assessment of broward county’s … and recommendations of broward county’s regional e911...

BrowardCounty Page1 ©Fitch&AssociatesAssessmentandRecommendationsofBrowardCounty’sRegionalE911 December2016

FINALREPORT–FINDINGSANDRECOMMENDATIONS

DECEMBER2016

ASSESSMENTOFBROWARDCOUNTY’SREGIONALE911

CONSOLIDATEDCOMMUNICATIONSYSTEMS

Submitted by:

FITCH & ASSOCIATES, LLC

2901 Williamsburg Terrace #G, Platte City, Missouri, 64079 816.431.2600 www.FITCHassoc.com

Broward County, Florida © Fitch & Associates, LLC Assessment of Regional E911 Page | 1 26 October 2015 Consolidated Communications System

26 October 2015

Response to Request for Proposal:

ASSESSMENT OF BROWARD COUNTY’S REGIONAL E911

CONSOLIDATED COMMUNICATIONS SYSTEMS

BID NO: N1388702R1

Prepared by:

2901 Williamsburg Terrace #G � Platte City � Missouri � 64079 P: 816.431.2600 � F: 816.431.2653

www.fitchassoc.com

BrowardCounty Pagei ©Fitch&AssociatesAssessmentandRecommendationsofBrowardCounty’sRegionalE911 December2016

ASSESSMENTOFBROWARDCOUNTY’SREGIONALE911

CONSOLIDATEDCOMMUNICATIONSYSTEMS

DRAFTFINALREPORT-FINDINGSANDRECOMMENDATIONS

TableofContents

INTRODUCTION ______________________________________________________________________________ 1

EXECUTIVESUMMARY_________________________________________________________________________ 2

RECOMMENDATIONS___________________________________________________________________________ 2OPTIONSDEVELOPMENT ________________________________________________________________________ 2

Figure1.ComparisonofCurrentandThreeOptions“Seats”Required_______________________________________3

OBSERVATIONALFINDINGS_____________________________________________________________________ 4

FINDINGSANDACTIONABLERECOMMENDATIONS__________________________________________________ 5

TECHNOLOGY________________________________________________________________________________ 5Findings—______________________________________________________________________________ 5

OPERATIONALOVERSIGHTANDSYSTEMGOVERNANCE ____________________________________________________ 6Findings—______________________________________________________________________________ 6

PERFORMANCEMEASURES_______________________________________________________________________ 6Findings—______________________________________________________________________________ 6

EFFECTIVENESSANDEFFICIENCY____________________________________________________________________ 7Findings—______________________________________________________________________________ 7

METHODOLOGY______________________________________________________________________________ 9

DISPATCHCENTERBESTPRACTICES _____________________________________________________________ 10Figure2.RequirementsforIAEDMedicalDispatchCenterAccreditation____________________________________11

CURRENTORGANIZATIONANDENVIRONMENT ___________________________________________________ 13

COUNTYDEMOGRAPHICS_______________________________________________________________________ 13CurrentandHistorical ____________________________________________________________________ 13

Figure3.BrowardCountyPopulationGrowth,2005to2015______________________________________________14ProjectedGrowthto2020_________________________________________________________________ 14

Figure4.PopulationProjectionsfor70+YearsAgeCohorts_______________________________________________15STAKEHOLDERRELATIONSHIPS____________________________________________________________________ 16

Participants ____________________________________________________________________________ 16TECHNOLOGYREVIEW _________________________________________________________________________ 17

Telecommunications _____________________________________________________________________ 18ComputerAidedDispatchSystem___________________________________________________________ 19RadioOperations________________________________________________________________________ 20DispatchFacilities________________________________________________________________________ 20

FINANCIALSTRUCTURE_________________________________________________________________________ 21Figure6.BSOExpenditureandBudgetHistory_________________________________________________________22

BrowardCounty Pageii ©Fitch&AssociatesAssessmentandRecommendationsofBrowardCounty’sRegionalE911 December2016

STAKEHOLDERINPUT_________________________________________________________________________ 23

METHODOLOGY _____________________________________________________________________________ 23StakeholderInput________________________________________________________________________ 23

STAKEHOLDERPERCEPTIONS–LEVEL1INTERVIEWS _____________________________________________________ 23PositiveIssuesRelatedtotheRegionalE911System ____________________________________________ 24

Figure7.GoalsfortheBrowardRegionalE911System __________________________________________________24Figure8.HistoryofCallTransfersBetween911Centers _________________________________________________24

IssuesofConcernRelatedtoRegionalCommunications__________________________________________ 25STAKEHOLDERPERCEPTIONS-LEVEL2&3INTERVIEWS___________________________________________________ 28

Teamwork _____________________________________________________________________________ 28PersonnelIntegration_____________________________________________________________________ 29InefficientProcedures/Processes____________________________________________________________ 29On-goingTrainingandAccountability________________________________________________________ 29QualityImprovement/Assurance____________________________________________________________ 29EquipmentFailuresandEmergencyProcedures________________________________________________ 29StaffingandWorkSchedules_______________________________________________________________ 30WorkEnvironment/Respect________________________________________________________________ 30

DISPATCHERANDMANAGEMENTSURVEYS ___________________________________________________________ 30SurveyParticipantDemographics___________________________________________________________ 31

Figure9.WorkLocationsofDispatcherandManagementSurveyRespondents_______________________________31Figure10.DispatcherSurveyRespondents’911WorkExperience _________________________________________32Figure11.ManagementSurveyRespondents’911WorkExperience _______________________________________32

SummaryofSurveyResults ________________________________________________________________ 33Figure12.Ibelieveweprovideagoodlevelofservicetocitizenswhocall911._______________________________33Figure13.Pleaseratethefollowing:Ibelieveweprovideagoodlevelofsupporttopublicsafetyfieldpersonnel.___33Figure14.CallersforemergencyservicesprovideaccurateinformationregardingtheADDRESSoftheemergency.__33Figure15.WhenIbeganmycurrentjob,theinitialtrainingIreceivedpreparedmewellforthework.____________33Figure16.Theongoingtraining,Ireceivecontinuestoenhancemyskills.___________________________________33Figure17.TheRegionalCommunicationsSystemisequippedandpreparedtohandlelargescaleemergenciessuchashurricanesormassshootingincidents._______________________________________________________________34Figure18.Theworkmethodsweutilizehelpimprovetheefficiencyinourwork._____________________________34Figure19.Thetechnologiesweutilizeimproveourefficiencycarryingoutourwork.__________________________34Figure20.Policiesandproceduresareeasilyunderstoodandapplied.______________________________________34Figure21.Icaneffectivelyusetechnologytolocatewirelesscallerswhodon’tknowtheirlocation.______________34Figure22.Thetechnologyandinformationsystemsweusearereliableandareappropriatetothejob.___________34Figure23.CADhasthetoolsIneedtohandleincidentsefficiently._________________________________________34Figure24.EquipmentproblemsarehandledappropriatelyandIgetfeedbackonproblemsIreport. _____________35Figure25.OtheroccupantsofthebuildingIworkattreatmewithrespect.__________________________________35Figure26.Uppermanagementsupportsouroperations._________________________________________________35Figure27.ThereiscleardivisionbetweentheCountyandBSOonwhomanagesthecommunicationscenter.______35Figure28.Dutyofficersandsitemanagersareavailableandwillingtohelpmewithproblemsorconcerns.________35Figure29.Ireceivefeedbackonmyjobperformance,includingpositiveacknowledgement.____________________35Figure30.Pleaseratethefollowing:Differentworkscheduleswillimproveourcurrentstaffingchallenges.________35Figure31.Publicsafetyfieldpersonneltreatthedispatchcenterpersonnelprofessionally._____________________36Figure32.ManagementgivesteammembersaclearpictureofthedirectionBSOCommunicationsisheaded. _____36Figure33.Managementunderstandsthedailyproblemswefacewithourjobs.______________________________36Figure34.Overall,Iamsatisfiedwiththejobbeingdonebymyimmediatesupervisor.________________________36Figure35.Managementencouragesotherstoproposenewandinnovativeideas.____________________________36

