% nevnte dette, som for eksempel "10" i stedet for "18" og vice
versa;
Monitoring behaviour
Trening i bruk av GPS
81% av brukereneer selvlærte. However, half the
users (50%) checked that other sources were needed for GPS training, while
49% of users checked that training from the user manual was sufficient,
and 1% checked that no training was required.
Training - Received Needed (n =172) (n =172) User manual 81% 49% Other 12% 50% None 7% 1%
25 brukere kommenterte at manualene kunne forbedres betraktelig.
Vanskeligheter med å lære GPS
Vurderingen av manualene var M = 4.7, SD = 1.6,
Tabell 10: Et utvalg av brukerkommentarer
om vanskeligheter i å lære hvordan GPS skal brukes.
Initially learning how to input data and utilise the many available functions. Lack of knowledge about the GPS thus limiting more effective use. I found that learning from the manual was very difficult. Once mastered operation is very imple. Took time to learn how to use it. Not used to computer technology There is nothing difficult about the GPS. The initial learning curve is the hardest. The manuals are written in such a way that it makes it more difficult to learn. If you are not young have not used computers they are very hard to learn how to use, it can take weeks. Understanding the instruction manual. I have not been used to computer language and also find instructions refer back to previous pages for continuation of operation and I tend to lose the trend. Working out how to use it by reading the user manual. Poor user manual
Table 12 shows some representative comments from users of perceived hazards when operating GPS in an inappropriate manner. These dangers could be minimised with human factors training to raise each individual pilot's awareness of these and similar issues. Complacency, failure to cross-check, over-reliance, risk taking in poor weather, and poor chart use were some of the issues raised as having happened to respondents. Many users believed these types of events were a very likely to occur without an understanding of the dangers involved when using GPS beyond its limitations through ignorance and a lack of training.
Tabell 12: Et utvalg av brukerkommentarer
om farer ved GPS bruk som kan unngåes ved oppmerksomhetstrening.
Antenne og strømforsyning kan være et problem
Portable GPS receivers are generally powered by an AA battery pack,
or by rechargeable NiCad packs. Connection to the aircraft powered cigarette
lighter is an optional accessory. Battery usage is high and flat batteries
can come at inopportune moments as warnings for this are poor. Panel mounted
receivers are permanently mounted in the instrument panel and are connected
to aircraft power and an external antenna. There are no trailing wires
or fiddly mounts to deal with. Power and aerial problems caused the greatest
GPS unreliability especially for handheld models. Some solutions offered
for power supply integrity were to have a battery life remaining indicator
such as found on mobile telephones; to indicate whether the unit is running
on internal or external power; to carry plenty of spare batteries; and
to have automatic switching to batteries if aircraft power is lost (which
some models did offer).
Lesbarhet dårlig under visse lysforhold
GPS controls were rated as adequate. Key size and spacing is constrained
by the small size of the unit necessary to maintain portability. Display
readability also suffered from the small unit size making the alphanumerics
difficult to read under certain ambient lighting conditions. The warning
features of the displays were not highly rated and it was mentioned that
they often interfered with other tasks as extra key pushing is required
to access the warning message and cancel it.
Navigasjonsinformasjon meget
god, men komplekse funksjoner vanskelige
The navigation functions offered to pilots by GPS received extremely
high ratings. Features such as accurate groundspeed, distance to go, and
course deviation indications make their task much easier than before. The
precision and reliability of the information gave the users great confidence
in the machine. This was reflected in the results that respondents use
GPS most of the time when flying, and that they are not as confident without
it. They were also less confident using the more complex GPS functions,
possibly because these require a greater depth of knowledge, and practice
to remember. Most keys control several functions because of space constraints
and this makes GPS use complicated for more than the basic functions. Thought
needs to be made in the design process for more intuitive steps to easily
access the required information.
Små taster og feil avlesninger et problem for sikkerheten
Input errors and misreading errors occur when using GPS. This is
mainly because of the small size of the control keys and the small display.
Incidents because of this have been reported and these have flight safety
implications. Design improvements and specific training may alleviate some
of these problems. The number of users reporting difficulties here were
small but the safety implications are considerable.
Automatisering og arbeidsbelastning.
Krever mer kunnskap og planlegging.
GPS is an example of the increasing cockpit automation occurring
in aviation. Wiener (1988) notes that two principle rationales for automation
may be questionable. Firstly, that automation requires less manual handling
and mental computation, leaving the pilot free for more effective supervision;
and secondly there is more time available for scanning for other aircraft.
Wiener (1988) suggests that while manual workloads have reduced, mental
workloads have increased, as have the range of skills required of pilots
to operate these sophisticated machines. He states that new aircraft technology
requires more programming, planning, sequencing, alternative selection,
and cognitive processing. Wiener says pilots perceive that automatic devices
require constant attention, therefore time needed to scan the displays
reduces time available to scan outside the cockpit.
