DTIC ADA519631: The Effects of Sleep Deprivation on Flight Performance, Instrument Scanning, and Physiological Arousal in Pilots

0.68 MB·English

by  Defense Technical Information Center

Form Approved
REPORT DOCUMENTATION PAGE
OMB No. 0704-0188
Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the
data needed, and completing and reviewing this collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing
this burden to Department of Defense, Washington Headquarters Services, Directorate for Information Operations and Reports (0704-0188), 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA 22202-
4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to any penalty for failing to comply with a collection of information if it does not display a currently
valid OMB control number. PLEASE DO NOT RETURN YOUR FORM TO THE ABOVE ADDRESS.
1. REPORT DATE (DD-MM-YYYY) 2. REPORT TYPE 3. DATES COVERED (From - To)
2009 Published Journal Article Jun 2002 - Jul 2006
4. TITLE AND SUBTITLE 5a. CONTRACT NUMBER
The Effects of Sleep Deprivation on Flight Performance, Instrument Scanning, and N/A
Physiological Arousal in Pilots 5b. GRANT NUMBER
N/A
5c. PROGRAM ELEMENT NUMBER
62202F
6. AUTHOR(S) 5d. PROJECT NUMBER
Fred H. Previc,1 Nadia Lopez,2 William R. Ercoline,3 Christina M. Daluz,2 7757
Andrew J. Workman,2 Richard H. Evans,4 and Nathan A. Dillon4 5e. TASK NUMBER
P9
5f. WORK UNIT NUMBER
04
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION REPORT
NUMBER
1Northrop Grumman 3Wyle Laboratories, Inc.
1000 Wilson Boulevard Life Sciences Group
Arlington, Virginia 22209-2278 1290 Hercules Drive
2Air Force Research Laboratory Houston, TX 77058
Human Effectiveness Directorate 4General Dynamics
Biosciences and Performance Division Advanced Information Services
Biobehavior, Bioassessment & Biosurveillance 5200 Springfield Pike
Brooks City-Base, TX 78235 Dayton, OH 45431
9. SPONSORING / MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR’S ACRONYM(S)
Air Force Materiel Command Biobehavior, Bioassessment & Biosurveillance Branch 711 HPW/RHP
711 Human Performance Wing Brooks City-Base, TX 78235
Air Force Research Laboratory 11. SPONSOR/MONITOR’S REPORT
Human Effectiveness Directorate NUMBER(S)
Biosciences and Performance Division AFRL-HE-BR-JA-2006-0021
12. DISTRIBUTION / AVAILABILITY STATEMENT
Approved for public release; distribution unlimited.
13. SUPPLEMENTARY NOTES
Published in the International Journal of Aviation Psychology, 19(4), 326-346, 2009. Public Affair approval No. 06-279; 9 Aug 2006.
14. ABSTRACT
The effects of 34 hr of continuous wakefulness on flight performance, instrument scanning, subjective fatigue, and EEG activity were
measured. Ten fixed-wing military pilots flew a series of 10 simulator profiles, and root mean squared error was calculated for various
flight parameters. Ocular scan patterns were obtained by magnetic head tracking and infrared eye tracking. Flying errors peaked after
about 24 to 28 hr of continuous wakefulness in line with peaks in subjective fatigue and EEG theta activity, and they were not directly
attributable to degradation of instrument scanning, which was very consistent across pilots and largely unaffected by the sleep deprivation.

15. SUBJECT TERMS
Fatigue, sleep deprivation, cognitive performance, EEG theta activity
16. SECURITY CLASSIFICATION OF: Unclassified 17. LIMITATION 18. NUMBER 19a. NAME OF RESPONSIBLE PERSON
U OF ABSTRACT OF PAGES Fred H. Previc
a. REPORT b. ABSTRACT c. THIS PAGE U 21 19b. TELEPHONE NUMBER (include area
U U U code)

Standard Form 298 (Rev. 8-98)