BrowardCounty Pageiii ©Fitch&AssociatesAssessmentandRecommendationsofBrowardCounty’sRegionalE911 December2016

Figure36.Managementeffectivelydealswithmisconductorunsatisfactoryperformance. _____________________36Figure37.Pleaserankthefollowingissuesinorderofimportance(1isyourtopconcernand5istheleastconcern._37Figure38.Pleaserankthefollowingissuesinorderofimportance(1isthemostsatisfiedtoyouand5istheleastsatisfied._______________________________________________________________________________________37

DATAANALYSES_____________________________________________________________________________ 38

SOURCESOFDATA____________________________________________________________________________ 38Background ____________________________________________________________________________ 38CADExport_____________________________________________________________________________ 38TelephonyExport________________________________________________________________________ 38RadioExport____________________________________________________________________________ 39

CADANDCDRTIMELINES ______________________________________________________________________ 40Relationship____________________________________________________________________________ 40

Figure39.RelationshipBetweenCDRandCADTimelines ________________________________________________40Validationof[Received]Timestamps ________________________________________________________ 41StatisticsforReceivedTimestamps__________________________________________________________ 42

Figure40.Validated[Received]Timestamps11/1/2015through12/31/2015________________________________42SUITABILITYOFPERFORMANCETARGETS_____________________________________________________________ 43

P1Intervals_____________________________________________________________________________ 44Figure41.NENARecommendation__________________________________________________________________45Figure42.AverageBusyHourBasedonTelephoneTraffic _______________________________________________46Figure43.ComparisonofORCATandFITCHPass/FailDeterminationBasedonAnswerDelays___________________47Figure44.AnswerDelaysatCentralPSAPon11/07/2015________________________________________________48

P2/P3Intervals__________________________________________________________________________ 49Figure45:EMDP2/P3Statistics&Performance________________________________________________________51Figure46:n-EMDP2/P3Statistics&Performance______________________________________________________52Figure47.P2/P3PerformanceforEMDCalls–TargetversusActual________________________________________53

LawEnforcementP2/P3___________________________________________________________________ 53Figure48.LawP2/P3Statistics&Performance ________________________________________________________54

P3Interval _____________________________________________________________________________ 54Figure49.P3PerformanceStatistics_________________________________________________________________55

P4Interval _____________________________________________________________________________ 55Figure50:ComparisonofP4Averagesand90thPercentiles(dataforNov-Dec2015)__________________________56

MODELINGCURRENTDISPATCHOPERATIONS_____________________________________________________ 57

RATIONALE_________________________________________________________________________________ 57MODELSOFDISPATCHOPERATIONS________________________________________________________________ 57

APCORETAINS__________________________________________________________________________ 57ErlangAnalyses _________________________________________________________________________ 58

Figure51.QueuingTheoryTriangle__________________________________________________________________58WORKSTATIONS_____________________________________________________________________________ 58

Figure52.WorkflowsandWorkstationsintheBSODispatchSystem_______________________________________59WAITINGQUEUES____________________________________________________________________________ 59DEFINITIONOF“ERLANGS”______________________________________________________________________ 60QUANTITATIONOFWORKLOADS __________________________________________________________________ 60

Figure53.RecordfromSummationDatabaseofTelephoneTraffic_________________________________________61Figure54.RecordfromSummationDatabaseofCADandRadioTraffic _____________________________________62Figure55.RecordfromAveragedDatabaseofTelephoneTraffic __________________________________________63

BrowardCounty Pageiv ©Fitch&AssociatesAssessmentandRecommendationsofBrowardCounty’sRegionalE911 December2016

Figure56.RecordfromAveragedDatabaseofCADandRadioTraffic _______________________________________64ERLANGMODELING___________________________________________________________________________ 65

ErlangTables___________________________________________________________________________ 65CurrentBSOStaffing _____________________________________________________________________ 65ModelofCentralIntakeWorkstation ________________________________________________________ 65

Figure57.ErlangModelofCentralIntake_____________________________________________________________66Figure58.ErlangModelCentralIntakeAdjusted_______________________________________________________68

ModelofCentralFIREAssignmentWorkstation________________________________________________ 69Figure59.ErlangModelCentralFIREAssignmentWorkstationBSO________________________________________70Figure60.ErlangModelCentralFIREAssignmentsWorkstation___________________________________________71

DISPATCHOPERATIONSMODELS—OPTIONS_____________________________________________________ 73Figure61.FunctionsRequiredInAllEmergencyServicesDispatchOperations________________________________73

PERFORMANCETARGETS _______________________________________________________________________ 73PERFORMANCETARGETSFORMODELLING____________________________________________________________ 74APPLICATIONTODISPATCHOPERATIONALMODEL/OPTIONS _______________________________________________ 75

Figure62.ChangesinLatenciesfor+1.28σSurgesinAllHours-of-Day______________________________________76Figure63.DependenceofAnswerDelaysonSurgesandDispatchersOnDutyforCentralIntakeat1800Hours _____77

DISPATCHMODELOPTION0_____________________________________________________________________ 77Figure64.DispatchModelOption0 _________________________________________________________________78Figure65.North,CentralandSouthPerformanceandDispatcherHours-OnDutyforDispatchModelOption0______79

DISPATCHOPERATIONSMODEL/OPTION1___________________________________________________________ 80Figure66.DispatchOperationsModel/Option1 _______________________________________________________80Figure67.PerformanceandDispatcherHours-OnDutyinDispatchModelOption1.___________________________81

DISPATCHOPERATIONSMODEL/OPTION2___________________________________________________________ 82Figure68.DispatchModelOption2 _________________________________________________________________82Figure69.PerformanceandDispatcherHours-OnDutyinDispatchModelOption2.___________________________83

COMPARISONOFOPERATIONSMODELS/OPTIONS ______________________________________________________ 84Figure70.ComparisonofOperationsModels/Options_________________________________________________84

RECOMMENDATIONS_________________________________________________________________________ 85

TECHNOLOGY_______________________________________________________________________________ 85Findings—_____________________________________________________________________________ 85

OPERATIONALOVERSIGHTANDSYSTEMGOVERNANCE ___________________________________________________ 87Findings—_____________________________________________________________________________ 87

Figure71.IdentifyingandResolvingOperationalIssues__________________________________________________90PERFORMANCEMETRICS _______________________________________________________________________ 94

Findings—_____________________________________________________________________________ 94EFFICIENCYANDEFFECTIVENESS___________________________________________________________________ 97

Findings—_____________________________________________________________________________ 97

BrowardCounty Pagev ©Fitch&AssociatesAssessmentandRecommendationsofBrowardCounty’sRegionalE911 December2016AssessmentandRecommendationsofBrowardCounty’sRegionalE911 December2016

Attachments

A. SchedulingMatrixSampleB. PerformanceMeasuresC. CalculationofAnswerDelaysD. ErlangMathematics&AssumptionsE. QuantitationofWorkloadsF. SamplePhoneRecordG. SampleRecordsfromFireCAD,LAWCAD,andRadioStatisticsH. CalculationofSurgesI. ErlangTablesofWorkstationsJ. MonthlyPerformanceReportFormat