Undersøkelsen viste at
GPS lettet pilotens opplevde arbeidsbelastning
The results of this survey suggest that GPS has significantly reduced
pilots perceived workload, possibly due to less requirement to read maps
and make mental navigation computations. Lookout has not increased as much
as might have been expected however and this may be because of extra time
scanning the unit. At least GPS does not appear to be encouraging more
time scanning the instruments at the expense of looking outside.
Få fløy i dårligere
vær med GPS
Users reported being tempted to fly in conditions they would not
consider without GPS and this was reflected in the rating of slightly more
bad weather flying. Other similar behaviours reported by a small proportion
of users included flying above cloud without an instrument rating which
has grave safety implications, especially if the pilot loses GPS information
as does occur for various reasons such as a power or reception interruption.
Navigasjon mer
nøyaktig med GPS, men gir ikke terrenginformasjon
As might have been expected users reported to be much more accurate
with their navigation, tracking, and awareness of their position relative
to airspace and terrain. This is a very positive feature for using GPS
and it should reduce the number of pilots becoming lost or running out
of fuel through poor route following. However, pilots may require specific
training to emphasise that GPS does not provide terrain information and
that it must be used in conjunction with maps to cross check the track
made good by the aircraft. The increased accuracy of flying direct tracks
VFR between popular waypoints raises the issue of "unofficial VFR routes"
being created. This may have the effect of "funneling" VFR traffic closer
together putting pressure on the concept of "see and be seen" for VFR separation.
Bekymringsfylt
at 6% bare bruker GPS. Viktig med kryss-sjekking
A concern from the results was that nearly 6% of users do not cross
check the GPS information, and that a further quarter of users at times
fly using GPS instead of a map. This over reliance on the machine could
have fatal consequences if the route data has been entered incorrectly.
This type of event was reported to occur, although most times it was corrected
through operators cross-checking their actions. Another concern raised
by the study was that the design of most GPS units makes it difficult to
easily check the entered track and distances of routes with the flight
plan. This is a basic requirement for safe navigation to prevent the machine
"dumbly and dutifully" taking the operator to where it has been programmed
but not intended to go. This is a design feature encouraging behaviours
with negative flight safety implications.
Trening i bruk av GPS er ønsket
The results show that little formal training is available for GPS
operational theory and practice, but that there is a desire by users for
such training courses to meet their perceived need. A significant proportion
of respondents indicated that the training they received on the use of
GPS was less than they desired. Typically they were left to teach themselves
using the manual, or had to obtain informal demonstrations from other experienced
users. Over a third of respondents considered that a formal training course
was required to enable proficient use of GPS, and over half the respondents
indicated that they would attend such a training course if it was available.
A significant proportion (10%) felt that a formal training syllabus and
examination was needed. Only a minority of respondents indicated that they
felt little or no formal training was required to use GPS competently.
As the majority of respondents in this survey were VFR pilots, suitable
training is likely to be perceived as even more important for IFR operations
with GPS.
The knowledge base of users for GPS has the potential for considerable improvement as users generally rated their knowledge of GPS theory as "good", as distinct from "very good" or "excellent". It is possible that formal training courses would increase the level of knowledge rated as necessary to operate GPS proficiently. It may be that while a "good" knowledge is rated as needed to operate GPS in a basic manner, a greater level of knowledge of both theoretical and practical aspects of operation will be perceived to be required to use GPS to its maximum capabilities, and to avoid falling into the human factors traps identified in this survey.
Learning difficulties identified by this survey such as coping with the computer jargon and concepts used by GPS, poorly written user manuals, and a steep learning curve can be overcome within an appropriate training course. Many users' expectations of GPS and its specific applications were inaccurate, and these misperceptions can be easily cleared up at the beginning of formalised teaching of GPS theory and operational practice.
Farer med GPS: complacency,
over-reliance, risk taking, "not in the loop"
Viktig med: situational awareness
og å ha en mental modell
The survey identified many human factors and flight safety implications
for inadequate training. These issues are typically not addressed by the
current learning methods, and they are less likely to be corrected through
individual experience as the operator becomes more proficient with the
GPS unit. Formalised human factors training to raise operator awareness
of possible hazards when operating GPS in an inappropriate manner is needed
to standardise operational practices and to minimise the dangers associated
with a lack of knowledge in these areas. Some of the issues identified
here include operator complacency, over-reliance, risk taking, and a lack
of basic navigation monitoring techniques. While respondents showed they
were extremely confident with the information they receive and the reliability
of GPS, there was a worrying indication that this was overconfidence, and
that precautions were not being taken in the event that GPS information
became unavailable. The reports that respondents were significantly less
confident navigating without GPS, and in using the more complex functions
available, could suggest that GPS is encouraging basic "pre-automation"
skill levels to decline while full proficiency with the equipment is not
being attained.
Viktige momenter som bør
være med ved treningkurs:
Kryss-sjekking, overvåking,
ha alternative planer hvis strømmen forsvinner,
ikke
bruke GPS som eneste navigasjonhjelpemiddel.