Prescribed by ANSI Std. Z39.18
THE INTERNATIONAL JOURNAL OF AVIATION PSYCHOLOGY,19(4),326–346
Copyright ©2009Taylor & Francis Group, LLC
ISSN: 1050-8414 print / 1532-7108 online
DOI: 10.1080/10508410903187562
The Effects of Sleep Deprivation
on Flight Performance, Instrument
Scanning, and Physiological Arousal
in Pilots
0
1
0
2

h
c
r
Ma Fred H. Previc,1Nadia Lopez,2William R. Ercoline,3

15 Christina M. Daluz,2Andrew J. Workman,2

6
:3 Richard H. Evans,4and Nathan A. Dillon4
8
1
: 1Northrop Grumman, San Antonio, Texas
t
A 2Brooks City-Base, San Antonio, Texas
]
y
r 3Wyle Laboratories, San Antonio, Texas
a
r
ib 4General Dynamics Advanced Information Engineering Systems,
L
l San Antonio, Texas
a
c
i
d
e
m
o
r
e
A
[ Theeffectsof34hrofcontinuouswakefulnessonflightperformance,instrument
:
By scanning,subjectivefatigue,andEEGactivityweremeasured.Tenfixed-wingmili-

ed tarypilotsflewaseriesof10simulatorprofiles,androotmeansquarederrorwascal-
d
oa culatedforvariousflightparameters.Ocularscanpatternswereobtainedbymag-
l
wn neticheadtrackingandinfraredeyetracking.Flyingerrorspeakedafterabout24to
o
D 28hrofcontinuouswakefulnessinlinewithpeaksinsubjectivefatigueandEEG
thetaactivity,andtheywerenotdirectlyattributabletodegradationofinstrument
scanning,whichwasveryconsistentacrosspilotsandlargelyunaffectedbythesleep
deprivation.
Fatigue due to sleep deprivation is considered a major risk to flight safety
(Borowsky&Wall,1983;Ramsey&McGlohn,1997;Tormes&Guedry,1975),
withsurveyssuggestingthatuptohalfofallpilotshaveactually“dozedoff”while
flying(Caldwell&Gilreath,2002).Fatiguedegradesnotonlybasiccognitiveper-
Correspondence should be sent to Fred H. Previc, 10906 Whispering Wind, San Antonio, TX
78230. E-mail: [email protected]
SLEEP, FLIGHT, AND SCANNING 327
formance (Caldwell, Caldwell, Brown, & Smith, 2004; Dinges et al., 1997;
Matthews,Davies,Westerman,&Stammers,2000)butalsoflightperformance,
including the ability to maintain designated flight parameters (Caldwell et al.,
2004; Caldwell, Caldwell, & Darlington, 2003; LeDuc et al., 1999; Morris &
Miller,1996).Whethertherisktoflightsafetyisdueprimarilytoareducedgen-
eral cognitive capacity or to flying-specific factors (e.g., stick control or instru-
ment scanning) has yet to be determined.
Sleepdeprivationexperimentshavegenerallyshownthatperformancereaches
anadirintheearlymorninghours,typicallyafter24hrofsustainedwakefulness
andatthetroughofthecircadiancycle,beforeimprovingslightlyasthesecond
dayprogresses(Caldwelletal.,2004;Caldwelletal.,2003).Thereboundonthe
seconddayispartlyduetotheeffectofthecircadiancyclethatrisesduringtheday
0
1
0 (Eddy&Hursh,2001)andalsototheimpendingcompletionoftheexperiment.In
2

ch previous simulator studies, helicopter pilots showed an earlier nadir than fixed-
r
Ma wing pilots, possibly because of their different daily schedules (Caldwell et al.,

5
1 2004; Caldwell et al., 2003).

6
:3 Onepossiblecorrelateofthedecrementsinflightperformanceisimpaired
8
1 ocularscanningoftheflightinstruments.Fivebasiceye-movementparameters
:
t
A thathavebeenrepeatedlystudiedinconjunctionwithfatigueare(a)blinkrate,