INTRODUCTION BrowardCountyretainedFITCH&Associates(FITCH)toconductanassessmentoftheCounty’sRegionalE911ConsolidatedCommunicationsSystem(RegionalE911System).Asanoverallgoal,FITCHistoinitiallyassesstheE911Systemthroughdatacollectionandbaselineassessments,externalbenchmarking,anddefinitionoffuturestateoptions.FITCHistoevaluatetheSystemagainstindustrybestpracticesandopineonthepertinenceandattainmentofpreviouslyestablishedgoals.TheRegionalE911Systemresultedfromtheconsolidationofeightsmallerpublicsafetyansweringpoints(PSAPs)afterextensivetechnicalreviewsandengagedpublicpolicydebates.TheSystem’sformalimplementationdatewasOctober2014.Sincestart-up,stakeholdershavemadeprogressinmeetinggoals;yet,therehavebeenconcernsabouttherelevantutilityofthecurrentperformancemetricsandtheSystem’sabilitytoquicklyachievealltheambitiousgoalsinitiallydefinedbythevariousstakeholders.Therewassignificantagreement/consensusintheearlystagesoftheconsolidationbasedonnumerousmeetingsandadoptionbyallpartiesofinterlocalagreements.However,itisfairtonotethatsuchconsensushasnowdissipatedinanumberofareas.TheCountycontractswiththeBrowardSheriff’sOffice(BSO)onaperformancebasistooperatetheRegionalE911Systemandprovidedispatchservices.BSOpersonnelreceiveanddispatchemergencyandnon-emergencycallsforpolice,fire,andemergencymedicalservices(EMS)withintheCounty,forallbuttwoofBrowardCounty’s31municipalities.BSOalsoprovidesteletype(queriesonly),whilethemunicipalitiesareresponsibleforanyservicesbeyondthatlevel.ThisreportrepresentsacompilationoffindingsfromthePhaseISystemassessmentandrecommendationsforthefuturestateoftheSystem-Phase2oftheproject.Phase1includedananalysesofqualitativeandquantitativedataasprovidedbytheCountyandotherstakeholders.Thisfinalphaseoftheproject/reportprovidesaseriesofspecificrecommendationsdesignedtoimproveoverallSystemeffectiveness,efficiencyandutilizationofindustrybestpractices.ThedocumentmergesbothPhasesintothisFinalProjectReport.AswithPhase1,FITCHconsultantsspentmanyhoursworkingwithBrowardCountyandBSOpersonnel.Wecontinuetobeimpressedwiththededicationoftheseindividualsandclearlyseethatallunderstandtheimportanceoftheirmissionandexpressadesiretoprovideexcellentservices.

BrowardCounty Page2 ©Fitch&AssociatesAssessmentandRecommendationsofBrowardCounty’sRegionalE911 December2016AssessmentandRecommendationsofBrowardCounty’sRegionalE911 December2016

EXECUTIVE SUMMARY TheExecutiveSummaryofthisreportbringsforwardthefindingsfromthePhase1reportandorganizesthemintofivegroupingsasfollows:

§ ObservationalFindings§ Technology§ GovernanceandOversight§ PerformanceMeasures§ EffectivenessandEfficiency

ObservationalFindingsarethosethatcommentontheRegionalSystemandrequirenofurtheraction.Theseareprimarilyobserved,positiveattributesoftheRegionalSystem.

Recommendations

TheotherfourcategoriesaboveaddressthefindingsfromthePhase1reportandprovideactionablerecommendations.Furtherdiscussion,intendedtoprovidecontextandfurtherexplanationfortherecommendations,shouldbereadindetailandareincludedinthereportsectiontitledRecommendations.Manyoftherecommendationswillrequiresignificantorganizationalandculturalchange.TheSystem,asawhole,willneedtorecognizethatchangemustbeplanned,andmoreimportantlyeachmajorchange,suchasimplementingEmergencyFireDispatch,mustbeexecutedtocompletionbeforeadditionalchanges,suchasEmergencyPoliceDispatch,areinitiated.Incertaincircumstances,lesssignificantchangeprocessescanbemanagedconcurrently.ThisiskeytotheSystem’sultimatesuccess.Throughoutthereport,recommendationstoincreaseresourcesincertainareasarebalancedbyrecommendationsintendedtoimprovetheoverallefficiencyoftheSystem.Intotal,thecurrentfundingofthesystemissufficient.Thereshouldbenoneedforfurtheroverallfunding.

Options Development

Thisreportevaluatesthreeoptionsforbothdispatchprocessconfigurationandthepersonnelneededtostaffworkstationsinordertoachieveadefinedperformancelevel.Theseoptions,andtheresourcesrequiredforthecurrentstaffingdeployedbyBSO,arereflectedbelow.

§ Option0—reflectsthecurrentdispatchconfiguration,butwithperformancetargetsadjustedtoachievenewrecommendedlevelsforcallintakeanddispatchpositions.

§ Option1—reflectsthestaffingrequiredforthenewperformancetargets,bututilizingasinglecallintakequeue.TheCountyrecentlymodifiedtheirsystemstonowutilizethesinglequeuefor911calls.

§ Option2—utilizesthenewperformancetargetsandsinglecallintakequeue,butfurtheroptimizesthefiredispatchoperationsbyconsolidatingthenumberofradiochannelsrequired.


Figure1belowreflectsBSO’scurrentstaffingasdescribedinBSOreportstitled“PSAPCallAnalysis”coveringtheperiodJanuary2015throughDecember2015.ThetablecomparesBSO’sreportedcurrentstaffingwiththehoursperdayrequiredforeachofthethreeOptions.Therequiredhoursnotedbelowforcurrentstaffingandeachofthethreeoptionsreflectthetotalhoursrequiredtobefunctioning‘inaseat’,bypositiontype,overasingle24-hourperiod.Figure1.ComparisonofCurrentandThreeOptions“Seats”Required

CurrentBSOModel FITCHOptions-Hours

All3PSAPs–"Seats"Required

StaticorVariablebyHour

CurrentStaffingHours

Option0

Option1

Option2

Intake 16.1 Variable 386 493 301 301

F/RDispatcher 15.0 Static 360 116 116 153

LawDispatcher 23.0 Static 552 235 235 235

Subtotal 54.1 1,298 844 652 689

PhoneSupport 3.0 Static 72 72 72 72

TTY 3.0 Static 72 72 72 72

Support 2.0 Static 48 48 48 48

Supervisor 7.0 Static 168 168 168 168

Subtotal 15.0 360 360 360 360

TOTAL 69.10 1,658 1,204 1,012 1,049

Reliefstaffingnotincludedabove1:

FireRelief 3LawRelief 4 IntakeRelief 5.3TheOptionsdefinetheneededpersonnelbyhour-of-dayandaddresssurgecapacityoftheRegionalSystem.IndevelopingtheOptions,thehoursneededinanypositiondonotincluderelieforstaffingmultipliers–onlytheactualnumberofseatsinthedispatchcenterthatmustbefilledatanyhouroftheday.2Itisunderstoodthat,asiscurrentpractice,BSOmustplanfor,andtheCountyfund,resourcestoallowfornecessarybreaks,trainingtimeandotheradministrativeneeds.Option2isthepreferredrecommendation.

1BrowardCountycurrentlyprovidesfundingtoallowforreliefstaffing.2Fitchuseshourstorepresentthehoursrequiredforpersonneltobeactivityoperatingattheirworkstationatagiventime,colloquiallyreferredtoas“buttsintheseat.”