A training course should be an effective means for alerting pilots to the physical difficulties that may be encountered when using a typical GPS receiver and in presenting strategies to deal with these. Cross checking procedures when entering data and other common input errors can be taught at an early stage rather than users finding out the hard way over time that their checking procedures were inadequate and that they have ended up going where they did not intend.
Training must emphasize the need for close monitoring of GPS like any other automated equipment and for the comparison of GPS data with other navigational sources. It is possible that the average VFR pilot will not have the instrument monitoring experience required to realise the importance and difficulty involved in this task. Users need to be made aware of the various reasons that GPS can fail and not to believe that the information will always be correct as appears to sometimes be the case. A naive common perception is that the major danger to trusting GPS is that it may be turned off without notice, however it is far more likely that the batteries will go flat or that the operator will program the GPS incorrectly.
Training should indicate the appropriate use for maps and charts in conjunction with GPS. Correct methods of flight planning and checking, monitoring progress with regard to airspace and terrain, and reversionary procedures should be considered.
While GPS equipment is generally well designed and easy to operate,
there are complex features that require practice to master, and there are
ways to use it that invite disaster should things go wrong. By instilling
over-confidence in the ability to find ones' way under any situation, inappropriate
decisions may be made to fly in conditions unsuitable should GPS information
fail for any reason. Pilots need to be aware of such situations, of the
likelihood and causes for losing GPS information, and of the means to ensure
they get the maximum benefit from their high-technology machine. Training
that promotes standardised operating procedures for VFR and IFR flight
with GPS, and that alerts users to possible problems and how to avoid them,
is a necessary step to ensuring flight safety is only improved with the
further implementation of GPS.
Behov for GPS trening av GA
piloter.
The study suggests that there is a need for a more formal approach
to GPS training at the GA level. Respondents expressed a strong desire
for structured training in the appropriate use of GPS. It was deemed to
be important that pilots fully understand the theory of how GPS functions
and how to use it effectively in conjunction with other navigation techniques.
Respondents found GPS manuals difficult to learn from and a good many reported
"learning" by trial and error. It is apparent that the power of GPS needs
to be balanced against safe learning opportunities and operational practice.
Human factors considerations are also required to standardise and improve
certain aspects of GPS equipment design.
Hva vil fremtiden bringe?
1: Elektroniske kart sammen
med GPS.
Dette er teknisk mulig. I USA har det blitt
laget en "talking map" for blinde bygd på sammenkopling av
GPS og elektroniske kart. Sammenkoblingen
av de to delene er relativ enkel for den som kan det.
Kunnskap av denne art finnes i Norge. Det
er et spørsmål om lønnssomhet.
2: Autopilot/wingleveler sammen
med GPS kunne gjøre IMC sertifikat aktuelt
Denne kombinasjonen kunne gjøre det
mer aktuelt å fly under IMC forhold, i det minste tillate
skygjennomgang etter å ha fløyet
på topp. England er det eneste
landet som har dette sertifikatet til nå iflg. mine opplysninger.
Verdt å diskutere?
REFERENCES
Airservices Australia, (1995). GPS set for introduction as Primary Means Navigation. Aviation Bulletin, 4, (6), 1, 4-5.
CAA, (1995a). Conditions of installation of global positioning system (GPS) equipment in New Zealand registered aircraft for use in VFR operations. AIC-GEN A28/95. Wellington: Aeronautical Information Service.
CAA, (1995b). Requirements for the use of GPS as an approved supplemental navigation aid when flying under IFR. AIC-GEN A30/95. Wellington: Aeronautical Information Service.
Nendick, M. (1994). GPS User Survey. Massey: University Print.
O'Hare, D., & St. George, R. (1994). GPS - (Pre)Cautionary Tales. Airways, 6 (2), 12-15.
Wiener, E. L. (1988). Cockpit automation. In E. L. Wiener & D.
L. Nagel (Eds.), Human factors in aviation. (pp. 433-461). San Diego:
Academic.
Control modes:
Kontrollnivåer:
3: Regelbasert normal atferd. En følger prosedyrer.
3: Tactical control mode: Rulebased normal behavior.
Foreløpige kommentar.
En kan se på en flytur som et sjakkspill. Jo mer erfaring en har og til bedre planlegging en har gjort, jo flere mentale modeller har en som gjør at en bedre kan hanskes med uventede situasjoner og overvinne problemer som måtte oppstå underveis. Hjelpemidler som GPS kan hjelpe oss til å ha mer nøyaktige modeller som igjen kan føre til tryggere flygning. Flygeren selv må innpasse all den informasjonen som mottaes i sin mentale modell som bør gi det beste grunnlaget for å gjøre de riktige handlinger.
En retningslinje ved utforming av hjelpemidler
som GPS, instrumentpaneller ol er at de må være tilpasset flygerens
måte å tenke på. I dag er det ofte motsatt. Kontrollorganer
bør også være utformet slik at en instinktiv reaksjon
er en riktig reaksjon, at en intuitivt ser hva som er den riktige handling.
Se Gibson,
Affordances and Aviationen mer teoretisk
diskusjon av dette siste punktet.
Bjarne Fjeldsenden 09.11.1998