Tropper, K. & Kallus, K. W. (2005).

Disorientation in VFR Pilots: Flight Performance and Psychophysiological Changes during

a Flight Simulator Training. Proceedings of the 13th International Symposium on Aviation Psychology (pp.573-578).
Oklahoma City.

DISORIENTATION IN VFR PILOTS: FLIGHT PERFORMANCE AND PSYCHOPHYSIOLOGICAL

CHANGES DURING A FLIGHT SIMULATOR TRAINING
Karin Tropper & K. Wolfgang Kallus
Department of Psychology, Karl-Franzens-University of Graz
Graz, Austria

Disorientation due to flying into instrument meteorological conditions (IMC) is a major safety hazard for VFR pilots
(VFR: visual flight rules) as confirmed by aviation accident databases. The objectives of our research are the
development and evaluation of systematic training programs to cope with different kinds of disorientation
phenomena and the analysis of the psychophysiological processes during dis- and reorientation. A study was
conducted using the multi-axial moveable flight simulator DISO (AMST Systemtechnik GmbH, Austria). 25 pilots
were randomly allocated to one of three testing groups (one control- and two experimental training groups). The
flight performance data confirm that participants with a training show better performance data in a test phase than
pilots without training. The simulation scenarios are of high impact: Heart rates are clearly increased in response to
more demanding segments of flight as e.g. during takeoff and landing. Analyses within the test profile “unusual-
attitude recovery” demonstrate – in addition to the expected increase of heart rate due to higher mental workload –
an important interaction: The increase is lower for pilots having received an unusual-attitude recovery training. First
EEG results illustrate changes in the alpha- and beta band due to changing strain. To sum up, this study tries to make
a contribution to basic research by analyzing psychophysiological processes as well as to applied science by
emphasizing the importance and effectiveness of orientation training programs for VFR pilots.

Introduction results are that the simulator illustrates disorientation

phenomena very realistically, that flight performance
Disorientation due to flying into instrument meteoro- increases after a disorientation recovery training, and
logical conditions (IMC) is a major safety hazard for that worse performance in simulator exercises – e.g.
VFR pilots. Analyses of aviation accident databases crash during the profile “Black hole approach” – is
confirm that in general aviation fatal aviation acci- accompanied by high physiological stress as indica-
dents are often classified as involving visual flight ted by increases in heart rate (Kallus & Tropper,
rules (VFR) into instrument meteorological condi- 2004). Based on these results, a study was designed
tions (e.g. Goh & Wiegmann, 2001; Véronneau & with adopted profiles for VFR pilots (Haug, 2003)
Evans, 2004). using again the multilevel multi-method approach for
Our concept to explain spatial and geographic orien- the evaluation of the training effects and the analysis
tation and disorientation bases on the model of antici- of cognitive, psychological and psychophysiological
patory action regulation from Hoffmann (1993) and processes.
the model of situation awareness (SA) from Endsley
(2000). “Situation Awareness is the perception of the Method
elements in the environment within a volume of time
and space, the comprehension of their meaning, and Design and subjects
the projection of their status in the near future” 25 VFR pilots (average age of 43 years, SD = 10.5,
(Endsley, 1995, p. 65). Situation awareness involves 23 men, 2 women, all owning a private flight license)
a correct appreciation of many conditions. The most were randomly allocated to one out of three testing
relevant aspects in aviation are three-dimensional groups. The experimental design is given in Table 1.
spatial awareness, system (mode) awareness, and task Table 2 shows an overview of the flight profiles.
awareness (Wickens, 2002). As correct orientation is Every participant completed three phases in the flight
a central factor of situation awareness, loss of orien- simulator. The nine pilots of the training group
tation leads to loss of situation awareness (LSA). attended the awareness training (“awareness”) during
The objectives of our research are the development phase I, followed by the training with orientation-
and evaluation of systematic training programs and unusual-attitude recovery exercises (phase II,
helping to cope with different kinds of disorientation “training”). The eight pilots of the awareness group
phenomena, using the multi-axial moveable also went through the awareness phase, but instead of
(continuous yaw, limited pitch and roll) flight simu- the training phase they completed the control condi-
lator DISO (Disorientation Trainer, AMST System- tion “free flight”. The control group (n = 8) went
technik GmbH, Austria). through two free flight phases instead of the training.
In a first study, 26 jet pilots participated. The main All 25 pilots passed the test (phase III) at the end of