OBSERVATIONAL FINDINGS BrowardSheriff’sOfficeisanAccreditedCenterofExcellenceasawardedbytheInternationalAcademiesofEmergencyDispatch(IAED).BSOutilizesemergencymedicaldispatcher(EMD)services–abestpracticefor911centers,andarequirementforongoingaccreditation.TheCountyhasimplementedasetofqualityassuranceandimprovementprocessesthatassistinobjectivelymovingtheSystemforward.TheCountyhasestablishedprocesscontrolandqualityimprovementprogramsthatincludemembersfromBSOandthecalltaker/dispatchunion.Thisallowsforeffortstoreviewandimplementservicequalityinitiatives.BSOisanaccreditedorganizationandutilizesqualityassuranceprocessesasprescribedbyIAED.Additionally,anumberofproceduresareinplacetomanagedispatchissues,reviewsituationsandimplementtrainingand/orpolicychangesforimprovement.Thenumberof911callersrequiredtobetransferredhasbeenessentiallyeliminatedundertheconsolidatedregionalsystem.Asaresult,totalcallprocessingtimeswerereducedbyapproximately30seconds.TheNationalFireProtectionAssociation(NFPA)setsabestpracticegoalof30secondstotransfercallersfromtheprimaryPublicSafetyAccessPoint(PSAP)toadifferentPSAP.Bycreatingaregionalsystem,thetransferstepiseliminatedandcallersonlyneedtoexplain/discusstheiremergencyonetime,notmultipletimesaswasthecase,historically.TheP1interval(thetimefromwhenthecallringsinthe911centeruntilthatcallisanswered),andP3interval(thetimefromwhentheassignmentdispatcherreceivesanemergencycallviathecomputer-aideddispatch(CAD)systemuntiltheyalerttheemergencyresponder),canbeaccuratelyevaluatedbasedoncurrentdataintheCADandtelephonysystem.BSOperformswellforthesedispatchintervals.Ontheotherhand,theP2interval(thetimefromwhena911callisanswereduntiltheinformationissenttoanassignmentdispatchertoalertfirstresponder)mustbecautiouslyevaluatedduetothetechnologyanddatalimitationsexistinginthecurrentCADand911systems.TheCountyhasindicatedthisissuewillberemediedwithimplementationofthenewCADinearly2017.


FINDINGS AND ACTIONABLE RECOMMENDATIONS

Technology

Findings —

TheCounty’sPSAPphonesystemandCADsystemarenoteffectivelylinkedtoallowcomprehensiveevaluationofSystemperformance.Formorethanhalfoftheincidentrecords,theeventintheCADcannotbelinkedtotheuniqueCallDetailRecord(CDR)thatinitiatedtheincident.Technologylimitationsresultedinonly25.6%ofCADrecordsconsideredvalidforuseinanalysisofP2/P3intervals.Countystaffisunabletodirectlyaccessphoneandradiosystemdata–therebylimitingtheirabilitytoanalyzesystemperformancebeyondthatpermittedbypre-designed/cannedreports,whichmakessomeoftherequiredreportingtediousanderrorprone.TheSystemutilizesemergencymedicaldispatching(EMD)software–abestpracticefor911centers.However,nosimilarprogramisutilizedforeitherfireorlawenforcementcalltypes.TheCADnetworkisredundantintheeventofafailure.However,itisnottestedonaregularbasis.Thisisasignificantdeficiencyandisinconflictwithbestpractices.

Recommendations —

TheCountyneedstoinsurethemissionsoftechnologydevelopmentandtechnologysustainmenthavedifferentfocusesandroles.Therefore,theCountyshouldprovideforaTechnologyDevelopmentTeamandaTechnologySustainmentTeamoverthenextfewyearsasnewtechnologiesareimplementedandthesystemcontinuestostabilize.AnabsolutepriorityfortheCountyistodevelopalinkbetween911phonerecordsandtheassociateCADincidentrecords.BSOshouldmaintainEMDcertificationtrainingforallcalltakersthroughtheInternationalAcademiesofEmergencyDispatch(IAED).CalltakerpersonnelshouldalsobetrainedandcertifiedasEmergencyFireDispatchers(EFD)andinthesystemimplementEFDinthenearfuture.Thesecertificationsareconsideredindustrybestpractice.Finally,lawenforcementagenciesshouldconsiderandevaluatetheefficacyofEmergencyPoliceDispatch(EPD)beingutilizedinthefuture.Thissystemisemergingasanindustrybestpractice.


Operational Oversight and System Governance

Findings —

BSO’soperationofthePublicSafetyAnsweringPointsischallengedwithsignificantmoraleproblemsembeddedinissuesofstaffing,trainingandmanagement.TheCountyhasinappropriatelymade,andpublicsafetyofficialsallowed,someoperationaldecisionstobehandledbytheCountythatshould,instead,bedeterminedbypublicsafetyofficials.Lowlevelsoftrustexistamongmajorstakeholders.Muchofthisisduetoroledefinitions.RelationshipsneedtoberedefinedinorderfortheSystemtomoveforwardeffectively.

Recommendations —

OperationalOversightandSystemGovernanceshouldberedefinedtostrengthentheroleofend-userswhilebalancingthelogisticalconcernsoftheOperator(BSO),andthefinancialandsystemgovernanceresponsibilitiesofBrowardCounty.Alternativeworkschedulesareavailableandshouldbeconsidered.AttachmentA,SchedulingMatrixSample,providessampleschedulesforconsideration.Fillingvacantpositionsinatimelymannerwiththegoalofmaintainingfullstaffingwillreduceexcessivemandatoryovertimeandtheassociatedstress.Thiswillallowresourcestoalignmorecloselytodemandpatterns,therebyimprovingefficiencyinthesystem.SupervisiononthePSAPdispatchfloorsshouldbeataratioofsixtooneasopposedtothecurrenttentooneratio.GreaterqualityassuranceprocessesaretobehandledbyBSOdispatchfloorsupervisors.Resourcesfordispatchertrainingshouldbeincreasedthroughreallocationofcurrentfunding.A“baselevelof911services”fundedbytheCountyshouldbemoreclearlydefinedbyutilizingthecurrentinterlocalagreementsandFITCH’smodelingofperformancelevelsasnotedincalltakingandradiopositions.Individualagenciesdesiringhigherlevelsofserviceshouldbeabletofundadditionalstaffinghoursortechnologyinordertoreceiveservicesspecifictotheirjurisdictionalneeds.TheRegionalSystem’smanagementandtechnologyshouldfacilitatetheseadditionalservicesaslongastheydonotdisruptthebaseservices.

Performance Measures

Findings —

Certainperformancemeasureshavebeenmisinterpreted,incorrectlyapplied,orareinconsistentwithcurrentindustrybestpractices.


TheCounty’suseofPASS/FAILtargetsprovideslittleinthewayofinformationforcontinuousqualityandperformanceimprovement.ThefailureofthecurrentPASS/FAILorYES/NO,P1busyhourtarget,isthatitprovidesnoguidanceastothelevelofsurgecapacitythatisfiscallyresponsibletobuildintothesystem.