]
ry which generally increases with sleep deprivation and fatigue (Lal & Craig,
a
br 2001;Morris&Miller,1996;Stern,Boyer,&Schroeder,1994);(b)pupildi-
i
L
ameter, which typically decreases with sleep deprivation (Morad, Lemberg,
l
a
ic Yofe, & Dagan, 2000; Ranzijn & Lack, 1997; Wilhelm, Wilhelm, Ludtke,
d
me Streicher&Adler,1998;Yoss,Moyer&Hollenhorst,1979);(c)saccadicve-
o
r
e locity,whichhasbeenshowntodecreasewithsleepdeprivation(Caldwellet
A
[
: al.,2004;DeGennaro,Ferrara,Urbani,&Bertini,2000;Rowlandetal.,2005;
y
B Russo et al., 2003; but see Morris & Miller, 1996); (d) mean saccade length
d
e
d (fixation distance), which increases with time-on-task (Lavine, Sibert, Gok-
a
o
nl turk, & Dickens, 2002); and (e) dwell time, which in at least one study was
w
Do shown to decrease with time-on-task (Lavine et al., 2002). In contrast to the
preceding basic eye-movement studies, no previous study has investigated
changes in pilot instrument scanning with extended wakefulness, although
therehavebeenseveralstudiesofpilotscanningbehaviorundernormalwake-
fulness(Bellenkes,Wickens,&Kramer,1997;Itoh,Hayashi,Tsukui,&Saito,
1990;Jones,Milton,&Fitts,1949).
Thechiefpurposeofthisstudywastoinvestigatechangesinflightperformance
duringextendedwakefulnessofover30hrandtodeterminetherelationshipbe-
tween fatigue-related flight performance decrements and both general changes
(e.g.,increasedblinkrate)andspecificchanges(e.g.,reducedscanningofspecific
instruments)inoculomotorbehavior.Anotherobjectivewastorelatechangesin
flightperformanceandinstrumentscanningtochangesinsubjectiveandobjective
fatigue and arousal, the latter being measured by the scalp-recorded electroen-
328 PREVIC ET AL.
cephalogram(EEG).Previousresearchhasdemonstratedthatincreasesinlow-fre-
quency(deltaandtheta)EEGactivitygenerallyparallelchangesinperformance
during sleep deprivation, fatigue, or both (Caldwell et al., 2004; Lal & Craig,
2001; Morris & Miller, 1996).
METHOD
Participants
TenpilotsfromtheUnitedStatesAirForce(USAF)participatedinthisstudyinan
off-dutycapacity.Eightofthe10pilots(allmale)wereactive-dutypilotsandthe
10 remaining 2 were reserve officers. The average age of the pilots was 34.2 years
0
2 (range=23–46years),withhalfofthepilotsover35yearsandhalfat30yearsor
h
c
r under.Theiraverageflightexperiencewas2,806hr(range=207hr–5,800hr),1
a
M
5 andthecorrelationbetweenflightexperienceandagewasalmostperfect(r=.96).
1
6 All pilots signed an informed consent document approved by the Brooks City-
3
:
8 BaseInstitutionalReviewBoardandwerecompensatedfortheirparticipationbe-
1

t: cause they served in an off-duty capacity.
A
] All pilots possessed at least 20/25 vision binocularly (they were allowed to
y
r
a wearcontactlensesbutnotglassesforcorrection),hadnormalvestibularfunction
r
b
Li asassessedbytheSharpenedRombergTest,andhadnopreviousevidenceofves-

al tibularsymptomssuchasdizziness,vertigo,anddisorientation.Nopilotsuffered
c
i
d fromsleepproblemsorseizures,andnonewascurrentlytakinganypsychoactive
e
m
ro medication(e.g.,antihistamines,antidepressants,sleepaids,etc.)orwasahabitual
e
[A smoker(i.e.,consumedmorethanonecigaretteperday)orcaffeinedrinker(i.e.,