573
the testing day. The simulator exercises were based flight path is approximately 33 nautical miles long
on a PC7 simulation. and it takes about 18 minutes to fly the whole circuit
(including takeoff, the flight maneuvers and landing).
As aid, the pilot gets a colored map of Peloponnes
Table 1: Experimental design into which the flight path is drawn. Additionally,
standardized headings are used by the instructor pilot
PHASE I PHASE II PHASE III to lead and help the pilot via radio connection. The
TRAINING instructor pilot also took on the tasks of an air traffic
GROUP Awareness Training TEST controller.
(n = 9)
AWARENESS Test profile VFR flight at minimal weather condi-
Control
GROUP Awareness TEST
(n = 8)
condition II tions, mountains. This profile begins with conditions
of bad visibility (10 km). It is planned to fly the same
CONTROL
Control Control route as during the instruction flight and the pilot is
GROUP TEST
condition I condition II explicitly instructed “to behave as in a real flight
(n = 8)
situation”. The visibility deteriorates further with
time (5 km). It is overcast and the mountains are in
Table 2: Overview of the simulator profiles clouds. It is not possible to fly the whole planned
circuit under VFR condition. Because visibility
PHASE I deteriorates gradually, it is expected that not all pilots
CONTROL become aware of the hazard and use visual flight
AWARENESS
CONDITION I rules into instrument meteorological conditions.
Cockpit Instruction Cockpit Instruction
Instruction flight at Instruction flight at Test profile Unusual-attitude recovery. Unusual-atti-
excellent weather excellent weather tude recovery means the process of returning the air-
conditions (WX) conditions (WX) craft to near straight and level from an unexpected
VFR flight at min. bank and / or pitch angle. The exercise is drawn from
WX, mountains jet pilots´ training courses. At the beginning of this
Free Flight I
VFR flight, moun- profile, the PC7 is already airborne. After about two
tains, clouds tilt
minutes, the instructor pilot takes over the control of
Passive spin Passive spin profiles: the PC7 and sets certain – standardized – flight para-
profiles: Gyrospin I Gyrospin I and
and Gyrospin II Gyrospin II meters via the external workstation. During the set-up
time the participant inside the flight simulator keeps
his eyes closed. After taking over the control from
the instructor pilot, the pilot in the simulator is re-
PHASE II PHASE III
quired to reach safe flight parameters (to recover) as
CONTROL
TRAINING CONDITION II TEST fast as possible. This exercise is conducted ten times.
VFR flight at min. VFR flight at
WX, mountains, min. WX, Procedure
visual and VOR Free Flight II mountains The examinations lasted five to eight hours per pilot.
VFR flight at min. Unusual-attitude Before and after each flight simulator phase, a two
WX, mountains, recovery minute resting measurement (baseline, eyes closed)
Radar Vectors was conducted. After each phase (outside the
Unusual-attitude simulator), the pilot took part in an extensive recon-
recovery training
struction interview concerning the flight profiles.

Dependent variables
Instruction Flight. The instruction flight takes place Aviation performance (observation data, instructor
under conditions of good visibility (about 80 km). It pilot ratings, time-measurements), psychological data
leads the pilot along a standardized flight path with (questionnaires for analyzing changes in subjective
the takeoff at Kalamata (Greece), leading to the physical and psychical state, reconstruction inter-
coast, along the coast, briefly across the sea, into views), and physiological variables (ECG, EEG,
terrain with mountains and finally back to the airport EOG, EDA) were measured.
of Kalamata. After passing the last of five turning
points and before landing in Kalamata, the pilot flies Some results concerning the following dependent
two maneuvers: an aileron roll and a looping. The variables are reported here:

574
- Flight performance: observation data Crashes (VFR flight, test)
- ECG: heart rate – deviation from baseline: Positive no crash
8
differences signify an increase in heart rate in 7 crash
comparison to the resting measurements. 6

n of pilots
5
- EEG: spontaneous activity 4
3
2
EEG was recorded by eight bipolar channels (posi- 1
tions of electrodes cf. Table 3; the ground electrode 0
TG (9) AG (8) CG (8)
was fastened to the forehead). The electrode impe-
dances were below 5 k ohms and the sample rate was
Figure 1: Crashes during the VFR flight of the test phase
128 Hz. Recorded data were subject to visual inspec- separate for pilots of the training group (TG), awareness
tion using the BrainVision software package of the group (AG), and control group (CG).
Company Brain Products GmbH (Munich). Seconds
with artefacts were excluded from further analyses.
The EEG from 1 second periods were submitted to Cloud layer entering (VFR flight, test)
spectral analysis using the Fast Fourier Transforma-
tion (full power spectra, Hanning window). After no cloud layer entering
8 cloud layer entering
averaging the absolute power values of the 1 seconds 7
6