Recommendations —

TheCountyshouldmodifythecurrentmonthlyperformancereportformatandreplaceitwithamonthlyreportthatfocusessolelyondataandprovidesnocommentary.The“busyhour”istoberedefinedinaprospectivemannerbasedonhistoricaldataandistobereassessedinnolessthan12-monthintervals.Thesechangesallowformeaningfulandactionableinformationexchangesandprovideuseragencieswithaneededlevelofoversight.TheCountyshouldpurchaseaperformancemeasurementsoftwarepackagethatwillprovideagencieswithreadyaccesstotheactivitiesandperformanceoftheirrespectivefieldunits,andsimultaneouslyallowtheCountyandBSOtoevaluatesystemperformanceatthemicroandmacrolevels.Onlytheperformanceonemergency/911incidentsshouldbeincludedintheperformancereports.Thecurrentpracticeofevaluatingduplicate911callsonasingleincidentskewsmeasurement.Thetruestructureofthereportshouldbetopresentthenumbersinawaythathighlightsthecallswhereresponsetimeisimportant.Somethoughtshouldbegiventopresentresponsetimesstartingwiththecallreceipttoemergencyservicearrivalonscene.Thiswillgivetheproperpresentationofthecaller’sexperience.Specialattentionshouldbepaidtohighpriorityincidents.Regardingreportingperformanceforvariouscallprocessingtimeintervals,oncethetechnologyissuesareresolved,theP2andP3intervalsshouldbereportedseparatelyandasacombinedmetric.Thereasoningisthat,particularlyforfireandemergencymedicalDeltaandEcholife-threateningcalls,fastandeffectivedispatchperformancecontributestopositiveoutcomes.MonthlyreportsshouldalsoreportP4(turn-outtimes)forfirerescueincidentsandP5(traveltime)forbothfirerescueandlawhighpriorityincidents.Ingeneral,dispatchcenterperformancemetricsaretofocusonoptimizingdispatchprocessesasmuchaspossible,withtheendresultbeingtogethelpmovingtoemergenciesasquicklyaspossible.Theprimaryobjectiveistocontributetothepotentialforpositiveoutcomesforpatientsandproperties.

Effectiveness and Efficiency

Findings —

CurrentPSAPs,trainingfacilityand“fleeto”planshavefacilitylimitations,especiallyrelatedtoadequatespace.


Theconsolidatedsystemiscapableofclosestunitresponsetolife-threateningemergencies,butprotocolsarenotyetinplacetoimplementthiscapability.Radiotrafficutilization,bybothfire/EMSandlawenforcementunits,iscomparativelyhigh.MDTs(mobiledataterminals)andMCDs(mobilecomputingdevice)arenoteffectivelyutilizedtoreduceradiotraffic.BSOcurrentperformanceindicatesoverstaffingincalltakerpositionsbasedonErlangmodeling.BSOcurrentperformanceindicatesoverstaffinginFireAssignmentpositionsbasedonErlangmodeling.

Recommendations —

CallprocessingstaffingshouldbeadjustedtoachieveP1/call-takingperformanceofbetweenthreetofivesecondsatthe90thpercentilebyadoptingtherecommendedworkstationfunctionalreorganizationasdetailedinthereportsectiontitled,DispatchOperationsModels–Options.Thisadjustment,inconjunctionwiththealreadyimplementedsinglequeforcallintake,providessignificantefficienciesinthecalltakingprocesswhilemaintaininghighlevelsofperformance.Fire-rescueagenciesshoulddevelop,approveandimplementcountywidenearestunitresponseprotocolsthatapplyirrespectiveofjurisdictionalboundariesinthoseincidentsinvolvinghighpriorityincidents(e.g.Delta&EcholevelEMDcalls).RecommendedprocesschangestoradiochannelusageincluderequiringincreasedusageofMobileDataTerminals(MDTs)byfieldresponders.OncetheCADisupgradedtoallowautomaticcomputerassignment/recommendationofresponseunitsforfire/rescuecalls,asingle“gatekeeper”function/firerescuealertchannelcanbeimplementedtomanuallyapprovetheassignmentconsistentwithOption2.Upondispatch,pre-definedtacticalradiochannelswouldbeusedformoreroutinefireincidentsandEMSincidents.Moresignificantincidents(structurefires,major/multipleunitresponses)wouldbeassignedadedicatedtacticalchannel.Thischangeinfirerescueradiooperationsprovidessignificantefficiencieswhilemaintaininghighlevelsofperformance.LawenforcementradiopositionsshouldbeconsolidatedtoincreaseefficiencyconsistentwithOption2.Long-termcapitalbudgetingprogramsshouldbeconsideredassoonaspracticaltoincludetwonewpurpose-specific911facilities.


METHODOLOGY TheSystemassessmenteffortderiveditsfindingsfromtwoperspectives.First,istheinputreceivedfromstakeholders,especiallyLevel1(elected,appointedandseniormanagementofficials)andLevel2(directors,managersandsupervisorypersonnel).InformationwasalsogleanedfromtheconsiderabletimeFITCHconsultantsspentdirectlyobservingoperationsinallthreeregional911facilities,inthefieldandfromdirectsurveysofdispatchpersonnel.ThesecondperspectiveisbasedonextensiveandsophisticatedanalysesofrawdataprovidedtoFITCHconsultants.Thedataincluded911centerphonerecords,computer-aideddispatch(CAD)recordsandradiosystemrecords.Fromthisinformation,FITCHwasabletoassesstheRegionalE911System’scurrentlevelofperformanceandmodelthatperformance.Todeterminestaffingneeds,BSOandtheCountycurrentlyutilizesastaffingestimatorandretentionratecalculatorknownasRETAINS,aproductoftheAssociationofPublic-SafetyCommunicationsOfficials(APCO).TheRETAINStitlestandsforResponsiveEffortstoAssureIntegralNeedsinStaffing.Theestimatorisrespectedasatoolforestimatingstaffingneedsandincludessomelevelofcomplexity.However,itsapplicationforBroward’sRegionalE911SystemissignificantlylimitedduetoBroward’sfluctuationsincallvolumeonanhour-by-hourbasisandthechangesinstaffingusedtomeetthosedemands.AneasilyoverlookedlimitationoftheRETAINSestimatoristhatitdoesnotincludespecificperformancetargetsaspartofthestaffinglevelcalculations.ThisisveryrelevantastheCountyandBSOoperateunderaperformancebasedcontract.Ofnote,therewereanumberofdatadeficienciesthatlimitFITCH’sabilitytocompletespecificprojectscopepoints.Forexample,whileCADdataforallofcalendaryear2015wasavailable,onlythreemonthsofphonerecordswereavailableduetoasystemupgrade.Fromthesetwodatasets,therewereonlytwomonthsofoverlapbetweenthephonerecordsandtheCADdata.Radiosysteminformationalsohadlimitationsthathampereddetailedsystemperformanceanalysis.Nonetheless,FITCHwasabletoconstructdetailedmodelsandwasabletodrawmeaningfulconclusions.Afullaccountingofdataissuesisdescribedindetailunderthereportsectiontitled,SourcesofData.OncetheSystemassessmentandmodelingofBSO’scurrentoperationswascomplete,FITCHconsultantsthenquantifiedtheoptimalnumberofcalltakersanddispatchersneededtomeetcertainperformancecriteriaintheRegionalE911System.Threeoptionsthatindicatestaffinglevelsandrecommendedperformancegoalsarepresentedforconsideration.Ultimately,thequantifyingofpersonnelisaparticipatoryprocessinvolvingstakeholderstosetnewparameters.Thisprocessiskeyindesigningadispatchcenterthatisbasedonnationalbestpracticesandlocalcompetencies.


DISPATCH CENTER BEST PRACTICES AccreditationbytheInternationalAcademiesofEmergencyDispatch(IAED)isthegoldstandardforemergencydispatchcentersandpublicsafetyagencies.AchievingandmaintainingstatusasanAccreditedCenterofExcellence(ACE)requirestop-notchsystemsforreportingandreviewingprocesses,andultimatelybenefitpatientsandthecommunity-at-large.Thegoalofaccreditationistoimprovepatientcareandclinicaloutcomes.IAEDprovidesthefollowingseparateaccreditationprocessesfordispatchpersonnel:

§ EmergencyPoliceDispatchCertification(EPD)§ EmergencyFireDispatchCertification(EFD)§ EmergencyMedicalDispatchCertification(EMD)

Eachcertificationareaprovidesstructuredcallprocessingfortherespectivediscipline.IAEDsetsout20pointsasaccreditationrequirements.Figure2belowarticulatesthe20IAEDpointsofexcellencethatmustbeformallydocumented,describedandverifiedaspartofthemedicaldispatchaccreditation/re-accreditationapplicationprocess.