:
y consumedmorethan100mgofcaffeineperday).Allpilotsrefrainedfromcaf-
B

ed feine,alcohol,andothermildstimulantsorsedativeswhilemonitoredathomeon
d
oa thenightbeforethesleep-deprivationperiodaswellasduringthe34hrofcontinu-
l
n
w ouswakefulnessinthelaboratory.Nineofthe10pilotscompletedasleeplogfor
o
D
the7dayspriortothestartoftheexperimentandhadtheirfinalnightofsleep(af-
terthefirstnightoftraining)monitoredbyawrist-activitymonitor.2Thesleepdu-
rationandqualityforthe3dayspriortothebeginningoftheexperimentwereana-
lyzed using the Fatigue Avoidance Scheduling Tool (FAST; Eddy & Hursh,
2001),andabaseline“wakingefficiency”scorewasderivedforeachpilottopre-
dicthislevelofalertnesspriortotheperiodofcontinuouswakefulness.Theaver-
age amount of sleep per night was 7.46 hr and was associated with an average
“wakingefficiency”scoreof90.59,withonlythreepilotsscoringbelow90(76.8,
85.2, 85.9).
1Three of our pilots had just completed undergraduate pilot training.
2One pilot misplaced his sleep log and failed to return it.
SLEEP, FLIGHT, AND SCANNING 329
Apparatus: Gyroflight Sustained Operations Simulator
This study was conducted in the Gyroflight Sustained Operations Simulator
(GSOS; Environmental Tectonics Corporation, Southampton, PA), a four-axis
flightsimulatorwithadditionalspatialdisorientation-producingcapabilities.The
GSOSwascolocatedwiththeAviationSustainedOperationsLaboratoryinBuild-
ing170atBrooksCity-Base,Texas.TheGSOSpossessesmotioncapabilitiesin
pitch(upto±25°),roll(upto±25°),andyaw(upto360°ofsustainedyaw).The
GSOSalsofeaturessubthresholdwashoutinpitchandrollaswellaslimitedheave
(upto±12cm).Ithasathree-channelhigh-resolution,noncollimated,out-the-win-
dowvisualdisplay,withatotalfield-of-viewof28°verticalby~120°horizontal.
The GSOS aeromodel replicates the T-6 aircraft, with which most of the pilots
10 werefamiliar,anditsreconfigurableinstrumentpanelwasalsodesignedtodepict
0
2 ascloselyaspossiblethepanelontheT-6aircraft.TheGSOSwasoperatedand
h
c
r monitoredfromacontrolstationinanadjacentroom,anditsphysiologicalrecord-
a
M
5 ing capability was configured to record eye movements and EEG.
1

6
3
:
8 Eye-Movement Recordings
1

:
t
A EyemovementswererecordedbymeansofEye-Trac6000(AppliedScienceLab-

]
ry oratories, Cambridge, MA), a head-mounted system that consists of a magnetic
a
br headtrackerandaninfraredeyetracker.The“FlockofBirds”headtracker(As-
i
L
cension,Burlington,VT)providessixdegree-of-freedomtrackingbymeansofa
l
a
ic 21-Hzmagneticpulsesignaldirectedtowardasensorattachedtothehead.Thepo-
d
me sitionoftheeyeintheorbitwassampledat60Hzusinganinfraredbeamandcam-
o
r
e eratomeasuretherelativeanglesofthepupilandcornealreflectance.Together,
A
[
: theheadandeyesignalsdeterminedgazewithanerrorof<0.5°duringcalibration.
y
B TheeyeandheadsignalswerethensenttoacomputerontheGSOSandrelayed
d
e
d throughtheGSOSslipringstoamonitorlocatedinthecontrolstation.Datawere
a
o
nl stored on a PC and analyzed later.
w
o
D
EEG Recordings
EEGswererecordedusingtheGRASS-TelefactorInstrumentsAurorarecording
system (West Warwick, RI) running TWin™ collection and analysis software.
EEGswererecordedfromgold-cupelectrodesattwosites(C andP )andwere
z z
referencedtolinked-mastoidelectrodes,whileanadditionalgroundleadwasat-
tachedtothescalp.EEGswererecordedwithcutofffilterssetat1Hzand70Hz
and were digitized at 200 Hz. Data were then stored on a PC and analyzed later.
Procedures
Flight profile. The GSOS flight profile, shown in Figure 1, consisted of
sevenmajorsegments:(a)takeoffat360°andclimbto8,000ft;(b)arightclimb-
330 PREVIC ET AL.
0
1
0
2

h
c
r
a
M

5
1

6
3
:
8
1

:
t
A

]
y
r
a FIGURE1 TheGyroflightSustainedOperationsSimulator(GSOS)profilewiththeseven
r
ib segmentsinwhichvariousmeasuresofflightperformancewererecorded:(a)takeoffat360°
L
l andwings-levelclimbto8,000ft;(b)rightclimbingturnto10,000ft;(c)wings-levelclimbto
a
ic 12,000ft;(d)rightlevelturnto180°;(e)wings-leveldescentto7,500ft;(f)leftdescendingturn
d
e to 4,000 ft; and (g) visual descent and landing at 360°.
m
o
r
e
A
[