n of pilots
periods of certain sections of measurement, the data 5
were combined to the standard bands of alpha (8-13 4
3
Hz) and beta (14-30 Hz). 2
1
0
Table 3: Positions of the 16 EEG electrodes (eight bipolar
TG (9) AG (8) CG (8)
channels, frontal to occipital regions)
Figure 2: Cloud layer entering during the VFR flight of the
Channel 1 F3 - FC´3
Channel 2 F4 - FC´4
test phase for pilots of the training- (TG), awareness- (AG)
and control group (CG)
Channel 3 FC3 - PC3
Channel 4 FC4 - PC4
Channel 5 C3 - P3
Channel 6 C4 - P4
Channel 7 P´3 - O1 Heart rate
Channel 8 P´4 - O2 Concerning the instruction flight at the beginning of
the testing day, the results of the two-factorial
ANOVA for repeated measures with the between
The rate of missing EEG-data is beyond five percent factor testing group exhibit a strong main effect of
for each channel (due to continuously artifacts – e.g. the section of measurement [F(15.2,319.4) = 21.4,
muscle activity – or technical problems). No missing p = .000]. (There are neither differences between the
EEG data have been replaced and to lose no testing groups nor is there an interaction.). As
additional data, only univariate analyses (power of illustrated in Figure 3, the different tasks within the
only one channel) have been calculated. flight are clearly reflected in the heart rate (beats per
minute, deviation from baseline). In average, the
heart rate is always above the baseline. The least
Results stressful sections are about between 90 seconds after
the takeoff and 30 seconds before the first flight
Flight performance maneuver (role). The first strong increase of the heart
During the test phase (VFR flight at minimal weather rate occurs before the takeoff; descriptively the be-
conditions, mountains), the pilots of the control group ginning of the ascent can be observed 30 sec. before
caused the highest number of crashes [Pearson-Chi2 the takeoff (TO), statistically (Tukey HSD post hoc
(df=2, n=25) = 10.96, p = .004, Figure 1]. Figure 2 tests, p < .05) it becomes significant 10 sec. before
illustrates that pilots of the training group show the TO. When the aircraft is safely airborne, the heart
tendency to enter the cloud layer less often than parti- rate decreases quickly within 30 seconds; the whole
cipants of the other two groups [Chi2 (df=2,n=25) = decrease takes about 90 sec. The flight maneuvers
4.99, p = .102]. aileron role and looping are also reflected in the heart
rate. Already 80 sec. before the landing (touchdown),
there is a strong increase in the heart rate which
reaches a maximum between the range of 10 sec.

575
before and 10 sec. after the touchdown, followed by a Within the test profile unusual-attitude recoveries,
rapid decrease within 20 seconds. there is a clear interaction between the section of
measurement and the testing group [F(17.7, 176.7) =
2.4, p = .002, Tukey HSD post hoc tests] in addition
Heart rate changes within the instruction flight to the main effect section of measurement [F(8.8,
32
176.7) = 12.2, p = .000, Figure 4]. While there are no
Training group group differences at the beginning of the profile (be-
Heart rate (dev. from baseline)

28
Awareness group fore flying the ten recoveries), the increase of the
24 role,
Control group heart rate is much higher in the control group than in
loop.
20 the training group.
16
12 EEG – Unusual-attitude recoveries (test phase)
8 As analyses illustrated no differences between the
4
three testing groups and for some calculations group
sizes were too low, the factor testing group has not
0
v Take- v vvv v Touch- been involved in the following calculations. In a first
off down step the absolute power of the EEG during waiting
82 sections of measurements with closed eyes for the command to recover from an
(76 x 10 sec. sections, 6 variable - indicated with v)
unexpected attitude (10 x 8 seconds, closed eyes),
was compared with the EEG during the resting
Figure 3: Heart rate changes (beats per minute – deviation measurements before and after the test phase (each
from baseline, means) separate for the three testing groups two minutes, eyes closed).
(TG: n = 8, AG: n = 8, CG: n = 8)
Table 4: Average power (µV-Square) in the alpha- and beta
band during the resting measurement before the test phase
Concerning the heart rate, no differences occur be- (RM5, 2 min., closed eyes), the time while waiting with
tween the three testing groups during the flight closed eyes for the command to recover within the test
profiles of the first two phases in the simulator. profile unusual-attitude recovery (Bef. Rec., 10 x 8 sec.)
and the resting measurement after the test phase (RM6),
and the results of the ANOVAs
Heart rate changes within the profile unusual-attitude
recovery (test) Bef
RM5 Rec. RM6 p-
Training group ALPHA (M) (M) (M) n ANOVA value
32 Awareness group
Control group
F3 - FC´3 2.6 2.1 3.1 17 F(2.0,32.0)=2.9 .070
28 F4 - FC´4 3.0 2.4 3.6 18 F(1.3,21.6)=5.7 .019
Heart rate (dev. f. baseline)