Figure2.RequirementsforIAEDMedicalDispatchCenterAccreditation3

Formallydescribeanddocumentthefollowing–1) Communicationcenteroverviewanddescription2) MedicalPriorityDispatchSystemTM(MPDS)versionandlicensingconfirmation3) CurrentAcademyEMDcertificationofallEMDpersonnelauthorizedtoprocessemergencycalls4) AllEMDcertificationcoursesareconductedbyAcademy-certifiedinstructors,andallcasereviewisconductedby

Academy-certifiedED-Qs5) FullactivityofQualityImprovement(QI)committeeprocesses.6) IAEDqualityassuranceandimprovementmethodology.7) ConsistentcaseevaluationthatmeetsorexceedstheAcademy’sminimumexpectations8) HistoricalbaselineQAdatafrominitialimplementationofstructuredAcademyQAprocesses(firstQISummaryReport,

ifavailable*)9) Monthlyaveragecaseevaluationcompliancelevelsforthecommunicationcenterforthesixmonthsprecedingthe

accreditationapplication,withcompliancelevelsatoraboveaccreditationlevelsforatleastthethreemonthsimmediatelyprecedingapplication

10) VerificationofcorrectcaseevaluationandQItechniques,validatedthroughindependentAcademyreview11) Implementationand/ormaintenanceofMPDSorientationandcasefeedbackmethodologyforallleadpersonnel12) VerificationoflocalpoliciesandproceduresforimplementationandmaintenanceoftheMPDS.Includeallpolicies

relatingtoEMDpractices13) Copiesofalldocumentspertainingtoyourcontinuingdispatcheducation(CDE)program14) SecondaryEmergencyNotificationofDispatch(SEND)orientation15) EstablishedlocalresponseassignmentsforeachMPDSDeterminantCode16) MaintenanceandmodificationprocessesforlocalresponseassignmentstoMPDSDeterminantCodes17) Thecommunicationcenter’sincidence(numberofoccurrences)ofallMPDScodesandlevelsforthesixmonths

immediatelyprecedingapplication18) AppointmentandappropriateinvolvementoftheMedicalDirectortoprovideoversightofthecenter’sEMDactivities19) Agreementtosharenon-confidentialEMDdatawiththeAcademyandothersfortheimprovementoftheMPDSand

theenhancementofEMDingeneral20) AgreementtoabidebytheAcademy’sCodeofEthics,CodeofConduct,andthestandardssetforthforanAccredited

CenterofExcellenceBrowardstakeholdersshouldappreciatethatAccreditationguaranteesthatalltheprocessesneededforhighqualitypatientcareareimplemented.HowpromptlytheyarecarriedoutisacomponentofperformanceindependentofAccreditation.TheIAED-ACEaccreditationrequirementscontainnotimemetrics.RequirementsforACEAccreditationarecomprehensiveandreflecttheeffortrequiredtoachieveandmaintainaccreditation.Evenforthebestdispatchcenters,accreditationistypicallyamulti-yearprocess.

3https://accreditation.emergencydispatch.org/resources/General/MEDICAL%20Accred-Re-Accred.pdf,June2016


TheBrowardSheriff’sOfficefirstaccomplishedaccreditationin2003.BSOmaintainedaccreditationandwasreaccreditedforthethree-yearperiod2015to2018.Ofnote,BSOusesonlythemedicaldispatchprotocolandisonlyaccreditedformedicaldispatch.

FINDING:BrowardSheriff’sOfficeisanAccreditedCenterofExcellenceasawarded

bytheInternationalAcademiesofEmergencyDispatch.

TheBrowardSheriff’sOfficehasalsorecentlybeenreaccreditedfortheircommunicationsservicesbytheCommissiononAccreditationforLawEnforcementAgencies(CALEA).Otherattributesofhighperformancedispatchcentersincludedailymeetingsofdispatchstafftoreviewthepriorday’sevents,refinedeploymentandreviewanyoperationalconcerns;regularsurveysbyemergencyprovideragenciestoincludequestionsregardingthedispatchprocess;continuousfeedbackloopsforimprovementthroughouttheorganization;andclinicaloversightregardingemergencymedicaldispatchingbyafull-timemedicaldirector,whohasdirectinvolvementwiththecenter’sperformanceandpersonnel.


CURRENT ORGANIZATION AND ENVIRONMENT AmeaningfulanalysisofthecurrentSystemrequiresanappreciationoftherecenthistoricalandcurrentorganizations,andtheirenvironment.Thefollowingsectionshighlightdemographictrendsimpactingdemandsforservice,existingrelationshipsamongstakeholdersandtechnologyimpactingSystemperformance.

County Demographics

Itimportanttounderstandtheutilizationofemergencyservicesfromahistoricalperspective.Firerescuedepartmentshaveseenasignificantincreaseinemergencyactivity.Whilereportedstructurefiresaredowndramatically,inthelastdecadealonethereisbeena40%increaseinoveralltotalemergencycallsbasedprimarilyonEMSandactivatedfirealarms.4Therefore,thefollowingdemographicinformationprovidesacontexttounderstandsomeofthedriversofsystemdemand.

Current and Historical

Today,BrowardCountyisamostlydeveloped,1,200+squaremileurbanareawithonly10.5squaremilesleftofdevelopableland.AccordingtotheUniversityofFlorida’sBureauofEconomicandBusinessResearch(BEBR),theCounty’stotalpopulationisestimatedat1,827,367asofApril1,2015.5Ofthe31municipalitiesinBrowardCounty,thethreelargestcitiesaretheCityofFortLauderdalewitharesidentpopulationof178,590,PembrokePines,166,611,andHollywood,149,728(July1,2015,USCensusdata).BrowardCounty’shistoricgrowthpeakedintheyear2000withanaverageannualgrowthof2.72%.Between2000and2005,averageannualgrowthhadslowedto1.44%,resultinginaresidentpopulationof1,739,487persons.Growthbegantoslowdueinparttosky-rocketinghousingcosts,followedbythe2008economicslump.In-migrationofresidentstypicallyfueledtheCounty’srapidpopulationgrowth.However,“excessivelyhighhousingcostsfollowedbydiminishingjobopportunities,reducedin-migrationandpopulationgrowthtoitssmallestlevelinsixtyyears.”6Nevertheless,thecontrastofaddedpopulationbetween2005and2010andthatexperiencedbetween2010and2015,issignificant.Figure3belowrepresentsthepopulationgrowthinfive-yearincrementsfor2005to2015.

4Ahrens,M.(2016).TrendsandPatternsofU.S.FireLoss.N.F.P.Association,NationalFireProtectionAssociation.5ProjectionsofFloridaPopulationbyCounty,2020-2045,withEstimatesfor2015,FloridaPopulationStudies,Vol.49,Bulletin174,January2016.UniversityofFlorida,BureauofEconomicandBusinessResearch.6Broward-by-the-Numbers,Number57,page1,July2009.BrowardCountyPlanningandRedevelopmentDivision,accessedJune2016.


