Towards an Integrated Pedestrian Safety Assessment Method

Abstract

http://www.ircobi.org/downloads/irc12/pdf_files/81.pdf

Full text


I. INTRODUCTION
Whilepassivesafetyassessmentiswellestablishedinregulationandconsumertesting,activesafety
assessmenthasonlyemergedrecently.Anintegratedsafetysystemisonewhichconsistsofbothactiveand
passivesafetydevices.Assessmentofthesesystemsisinitsinfancyandisthesubjectofcurrentresearch.The
currentchallengeistodefineamethodologywhichintegratesactiveandpassiveassessmentsandtakesinto
accounttheinfluencethattheactivesafetysystemhasontheboundaryconditionsforthepassivesafety
system.Thecurrentstudyisfocusedondevelopingamethodologytoassessintegratedpedestriansystems.
Somepreviousresearchworkhasbeenperformedtodevelopmethodologiestoassessthesesystemsbutto
datenomethodologyhasbeendevelopedwhichintegratestheactiveandpassiveassessmentsfullyandtakes
intoaccounttheeffectoftheactivesafetysystemonthepassivesafetysystemboundaryconditions.The
currentresearchworkisdescribedbelowandkeyfeaturesofitaresummarisedintable1.
EuroNCAPassessestheprotectionofferedforpedestrianswithcomponenttestsatfixedtestspeedand
impactangles[1][2].Anassessmentforactivesafetysystemsiscurrentlyunderdiscussion.Testmethodsfor
activesafetysystemsareproposedby,besidesothers,AEB[3]andvFSS[4].Acombinedassessmentcannotbe
thestraightforwardcombinationofbothbecausethebenefitoftheactiveandpassivesystemaremeasuredin
differentunits:e.g.impactspeedreductionforactivesafetysystemsandinjurycriteriameasurementsforthe
passivesafetysystemcomponenttestwhichcanberelatedtoinjuryriskofthetestedbodyregion.
AcombinedassessmentmethodhasbeenproposedbySchramm[5]andRothandStoll[6].Itmakesuseof
separateactiveandpassivesafetytestsatareferencespeed(40km/h)andcalculatesMAIS2+*1injuryrisk
reductionasafunctionofspeedreductionandEuroNCAPpedestrianpointscore.Thiscombinedassessmentis,
however,notafullyintegratedassessment.Activesafetysystemsfunctionalityisassumedtohavenoinfluence
ontheconfigurationofthepassivesafetytests,forexamplenochangeinimpactareaorangleisexpected.This
dependencehasbeendemonstratede.g.byPengetal.[7].Furthermore,expectedbenefitsofthesystemsare
added,whichrequiresindependenceoftheeffectivenessofpassiveandactivesafetysystems.Thiscannot
necessarilybeassumedtobethecaseastwotechnologiesmightaddressthesametypeofinjuryasshownby
Rosenetal.[8].
VeRPS[9]hasbeenproposedasanassessmentmethodwhichcombinessimulationandcomponenttests.
Simulationisusedtodetermineimpactpoints,speed,angle,etc.forthespecificvehiclebecauseithasbeen
shownthatcargeometryandstiffness,besidesothers,influencetheseparameters.Componenttestsareused
tomeasureimpactorresponseandrelatethistoinjuryriskpotential.Thismethodologycouldbeextendedinto
anintegratedassessment.Theimpactspeedatwhichthevehiclespecificsimulationsarecarriedoutcouldbe
alteredandinteractionbetweenactiveandpassivesafetysystemscouldbereflectedinanalteredpassive
safetytest.However,alargenumberofsimulationsandtestswouldbeexpectedtoberequiredwhichmaynot
bepracticable.
Hamacheretal.[10]hasdevelopedanextensionfortheVeRPS‐Indextoincludeactivesafetytechnology.
Impactspeedreductionfromareferencetestspeedandchangesinpassivesafetyboundaryconditionsare
considered.Forsixdifferentgeometricalvehicleclasses,towhichtheassessedvehiclehastobeassigned,
adjustmentfactorsforheadimpactspeedandanglearederivedfrommulti‐bodysimulationasthemaximum
*1AIS:AbbreviatedInjuryScale,astandardmeasureforinjuryseverity[18]andMAIS:MaximumAIS,measureforoverallinjuryseverity
usingthemaximumAISofallinjuriessustained.Itreachesfrom0(uninjured)to6(currentlynottreatable).