24 FC3 - PC3 16.5 7.9 16.5 18 F(1.4,23.1)=10.1 .002

20 FC4 - PC4 17.4 8.0 19.5 19 F(1.1,20.6)=9.0 .005
16 C3 - P3 23.7 14.5 23.0 19 F(1.7,30.2)=9.2 .001
12 C4 - P4 20.6 12.4 20.5 19 F(1.6,29.5)=9.1 .002
8 P´3 - O1 38.4 40.9 44.1 20 F(1.3,23.8)=0.5 .517
4 P´4 - O2 40.9 39.1 42.8 19 F(1.3,23.4)=0.7 .451
0
1. Rec
2. Rec
3. Rec
4. Rec
5. Rec
6. Rec
7. Rec
8. Rec
9. Rec
10. Rec

Bef
RM5 Rec. RM6 p-
BETA (M) (M) (M) n ANOVA value
36 sections of measurement F3 - FC´3 1.3 1.8 1.6 14 F(2.0,26.0)=1.2 .326
F4 - FC´4 1.2 1.5 1.3 16 F(1.3,19,.6)=1.0 .342
Figure 4: Changes in the heart rate (beats per minute – FC3 - PC3 4.3 3.6 4.5 15 F(2.0,28.0)=3.4 .049
deviation from baseline, means) during the test profile FC4 - PC4 4.1 3.4 4.2 17 F(1.4,23.1)=2.7 .103
unusual attitude recovery (ten recoveries) for the three C3 - P3 4.1 3.6 4.2 16 F(2.0,30.0)=1.3 .291
testing groups (training group: n = 9, awareness group: n =
7, and control group: n = 7); each recovery exercise takes C4 - P4 3.8 3.5 3.6 17 F(1.9,30.6)=0.5 .584
about 13 seconds, the whole profile about 12 minutes. P´3 - O1 5.2 5.3 5.3 19 F(1.7,30.6)=0.1 .938
P´4 - O2 5.1 5.4 5.1 18 F(1.2,21.1)=0.3 .615

576
The results demonstrate no changes in the absolute Discussion
power of the alpha band at the parieto-occipital
positions P´3-O1 and P´4-O2. But concerning all The results of the flight performance data confirm
other measurement positions (frontal to parietal), the positive training effects, especially for the test profile
alpha occurring during anticipating the recovery “VFR flight at minimal weather conditions, moun-
exercises is clearly decreased compared to a resting tains”. Pilots with a training behave less risk prone,
measurement. For the beta band, a low decrease at whereas pilots without any kind of orientation
FC3-PC3 could be detected (Table 4). training do often not turn back at an appropriate
moment. They enter the cloud layer more frequently
In a second step, the periods before recovering (10 x and lose orientation, which finally can lead to a crash
8 seconds, closed eyes) and during recovering (10 x 5 into the mountain or into the ground by trying to stay
seconds after controls have been handed over from under the cloud layer without realizing that the
the instructor pilot to the participant in the simulator, mountains are in clouds. This happened despite the
eyes opened) were compared. As expected, there are fact that the pilots had a map (including the geogra-
of course very big decreases in the alpha band – phical data of the region etc.), that they had flown the
especially over posterior regions, but at the two route already under conditions of good visibility (in-
anterior channels, there are no changes in the alpha struction flight), and that they always had the possi-
band. Concerning the beta band, there is a significant bility to get weather information from the “air traffic
increase of power at F3-FC´3 and decreases at controller” (i.e. form the instructor pilot at the exter-
posterior regions. nal work station of the simulator). As many accident
reports, this fact highlights the problem of deteriora-
Table 5: Average power values (µV-Square) in the alpha- ting visibility conditions: Some VFR pilots do not
and beta band while waiting with closed eyes for the recognize the ensuing danger which can lead to fatal
command to recover (Bef. Rec., 10 x 8 sec.) and while crashes, even in regions well known to the pilots.
recovering (Rec. 10 x 5 sec. after controls have been taken The simulation scenarios are of high impact for the
over), and the results of the T tests
pilots, as could be demonstrated by the changes in the
heart rate. As example the data of the instruction
Bef. Rec p-
flight have been presented. The clear increases
ALPHA Rec. (M) (M) Diff df t value
caused by the takeoff and the landing procedure are
F3 - FC´3 1.8 1.9 .2 13 .6 .548
similar to the published results concerning changes
F4 - FC´4 2.0 1.7 -.3 13 -1.2 .259 during flight (e.g. Hankins & Wilson, 1998; Wilson,
FC3 - PC3 8.0 2.6 -5.5 14 -2.5 .026 2002). Veltman (2002) compared psychophysiolo-
FC4 - PC4 7.5 2.3 -5.2 14 -2.3 .037 gical reactions during simulator and real flight and
C3 - P3 11.3 2.0 -9.4 15 -2.8 .013 could confirm similar results for heart rate, heart rate
C4 - P4 1.3 2.1 -8.2 15 -2.5 .026 variability, and respiratory frequency.
P´3 - O1 35.8 3.2 -32.5 16 -3.1 .007 Our analysis of the heart rate within the test profile
P´4 - O2 36.1 3.5 -32.7 15 -2.9 .011
unusual-attitude recovery demonstrates the expected
increase of heart rate due to increasing mental work-
load. Additionally, the results illustrate a significant
Bef. Rec p-
BETA Rec. (M) (M) Diff df t value lower increase of the heart rate for pilots having
received an unusual-attitude recovery training. As a
F3 - FC´3 1.2 1.7 .5 11 2.9 .015
conclusion, the effects of the evaluated training
F4 - FC´4 1.3 1.6 .3 13 1.3 .220
program can be described as increasing flight per-
FC3 - PC3 2.8 2.6 -.2 12 -.8 .423 formance together with reducing stress in demanding
FC4 - PC4 3.5 2.3 -1.2 13 -2.0 .071 flight situations. First EEG results show changes in
C3 - P3 3.5 2.4 -1.1 14 -2.9 .012 the alpha- and the beta bands due to changing strain
C4 - P4 3.5 2.1 -1.4 14 -2.2 .042 in the simulator.
P´3 - O1 5.2 3.0 -2.2 16 -3.2 .005 To sum up, this study makes a contribution to basic
P´4 - O2 5.3 3.0 -2.3 14 -3.9 .002 research by analyzing psychophysiological changes
as well as to applied science by emphasizing the
importance and effectiveness of orientation training
programs for VFR pilots.