Figure3.BrowardCountyPopulationGrowth,2005to2015

Atotalof32,573residentswasaddedtoBroward’spopulationbetween2005and2010,but104,201residentswereaddedtothepopulationbetween2010and2015,representing5.9%growthforthatperiod.7

Projected Growth to 2020

TheUniversityofFlorida’sBureauofEconomicandBusinessResearch(BEBR),providesannualpopulationforecastsforthestateandforallFloridacounties.BEBR’sprojectionsofoverallpopulationgrowthinBrowardCountyisexpectedtoslowintheoutyearsto2020.BEBR’sJanuary2016populationprojectionsforBrowardCountyareprovidedas“low”,“medium”and“high”.Themediumprojectionsarethoughttogenerallyprovidethemostaccurateforecastsoffuturepopulationchange.BEBRforecastsBroward’smediumpopulationat1,914,500asofApril1,2020,whichrepresentsa2.04%increaseover2015.TheUSCensus’AmericanCommunitySurveyfor2007-2011,notesthatBrowardCountyisanetexporterofworkersinthedaytimeduringtheworkweek.WhilethecitiesofFortLauderdaleandPompanoBeachexperienceasignificantnetincreaseintheirdaytimepopulations,suburbanareastendtolosepopulationinthedaytimeduetomanyworkerscommutingoutofthearea.Downtownareasgenerallyseeasignificantincreaseindaytimepopulation.ThegreatestdaytimegainsareseeninthemunicipalitiesintheeasternpartoftheCounty,suchasFortLauderdaleandPompanoBeach.8

7PopulationdatawasderivedfromtheBrowardCountysourcenotedinthepreviousFootnoteandwasusedinsteadofUSCensusdataasitismorecomplete.CensusdataandBrowardCounty’sestimatesandprojectionsarerelativelysimilaranddonotrepresentasignificantdisparity.8Broward-by-the-Numbers,Number60,page1,March2013.BrowardCountyPlanningandRedevelopmentDivision,accessedJune2016.

1,739,4871,772,060

1,876,261

1,650,000

1,700,000

1,750,000

1,800,000

1,850,000

1,900,000

2005 2010 2015

BrowardCounty- TotalPopulation


TheBEBRpopulationforecastsincludedatabyagegroups.Ofparticularinterestistheagecohortof70+years,whichsignificantlyimpactstheneedforhealthcareservicesand,inparticular,emergencymedicalservices.BEBR’sdataregardingagecohortsof70+years,providesestimatesfor2012andprojectionsfor2015and2020.Theinformationispresentedbelow.Thistrendingdemographicwillhaveaconcurrentimpacton911servicesaswell.Figure4.PopulationProjectionsfor70+YearsAgeCohorts

Theprojectednumbersincreaseovereachfive-yearperiodandforeachagegroup,exceptfortheagecohortof80to84years.Overall,thenumberofBrowardresidentsovertheageof70years,isexpectedtoincreasebyapproximately41,700individualsor15.4%asestimatedbetween2012andprojectedfor2020.Thefigurebelowrepresentsthegrowthfortheentireagegroupof70+years.

Figure5.PopulationProjectionsforResidentsAge70+Years

010,00020,00030,00040,00050,00060,00070,00080,00090,000

2012 2015 2020

PopulationProjections:ResidentsAge70+Years

70to74 75to79 80to84 85+

184,061 195,592225,755

0

50,000

100,000

150,000

200,000

250,000

2012 2015 2020

ProjectedPopulationGrowth70+AgeGroup


Intuitivelythereexiststhesensethatasthesizeoftheoldercohortincreases,thenumberofagerelatedemergencyeventswillalsoincrease.Theincreasednumberofpeopleinthe70+agegroup,inparticular,isexpectedtodrivedemandforemergencymedicalservices.Thecriticalquestionis,byhowmuch?Fourstudiesprovideinsightintotheimpactofsuchademographictrend.First,theDepartmentofEmergencyMedicine,UniversityofNorthCarolinaatChapelHill,ChapelHill,NorthCarolina,conductedaretrospectivestudyof2.7millionEMStransportstoemergencydepartmentsacrossNorthCarolinain2007.Amajorfindingofthisstudywasthatindividuals65yearsofageorolderaccountedfor38%ofallEMStransportstoNorthCarolinaemergencydepartments.9AsecondstudysupportedbyFlorida’sPinellasCountyMentalHealthandSubstanceAbuseTaskForce,withcooperationofthePinellasCountyDataCollaborative,evaluatedtheagedistributionofemergencymedicaltransportsinPinellasCounty,FL,fromJuly1999throughJune2000.StatisticsfromthesummermonthsinPinellasCountyareequallyrelevanttoBrowardCounty.Duringthisseason,thestatisticsreflecttheeffectsofthestablydomiciled,localpopulation.Distortionsduetotheinfluxofwinter“snowbirds”areabsent.AccordingtotheUnitedStates2000Census,PinellasCountyhad22%ofitsdomiciledpopulationinthe65+cohort.Duringthesummermonths,whenthereisnopopulationdistortionduetosnowbirds,atleast50%ofallemergencymedicaltransportsinvolvedthe65+cohort.InPinellasCounty,theonefifthofthedomiciledpopulationinthe65+cohortaccountedforonehalfofallemergencymedicaltransports.SimilarobservationsregardingageandemergencymedicaltransportsweremadeinsmallerandearlierstudiesinForsythCounty,NorthCarolinain1995,andinDallas,Texasin1990.10FITCHbelievesthatthedemandforemergencymedicalservicesinBrowardCounty,likeitsFloridawestcoastneighbor,PinellasCounty,willbedrivendisproportionatelybythe65and70+yearoldcohorts.

Stakeholder Relationships

Participants

Ofthe31municipalitiesinBrowardCounty,allbuttwo,CoralSpringsandPlantation,areparticipantsintheRegionalE911System.TheSystemistheresultofa2002CharteramendmentthatcalledforcoordinationbetweentheCountyandmunicipalitiestoestablishacountywidecommunicationsinfrastructureforfireandemergencymedicalservices.Aprimaryoutcomeofconsolidationwasto

9TFPlatt-Mills,BLeacock,JGCabañas,FSShofer,SAMcLean,PrehospitalEmergencyCare,2010Jul-Sep;14(3):329-333.doi:10.3109/10903127.2010.481759.“Emergencymedicalservicesusebytheelderly:analysisofastatewidedatabase.”http://www.ncbi.nlm.nih.gov/pubmed/20507220.10JLWofford,WPMorgan,MDHeuser,ESchwartz,RVelez,MBMittelmark,AmJEmergMed,1995May,13(3):297-300.“Emergencymedicaltransportoftheelderly:apopulation-basedstudy”andCEMcConnel,RWWilson,SocSciMed,1998Apr,46(8):1027-1031.“Thedemandforprehospitalemergencyservicesinanagingsociety”.


enableclosestunitresponsestolife-threateningemergenciesandprovidesupportforregionalspecialtyteams.11Thecurrentsetofstakeholderscanbemorereadilyidentifiedasfollows:

§ BrowardCounty,withlegislativeandfinancialresponsibilitiesfortheSystem,§ BSOasthecontractedOperatoroftheSystem,supplyingpersonnelanddirectmanagementof

thethreepublicsafetyaccesspoints(PSAPs)locatedthroughouttheCounty,and§ MunicipalfirerescueandlawenforcementagenciesasendusersoftheRegionalE911System’s

services,andBrowardSheriff’sOffice(BSO)asanenduserpublicsafetyagency.

Aswillbehighlightedfromstakeholderinput,relationsamongthethreemajorstakeholdersarenotoptimal.ThiswasemphasizedbyrecentfindingsfromfacilitatorsworkingwithCountyandBSOstaff.Theyconcludedthat the designofoneteamreportingerrorsontheotherteam’sworkdoesnotsupportacollaborativerelationshipbetweentheCountyandBSO.Theyindicatedthatthetwoteamsarenotpositionedtobecollaborativeinreachingthesamegoalandwilllikelycausemoreexpendedenergyandtimeindefendingtheirrespectiveperspectives.Thefacilitatorsrecommendedfocusingontheredesignoftheexistingworkingmodeltosupportacollaborativeworkingteam.