1NilsLubbeisEngineeratTechnicalAffairsPlanningdivisionofToyotaMotorEurope(TME),Belgium,1930Zaventem,HogeWei33
(phone:+3227123969,nils.lubbe@toyota‐europe.com).2MervynEdwardsisHeadofStructuralCrashworthinessatTransportResearch
Laboratory(TRL).3MarcusWischisEngineeratPassiveSafetyandBiomechanicsdivisionoftheFederalHighwayResearchInstitute
(BASt),Germany.
NilsLubbe1,MervynEdwards2,MarcusWisch3
TowardsanIntegratedPedestrianSafetyAssessmentMethod
IRC-12-81
IRCOBI Conference 2012
- 761 -

valuesfromseveralstaturesandimpactlocations.Thelikelihoodofheadimpactatdifferentpositionsis
likewisecalculatedaprioriforthevehicleclass.Both,kinematicadjustmentfactorsandimpactprobabilitiesare
availablefor4differentimpactspeedsaswellasforanadultandchildpedestrian.Theresultingindexvalueis
animpactspeeddependentprobabilityweightedriskofAIS3+headinjury.Aseparatesimplifiedleginjuryindex
isproposedsimilarly.Thismethodologyisanintegratedassessmentasboundaryconditionsforthepassive
safetyassessmentareadjustedbasedontheinfluenceofactivesafetysystems.
Recently,Hutchinsonetal.[11]proposedanassessmenthighlightingtheimportanceofconsideringnotonly
onereferencespeedbutallimpactspeeds.Thisistoavoidsub‐optimization:Adesignoptimalforasingle
referencespeedmightnotbeoptimaloverallimpactspeeds.ThebenefitmeasureusedforillustrationisAIS2+
injuryrisk.However,thoseauthorsbelievethat“abetteroptionwouldbetotakeintoaccountalllevelsof
injury,withtheirdifferentlevelsofseriousness,andarriveatanaverage‘cost’”.

Thisshortcommunicationpapershowstheprinciplesandstepstobetakentodevelopanextensiveandfully
integratedpedestriansafetyassessmentmethod.

Table1:Elementsofexistingpedestriansafetyassessmentmethodsrelevantforanewintegratedassessment
EuroNCAP[1][2]AEB[3]vFSS[4]Schramm [5]
RothandStoll[6]VeRPS[9]Hamacheretal.[10]Hutchinsonetal.
[11]
Activesafety
tests‐
Warningand
autonomous
braking
Warningand
autonomous
braking
Bysimulation‐
Notspecified
(externaltests
protocol)
‐
Testspeed‐
Increasing
from10km/h
tonoeffect
40km/hVarious‐40km/h‐
Test
scenarios‐ 3daytime
crossing
4daytime
crossing Crossing&stationary)‐Notspecified‐
Outcomefor
eachscenario‐ Impactspeed
reductions
Impactspeed
reduction
Impactspeedreduction
transformedtoMAIS2+
injuryreductionfrom
globalinjuryrisk–
impactspeedcurve
‐ Impactspeed
reduction‐
Passive
SafetytestsComponenttests‐ ‐ComponenttestsComponent
testComponenttestsComponenttest
Bodyregions
tested
Childhead,
Adulthead,
Upperleg,Lower
leg
‐ ‐ AsEuroNCAPHeadHeadandlowerlegHead
Testspeed 40km/h‐‐AsEuroNCAP
Impactpoint
specificfrom
simulationwith
45km/h
impact
AsEuroNCAP
40km/h–
calculationoftest
valuesforwhole
incidencespeed
distribution
ImpactanglesFixed‐‐
AsEuroNCAP
(nochangesdueto
activesafety)
Impactpoint
specificAsEuroNCAP
AsEuroNCAP
(nochangesdue
toactivesafety)
Impactarea
andpoint
distribution
Fixedimpact
area.
Uniform
distribution
withintestarea
‐ ‐
AsEuroNCAP
(nochangesdueto
activesafety)
Vehiclespecific
bysimulation
(potentially
changedby
activesafety)
AsEuroNCAP
AsEuroNCAP(no
changesdueto
activesafety)
Outcome
Pointscore(body
regionspecific
thresholdsbased
oninjuryrisk
curves)
‐ ‐
Pointscoretransformed
toMAIS2+risk
reductionwithinjury‐
shiftmethod
Riskreduction
atchosenAIS
level
InjuryriskatAIS3+
level
Costfunction
frominjuryrisk,
examplerisk
reductionat
AIS2+level
Integration‐ ‐ ‐
Additionofactiveand
passivesafetyMAIS2+
reduction
Asoutlook:
Changein
impactspeed
forvehicleand
impactpoint
specific
simulations.
Passivetestsadjusted
forheadimpact
velocityandangle
vehicleclassspecific,
weightedbyimpact
speedadjusted
impactprobabilities
Asoutlook:
Changein
incidence
distribution.