:
y
B ingturnto10,000ftand235°;(c)awings-levelclimbto12,000ft;(d)arightlevel
d
e
d turn to 180°; (e) a wings-level descent to 7,500 ft; (f) a left descending turn to
a
o
nl 4,000ftand90°;and(g)visualdescentandlanding.3Theflight,whichrequired
w
Do about19mintocomplete,simulatedatransitionfromadusktakeofftoanighttime
landing and was performed mostly in instrument meteorological conditions
(IMC).TheexceptionstoIMCwereduringabriefperiodaftertakeoff,duringa
smallsectionofthewings-levelclimbwhilepilotssearchedfortraffic,andduring
theturntofinalapproachfollowedbythevisualapproachandlanding.Oneach
segment,thepilotwascommandedtomaintainasetofpreviouslyspecifiedcon-
trolorperformanceparameters,includingairspeed(allsegments),heading(Seg-
ments1,3,and5),verticalvelocity(Segments2,3,5,and6),bank(Segments2,4,
and6),andlongitudinalbearingandglideslope(Segment7).Onodd-numbered
flightsthepilotflewasalreadydescribed,andoneven-numberedflightsthepilot
3Inanadditionalsegment,whichwastoovariabletoallowcommandedparameters,thepilotwas
required to achieve and maintain a heading at 45° to intersect final approach.
SLEEP, FLIGHT, AND SCANNING 331
flewamirrorprofile,beginningwithaclimbtotheleftfollowedbyawings-level
climbat125°ratherthanaclimbtotherightfollowedbyawings-levelclimbat
235°.TheGSOSprofilewasdesignedtobesemiautomatedandrequiredtheoper-
atortodirectlyinstructthepilotonlyduringgrossflighterrorssuchasthewrong
turning direction, wrong course, or wrong heading.
On four of the flights (1, 4, 7, and 10), seven spatial disorientation conflicts
were inserted in during various segments of the flight. These conflicts were de-
signedtotesttheeffectsofsleepdeprivationonspatialdisorientation(Previcetal.,
2007). They involved either motion illusions (an excess pitch sensation during
takeoff in Segment 1, a Coriolis illusion during head tilt in Segment 4, and
postrotatorysensationsfollowingrolloutfromtheturnsattheendofSegments2,
4, and 6) or visual illusions (a sloping cloud deck in Segment 3 and a narrow,
0
1
0 up-slopingrunwayinSegment4).Pilotswereinstructedtoreportanydiscrepan-
2

ch cies or conflicts with their instruments but were not informed in advance of the
r
Ma specificillusions.Onlytwooftheconflictswereshowntoinfluenceflightperfor-

5
1 manceduringmeasuredepochs:theslopingclouddeckonbankduringthelong

6
:3 wings-levelascentinSegment3andtheillusoryrunwayonglideslopeduringthe
8
1 landinginSegment7.However,theclouddeckeffectwasveryslight(<1°)and
:
t
A waspresentforonlyasmallportionofSegment3,andglideslopeinSegment7

]
ry waslaterremovedfromthecompositeerrormeasure.Anyremainingeffectofcon-
a
br flictversusnonconflictflightswasdeterminedtobenonsignificantinaprelimi-
i
L
naryanalysis,sothatthedatawerethencollapsedacrossconflictandnonconflict
l
a
ic flights.Becausetransitionsbetweenflightmaneuvershadtobeeliminatedwhile
d
me pilots were in the process of attaining their commanded flight parameters, only
o
r
e about 50% of the total flight was used for data analysis.
A
[
: Root mean squared error (RMSE) was used as the measure of flight perfor-
y
B mance. The RMSE values were calculated using Equation 1:
d
e
load RMSE=∑n (i−c)2/n, (1)
n
w
Do i=1
whereiis the observed value andcis the commanded flight parameter.
EachsegmenthadthreemeasuresexceptSegment1,whichonlyhadairspeed
andheading,andSegment7,becauseslopewasremovedduetotheeffectsofthe
spatialdisorientationconflictsonfouroftheflights.TocomputecompositeRMSE
valuesfordifferentflightsegmentsandparametersaswellasfortheentireflight,
allRMSEvaluesweredividedbythebaselinevalue(theRMSEinFlight1)and
then converted to log units before averaging. The grand composite average was
based on a total of 19 individual values.
Eye-movement calibrations and recordings. Eye movements were re-
cordedduringallflights.Priortothebeginningoftheexperiment,theheadtracker
wascalibratedbyplacingthemagneticsensorinapointerrodthatwasaimedat
332 PREVIC ET AL.
differentportionsoftheGSOSinstrumentpanel.Justpriortoeachflight,andafter
thepreflightrestingEEGrecordingssession,pilotssatdownintheGSOSandthe
infraredcamerawasadjustedtoobtainagoodimageoftheeye.Then,theGSOS
doorwasclosedandtherestoftheeyecalibrationwasmonitoredfromtheGSOS
controlroom.Duringthenine-pointeyecalibration,thepilotscannedmostofthe
GSOSinstrumentpanelwhilehisheadremainedstill.Acalibrationwasconsid-
eredsuccessfulifthecalculatedgazewasnomorethan1°offatanysinglepoint
(andlessthanthatonaverage).Typically,thecalibrationhadtoberepeatedoneor
moretimeswhiletheoperatoroptimallyadjustedtheilluminationandsensitivity
oftheinfraredcamera.Initialcalibrationsweresuccessfulon98%offlights;how-
ever,manyeye-movementrecordswerelaterdiscardedbecauseeitherthepupilor
cornealimageswerelostformorethan15%ofthesamplesduringtheflightoron
0
1
0 visual inspection there was too much variable drift in the eye-movement record
2