577
Acknowledgments 197-241). Virginia: American Institute of Aero-
nautics and Astronautics.
The research was granted by the Austrian Industrial
Research Promotion Fund (FFF). Wickens, C. D. (2002). Situation Awareness and
Workload in Aviation. Current directions in Psy-
chological Science, 11(4), 128-133.
References
Wilson, G. F. (2002). An Analysis of Mental Work-
AMST Systemtechnik. Flight simulators and tech- load in Pilots During Flight Using Multiple Psy-
nical parameters. Retrieved February 10, 2005, chophysiological Measures. International Journal
from http://www.amst.co.at. of Aviation Psychology, 12(1), 3-18.

Endsley, M. R. (1995). Measurement of Situation

Awareness in Dynamic Systems. Human Factors,
37 (1), 65-84.

Endsley, M. R. (2000). Theoretical Underpinning of

Situation Awareness: A Critical Review. In
Endsley, M. R. & Garland, D. J. (Eds.), Situation
Awareness Analysis and Measurement (pp 3-32).
New Jersey: Lawrence Erlbaum Associates.

Goh, J. & Wiegmann, D. A. (2001). Visual Flight

Rules Flight into Instrument Meteorological
Conditions: An Empirical Investigation of the
Possible Causes. International Journal of Avia-
tion Psychology, 11(4), 359-379.

Hankins, Th. C. & Wilson, G. F. (1998). A Compa-

rison of Heart Rate, Eye Activity, EEG and
Subjective Measures of Pilot Mental Workload
During Flight. Aviation, Space, and Environ-
mental Medicine, 69(4), 360-367.

Haug, T. (2003). AMST SD Private-Pilot Flight Trai-

ning. Instructor Pilot Checklist. Ranshofen: Un-
published Training manual.

Hoffmann, J. (1993). Vorhersage und Erkenntnis.

[Prediction and Recognition] Göttingen: Hogrefe.

Kallus, K. W. & Tropper, K. (2004). Evaluation of a

Spatial Disorientation Simulator Training for Jet
pilots. International Journal of Applied Aviation
Studies, 4(1), 45-55.

Veltman, J. A. (2002). A Comparative Study of

Psychophysiological Reactions During Simulator
and Real Flight. International Journal of Aviation
Psychology, 12(1), 33-48.

Véronneau, S. J. H. & Evans, R. H. (2004). Spatial

Disorientation Mishap Classification, Data, and
Investigation. In F. H. Previc & W. R. Erocline
(Eds.), Spatial Disorientation in Aviation (pp.

578