IRC-12-81
IRCOBI Conference 2012
- 762 -

II. METHODS
AnintegratedpedestriansafetyassessmentmethodologyisbeingdevelopedwithintheEuropean
Commission7thframeworkprojectAsPeCSSusingaliteraturereview,accidentdataanalysis,computer
simulation,hardwaretestingandvalidationagainstreal‐worlddata[12].Themainaimistodevelopan
assessmentthatisrelatedtothebenefitthatthesystemwillofferinreal‐worldimpactsinordertoensurethat
itismeaningful.Otherobjectivesforthedevelopmentoftheassessmentinclude:
 Afullyintegratedassessmentisnecessarytoevaluatepotentiallyrelevantinteractionsofsafety
systems.Theintegratedassessmenthastomeasurebenefitsofsafetystrategieswhichreduce
impactspeedsandreduceinjuryriskforgivenimpactspeeds.Somepossibilitiesforanintegrated
assessmentmethodhavebeenshownbyHamacheretal.[10].
 Themeasureofbenefitneedstobeclearlydefinedandtherationalesforthechosenmeasure
motivated.Themethodologyneedstoconsiderallthecasualty’s(AIS2+)injuriesandnotjustthe
maximumAISinjurybecauseitisthecombinationofalltheinjurieswhichdeterminestheoutcome
forthecasualty.LowAISinjuriescanhaveasignificantriskoflongtermconsequencesandcan
thereforebecostly[13].Thederivationofacostfunctionforallinjurylevelsandbodyregionsfora
Europeanpedestrianpopulationhasnotyetbeenattempted,howeverdatafortheUSAexistsinthis
detail[14].Thebenefitneedstobeexpressedasasinglenumberwhichisindicativeoftheoverall
benefitofthesystem.Thiswillenableeasycomparisonoftheassessmentsofsafetysystemswith
differentstrategies.
 ArelevantrangeofimpactspeedsneedstobeconsideredbasedonthedescriptionofKullgren[15]
ofroadtrafficinjuriesanddependentonincidenceratesandinjuryrisk.Thisfollowstheapproachof
Hutchinsonetal[11].Itwasshownthatpedestrianprotectionhasacontributionatspeedsbeyond
regulatorytesting,withafatalityriskof50%at75km/h[16].Asingletestmightencouragesub‐
optimizationbecausethestructuretestedmightnotbedevelopedtoofferprotectionathigher
speeds[11].Similarly,thefullpotentialatlowerspeedsmightnotbereached.Assessingprotection
offeredatrelevantspeedswiththecorrespondingincidencerateswillallowaglobaloptimumfor
protectiontobereached[11].
 Injuriessustainedbyallbodyregionsandfromgroundimpactsneedtobeconsidered.Thebenefitof
improvedprotectionforaspecificbodyregionmightbeoffsetbyworseprotectionforanotherbody
region.Groundimpactmayreducetheeffectivenessofpassivesafetymeasuressimilartothewaya
notaddressedaccidentscenarioreduceseffectivenessofactivesafetysystems.Appropriate
componenttests(e.g.withheadformandlegform,butpotentiallyalsonewtoolstotestcurrently
“untestable”areas)willbeidentified.
 Boththeimpactareaaswellasimpactpointdistributionneedtobealignedwithactualimpact
probabilitiestoassessvehiclestructuresaccordingtotherisktheyimposetopedestrians.Impact
probabilitiesofparticularpedestrianbodyregionshavebeenshowntodepend,besidesother
variables,onimpactspeede.g.[7][8].Thisinfluenceneedstobeexplicitlymodelled.Forthis,full
humanbodymodelsimulationswillbecarriedout.
 Furthermore,theinfluenceactivesafetyinterventionmighthaveonimpactkinematicsneedstobe
analysedbyfullhumanbodysimulationandreflectedinthemethodology.Bothhavebeen
incorporatedinthemethodproposedbyHamacheretal.[10].
 Theassessmentmethodologyneedstobeaccurateandvalidatedagainstreal‐worlddata,aswellas
simpleandusableforvehicleassessment.