ch from the beginning to the end of the flight (see below).
r
Ma The eye-movement data collected during each of the seven segments corre-

5
1 spondedmostlywiththeperiodsinwhichflightperformancedataweregathered.

6
:3 Unlike the flight performance data, however, the eye-movement data were re-
8
1 cordedevenwhilethepilotrolleddowntherunwayinSegment1andtheywerere-
:
t
A cordeduntiltouchdowninSegment7.Therewereatotalof22measuresobtained

]
ry from the eye-movement recordings. Five of these were basic measures: average
a
br pupildiameter,averageblinkrate,meanfixationduration(definedaseyeposition
i
L
remainingwithin1SD[0.5°]foratleastsixconsecutivesamples,or83msec),av-
l
a
ic eragesaccadelength,andpercentageofdwelltimesgreaterthan2sec.4Unfortu-
d
me nately,saccadicvelocitycouldnotbemeasuredbytheEye-Trac6000systembe-
o
r
e cause of the relatively slow head sampling rate. There were also 17 measures
A
[
: relatedtothepilot’sinstrumentscan.Theseincludedthepercentageofdwellson
y
B eachoffiveflightinstruments—theelectronicattitudedirectorindicator(EADI),
d
e
d airspeedindicator,altimeter,horizontalsituationindicator(HSI,alsoknownasthe
a
o
nl headingindicator),andtheverticalvelocityindicator(VVI)—aswellastheper-
w
Do centageofdwellsofftheinstrumentpanelaltogether.Therewerealso10measures
of transitioning to and from the five flight displays as well as a measure of
transitioningtoandfromtheinstrumentpanelasawhole.Indeterminingthedwell
andtransitionpatternsforthefivemajorflightdisplays,theiroutlinesonthein-
strumentpanelspaceweremappedtothecalibrationspacefortheeyetrackerand
superimposed on the scan pattern from each flight, as shown in Figure 2.
Forreasonsthatareunclear,therewasaslightdriftofgazepositionrelativeto
thecalibrationinmostoftheinflighteye-trackingrecords.However,thedriftwas
almostalwaysconstantwithinagivenflight,whichallowedustoperformasingle
4Dwelltimereferstotheamountoftimespentcontinuouslyinoneareaofinterest,whetherinmul-
tiplefixationsornot.The2seccriterionforlongdwelltimeswasbasedonthefactthat,inpreliminary
data, less than 10% of all dwell times were greater than that value.
SLEEP, FLIGHT, AND SCANNING 333
0
1
0
2

h
c
r
a
M

5
1

6
3
:
8
1

:
t
A

]
y
r
a
r
b
Li FIGURE2 Arepresentativescanpatternoveranentireflight,withtheoutlinesofthefive

al designated flight instruments shown in white boxes.
c
i
d
e
m
ro recenteringoftheentirescanpatternforeachflightpriortoanalysisandretainall
e
A
[ butfourrecordswithexcessivevariabledrift.Therecenteringwasconductedby

:
By twooftheexperimenters(NLandWRE)beforeanydataanalysiswasperformed.