III. RESULTS
Anoutlineassessmentmethodologyhasbeendeveloped.Itconsistsoffivestepsaslistedbelow(seefigure
1).
1. Activesafetytesting:Exposure/velocitycurveshift
Driverwarningandautonomousemergencybrakingsystemswillbeassessedwithrespecttotheirabilityto
reduceimpactvelocity.Changestoimpactkinematicsduetothisinterventionwillbenotedforpassivesafety
testing.Analysisofaccidentdatawillbeusedtodefinerepresentativetestscenarios,similarto,butnot
necessarilyequivalenttothosedevelopedbyotherprojectssuchasAEB[3]orvFSS[4].Thetestscenarioswill
beweightedcorrespondingtotheircontributiontoinjuryoccurrence.Fromeachtestscenariothetypicalspeed
IRC-12-81
IRCOBI Conference 2012
- 763 -

reductionoverthewholerangeofimpactspeedswillbederived,inamannersimilarto[3].Usingthis
information,theexposure–velocitycurvesforthecorrespondingaccidentscenariowillbeadjustedtoaccount
fortheeffectoftheactivesafetysystem.
2. Passivesafetytesting:Impactormeasurement
Testswillbeconductedatoneorseveralspeedsandimpactanglestoestimateimpactorinjurycriteria
measurementsfortherelevantvehiclespeedsidentifiedinstep1.Fortheheadformimpactorthiswillinvolve
estimationofheadimpactvelocityfrompedestrianimpactspeed.Asthenumberoftestswillbelimited,a
modelforspeeddependencyforeachimpactormightbeneeded.Forheadimpactortesting,recentwork
proposedsuchamodel[11][17].Impactpointswillbechosenaccordingtotheimpactdistributionofthe
pedestrianpopulation.Boththeboundariesofthetestareaaswellasthedistributionwithinneed
investigation.Forthelateraldirection,uniformpointdistributionwillbeassumed(ase.g.in[10]).Forthe
longitudinaldirectionapotentialshiftofimpactprobabilitiesandimpactkinematicswithimpactspeedhasto
beconsidered,takingintoaccountlimitedrepeatabilityandreproducibilityduetothecomplexityofvehicleto
pedestrianaccidents.
3. Calculationofinjury:Injuryrisk
Injurycriteriameasurementsfromstep2willbeconvertedintoaninjuryestimatefortestedbodyregions
usinginjuryriskcurvesandvelocity‐exposuredatafromstep1.Availableinjuryriskcurvesmayneedtobere‐
analysedtomatchwiththepedestrianpopulationexposed.Gender,age,stature,andimpactvelocitymightbe
influential.Furthermore,injuryriskcurvesneedtobemadeavailableforallinjuryseveritylevels.
4. Calculationofcost:Socio‐economiccost
Injuryrisksfortestedbodyregionswillbeconvertedintocostsforindividualinjuries.Thiscouldbedonein
oneoftwoways.Firstly,relevantcostscouldbeestablishedfromEuropeaninsuranceandaccidentdata.The
largestcontributorstocostareusuallyproductionlossandthevaluationoflostqualityoflife,whilemedical
treatmentcostremainscomparablylow.Alternatively,USharmdata[14]mightbeweightedtowardsEuropean
pedestriancausalities.
5. Vehicleassessment:Weightingandsumming
Inthelaststep,costswillbeweightedtoaccountfornon‐testedbodyregionsandgroundimpact.These
costswillbesummedtogiveoverallsocio‐economiccostofvehiclefittedwithactiveandpassivesafety
systems.Thistotalcostwillbesubtractedfromabaselinecostrepresentingatypicalvehicletoexpressthe
socio‐economiccostintermsofasavingorbenefit.
Figure1:AsPeCSSmethodology

IRC-12-81
IRCOBI Conference 2012
- 764 -

IV. DISCUSSIONANDCONCLUSIONS
Thispaperdescribesaconceptfortheassessmentofanintegratedpedestriansafetysystem.Itincludesthe
assessmentofactiveandpassivesystemsinanintegratedmannerandtakesintoaccounttheeffectthatthe
activesystemhasontheboundaryconditionsforthepassivesystem.Itisbasedonanestimateofthereduction
ofpedestrianinjurythatwouldbeseeninthereal‐worldandhenceismeaningful.Furtherdevelopmentwill
includevalidationandcalibrationagainstrealworlddata,uncertaintyassessmentandpossiblysimplificationfor
usebystakeholderssuchasEuroNCAP.Inthefuture,hardwaretestingcouldbereplacedbysimulationtodo
moretestsinashortertime.
V. REFERENCES
[1]EuroNCAP,"AssessmentProtocol‐PedestrianProtectionVersion5.1,"2011.
[2]EuroNCAP,"PedestrianTestingProtocolVersion5.3.1,"2011.
[3]AEBtestprocedures.(2011),http://www.thatcham.org/safety/pdfs/AEB_test_procedures_nov11.pdf
[4]Niewöhner,W.,etal.,"ProposalforaTestProcedureofAssistanceSystemsregardingPreventive
PedestrianProtection,"inProceedingsofESVconference,WashingtonDC,USA,2011,paper11‐0393.
[5]Schramm,S.,"MethodezurBerechnungderFeldeffektivitätintegralerFußgängerschutzsysteme,"2011.
[6]Roth,F.andStoll,J.,"Integratedpedestrianprotection,"inPraxiskonferenzFußgängerschutz,Bergisch‐
Gladbach,2011.
[7]Peng,Y.,etal.,"Adulthandchildheadimpactconditionsasafunctionofvehiclefrontendgeometry,"in
IRCOBIconference,Krakow,2011,pp.277‐281.
[8]Rosén,E.andFredriksson,R.,"Integratedpedestriancountermeasures–Potentialofheadinjuryreduction
combiningpassiveandactivecountermeasures,"SafSci,pp.400‐407,2012.
[9]Kühn,M.,etal.,"Assessmentofvehiclerelatedpedestriansafety,"inESVconference,WashingtonDC,
2005,paper05‐0044.
[10]Hamacher,M.,etal.,"Assessmentofactiveandpassivetechnicalmeasuresforpedestrianprotectionat
thevehiclefront,"inESVconference,WashingtonDC,2011,paper11‐0057.
[11]Hutchinson,T.,etal.,"Pedestrianheadformtesting:Inferringperformanceatimpactspeedsandfor
headformmassesnottested,andestimatingaverageperformanceinarangeofreal‐worldconditions,"
TrafInjPrev,p.inpress,2012.
[12]http://www.aspecss‐project.eu/,2012.
[13]Malm,S.,etal.,"RiskofPermanentMedicalImpairment(RPMI)inRoadTrafficCrashes,"inAnnAdv
AutomotMed,2008,pp.93‐100.
[14]Blincoe,L.,etal.,"TheEconomicImpactofMotorVehicleCrashes2000,"2002.
[15]Kullgren,A.,"VailidityandReliabiltyofVehicleCollisionData.CrashPulseRecodersforImpactSeverityand
InjuryRiskAssessmentsinReal‐LifeFrontalImpacts,"Stockholm,1998.
[16]Rosén,E.andSander,U.,"Pedestrianfatalityriskasafunctionofcarimpactspeed,"AccidAnalPrev,pp.
536‐542,2009.
[17]Searson,D.,etal.,"HeadformimpacttestperformanceofvehiclesundertheGTRonpedestriansafety,"
2009.
[18]AAAM,"AbbreviatedInjuryScale,"2005.
VI. ACKNOWLEDGEMENTS
TheresearchleadingtotheseresultshasreceivedfundingfromtheEuropeanUnionSeventhFramework
Programme(FP7/2007‐2013)undergrantagreementn°285106.
TheauthorswouldliketothankallpartnersoftheAsPeCSSprojectTask1.3forthefruitfuldiscussionsand
valuableinputtothedevelopmentofthisintegratedpedestriansafetyassessmentmethod.
IRC-12-81
IRCOBI Conference 2012
- 765 -