ed Somerecordswithvariabledriftbutnormalblinkrateswereretainedforthebasic
d
a
o oculomotor analyses, while other records showing blink rates greater than 1/sec
l
n
ow (i.e.,greaterthan3SDaboveourmeanof0.35/sec)werediscardedfromthebasic
D
oculomotoranalysesbutretainedforthescan-patternanalysis.Intheend,71of
100recordswereretainedforthescan-patternanalysisand64of100recordswere
usedinthebasicoculomotoranalyses,includingatleasttwoeachfromtheearly
and late flights of 9 of the 10 pilots.5
EEG recordings and analysis. Electrodeswereattachedtoeachpilotafter
the final training session. Each placement site was cleaned with acetone, after
which electrodes were attached to the scalp with collodion and then filled with
5Onlyoneeye-movementrecordwassalvageablefromthetenthpilot,butitwasnotstatisticallyan-
alyzablebecausetherewerenootherrecordsfromthatpilot.Hence,allofthatpilot’sgazedatawere
discarded.
334 PREVIC ET AL.
electrolytegel.Afterimpedanceswerecheckedanddeterminedtobeatacceptable
levels(<5KΩ),theEEGelectrodesremainedonthroughouttheentire34hrof
continuous wakefulness.
Pre-andpostflightrestingEEGswererecordedfor4mineach,withtheeyes
open (2 min) and the eyes closed (2 min). Preflight EEGs were recorded in the
GSOScontrolareajustbeforethepilot’sgazewascalibrated,whereaspostflight
EEGs were recorded immediately following each flight in the GSOS and moni-
toredfromthecontrolstation.In-flightEEGswerealsorecorded,butthesewere
contaminatedbytheinteractionoftheeyetracker’smagneticpulsegeneratorand
themovementoftheGSOS,aswellasbypilots’eyeandlimbmovements.Unfor-
tunately,postflightdatawerenotrecordedproperlyfor19of100flights,sothat
onlythepreflightrestingEEGsweresubjectedtoanalysis,althoughonepilotwith
0
1
0 three missing records was removed even from that analysis.
2

ch Atotalofthree3-secepochs,eachselectedbecausetheywerefreeofanyobvi-
r
Ma ous muscle or blink artifacts, were selected from each EEG record. EEGs from

5
1 eachelectrodesiteandconditionwereanalyzedseparatelyfortheirFourierampli-

6
:3 tude in each of three bands: deltas (1.5–3.0 Hz), theta (3.0–8.0 Hz), and alpha
8
1 (8.0–13.0 Hz).
:
t
A

]
ry Overall schedule. Pilotsarrivedthenightbeforethebeginningofthecon-
a
br tinuouswakefulnessperiodforinitialtrainingontwoversionsoftheflightprofile
i
L
(neitherofwhichwasembeddedwithspatialdisorientationconflicts)andonthree
l
a
ic cognitivetests:thePsychomotorVigilanceTest(PVT),theMulti-AttributeTask
d
me Battery(MATB),andtheOperationSpanTask(OSPAN).Thesecognitivetests
o
r
e andtheircorrelationswithchangesinflightperformanceduringcontinuouswake-
A
[
: fulness are described elsewhere (Lopez, Previc, Fischer, DaLuz, & Workman,
y
B 2009).Duringalltrainingflights,pilotswereoutfittedwiththehead-mountedop-
d
e
d tical device and their gaze measurement was tested for its adequacy.
a
o
nl Aftertheirmonitoredsleep,pilotsarrivedbackinthelaboratoryateither0730or
w
Do 0830andflewtheGSOSprofileforafinalpracticeflightandreceivedadditional
trainingonthecognitivetests.AtmidmorningonDay1ofcontinuouswakefulness,
eachpilothadEEGelectrodesattachedand,afteraperiodofrestandlunch,began
the experiment. During an experimental session, two pilots were run in tandem,
withPilot1’sfirstflightbeginningat1200andPilot2’sfirstflightbeginningat
1300.Successiveflightswererunat3-hrintervals(i.e.,1500,1800,2100,etc.,for
Pilot1;1600,1900,2200,etc.,forPilot2).6Pilotsarrived5minpriortoeachflight
so their resting EEG could be recorded and their eye-tracking calibration com-
pleted, and they remained 5 min following each flight for the postflight resting
EEG measurement. The complete sequence of flights is shown in Table 1.
6Becausetheflighttimeswerestaggeredby1hrforthetwopilots,thesessiontimesinthedatafig-
ures (next section) are listed as 1230, 1530, 1830, and so on.