FINDING:Lowlevelsoftrustexistamongmajorstakeholders.Muchofthisisdueto

roledefinitions.RelationshipsneedtoberedefinedinorderfortheSystemtomove

forwardeffectively.

Itisnoteworthythatsincetheconsolidationeffortbegan,currentstakeholdershaveengagedinasustaineddiscourseon911services–somethingthatwasrarelydiscussedbefore.TheoutcomeofthisdialogisthatthenewSystem,withgreatlyincreasedscrutiny,isnowidentifyingandaddressinglong-standingissues.Itislikelytheseissuesexistedbefore,butindividualPSAPsdidnothavethetransparencythatexistsnow.FITCHnotedthatattemptstoobtainspecifichistoricalperformancedatafromvariouscommunitieswasunsuccessfuleitherbecauseofaninabilityorunwillingnesstoprovidesuchinformation.Fromoneperspective,thetensionthatexistsnowcanbeseenasanoutcomeofthetransparencyandprogressthatisnowbenefitingthepublicandfirstresponders.

Technology Review

Thetechnologyassessmentwasaccomplishedthroughdiscussionswithtechnicalsupportpersonnelanddirectobservationson-siteatthedispatchconsoles.Theassessmentfocusesontelecommunications,thecomputeraideddispatch(CAD)systemandradiooperationstechnologies.Belowarethekeyissuesobservedintheinitialassessment.

11BrowardSheriff’sOfficeRegionalAgreementsaccessedthroughBroward.org,RegionalCommunicationsandTechnology,BrowardCountyRegional911andBrowardCountyCharter,RevisedNovember4,2008,ArticleV.PublicSafety,Section5.03(A).


Telecommunications

TheRegionalE911SystemcurrentlyoperatesonanIntradoPower911telephonesystem(version5.5),witharedundantnetwork.AutomaticCallDistribution(ACD)andprerecordedansweringisinuseindependentlyateachfacilityfor911calls,whilenon-emergencycallsaredistributedacrossallthreePSAPs.FITCHconsultantswereadvisedthatasinglequeueforhasbeenimplementedwhichnetworksallthreeCountyPSAPfacilities.AnAutomaticCallDistribution(ACD)systemhasbeenimplementedacrosstheentirenetworkfor911calls,whichwillsignificantlyimprovetheefficiencyofcallhandlingandavoidunnecessarydelays.ThedesignoftheACDNetworkEnhancementincludeslocalpreferencecallhandlingbeforeroutingcallstotheregionalqueue.PartofthedatarequiredtodescribethetotaltimelineofasingleincidentresidesinthephonesystemandtheotherpartresidesintheCAD.AmajorissueidentifiedbyFITCHisthefailureofthephoneandCADsystemstoeffectivelylinkrecordsassociatedwithasingleincident.TheCountyrecentlyacknowledgedthelackofthislinkageasanissueofconcern,andhasindicatedtheyarecurrentlyundertakingeffortstoeffectivelyaddressthisissue.Afteraconcertedeffortwiththedataprovided,FITCHwasabletolinkincidents,butforfewerthan50%oftheincidentrecords.ThistechnologydeficitsignificantlylimitstheabilitytocalculatetheP2/P3callprocessingintervals.MostimportantisthattheSystemcannotreliablyanswerthefundamentalquestionofhowlongittakesbetweenwhenacallismadeto911andwhenhelparrives.

FINDING:County’sPSAPphonesystemandcomputer-aideddispatch(CAD)systems

arenoteffectivelylinkedtoallowcomprehensiveevaluationofSystem

performance.

Duringdatacollection,therewerechallengesinobtainingdirectaccesstothephonesystemandtheradiosystemdatatables.Countystaffreportedtheyprocesstheirreportsthroughastandardizedreportinginterface,andlackdirectaccesstophonesystemdata.Countystaffdidadvisefundingisavailabletopurchasethenecessarysoftwaretoallowdirectaccesstocriticalsystemdata.

FINDING:Countystaffisunabletodirectlyaccessphoneandradiosystemdata–

therebylimitingtheirabilitytoanalyzesystemperformancebeyondthatpermitted

bypre-designedreports(a‘canned’reportingsystem)whichmakessomeofthe

requiredreportingtediousanderrorprone.


Computer Aided Dispatch System

ThecurrentMotorolaPrintTracCADsystem,originallydeployedin1994,serveseachofthethreePSAPfacilities.Forsomeagencies,suchastheCityofFt.Lauderdale,thisCADisbelievedtobeabackwardsstepintechnology.TheCountyhasacknowledgedtheageoftheircurrentsystem,andsomehistoricalproblemswithnetworkstability.Forthesereasons,consultantswereadvisedthatthereareplanstoupgradetoaMotorolaNextGenCADinthenearfuture,currentlyreportedasearlyas2017.TheCounty,BSOandendusersarecollaboratingtoidentifyimprovementsinthenewCADinordertoimprovetheoverallSystem.Atpresent,theCountyisriskaversetoroutinelytestingtheredundantnetworkdesignbecausetransferringsystemprocessingtotheCADdisasterrecoverysystemrequiresmanualintervention,andcantakeuptofourhourstocomplete.Asnotedabove,thecurrentCADdoesnothaveaneffectivemethodtoassociaterecordsfromthephonesystemtotheappropriateCADrecord.

FINDING:TheCADnetworkisredundantintheeventofafailure.However,itisnot

testedonaregularbasis.Thisisacurrentdeficiencyandisinconflictwithbest

practices.

For911personneltoeffectivelydispatchemergencyresponders,twoessentialpiecesofinformationarerequired–whereistheemergency,andwhatistheemergency.DirectobservationsandanalysisofCADdatareflectthatthecurrentabilitytoobtainanaccurateincidentlocationishamperedbyanumberofissues.Operatorsstruggletoquicklyobtainandvalidatethecaller’s/incidentlocation.Thisproblemwasidentifiedpriortothisstudyandanumberofmitigatingstrategieshavealreadybeendeployed,mostlyrelatedtocalltakertraining.Inparticular,analysisbytheCountyandBSOnotethatcalltakerswho‘deviate’fromrecommendedprocesses,especiallyinmedicalcalls,takelongertoprocessthecalleffectively.911personnelreported,andFITCHpersonnelobserved,inconsistentperformanceofmappingtechnologythatdecreasedthecapacitytoquicklylocate911callers.Thereareanumberoftechnologysolutionsthatwillhelpimproveaddressing,andthereforeoverallcallprocessingtimes.BrowardRegional911SystemdispatchersarecertifiedasEmergencyMedicalDispatchers(EMD)andassuchprovidepre-arrivalinstructionstocallersinneed.Aspartofthatprocess,BSOmaintainsaqualityassurance(QA)programthatincludesspecializedQApositionsandPriorityDispatch’sAQUAsoftwarethatmeasures,analyzesanddocumentscallprocesses.Thesoftwareassistsinpinpointingtrainingneedsanddocumentscontinuousimprovementefforts.TheQAprogramshouldmeetcriteriaidentifiedinDispatchCenterAccreditationRequirementsnotedintheDispatchCenterBestPracticesreportsection.CompliancewithcertainoftheserecommendedstandardsarereportedbyBSOtotheCountyforinclusioninmonthlyreports.TherealsoexistoptionstohavethisQAreviewdonebyexternalpartiesto


ensureobjectivity.WhiletheuseofEMDisabestpractice,theuseofsimilarfireandlawenforcementsystemsarenotbeingutilizedwithintheBrowardsystem.

FINDING:TheSystemutilizesemergencymedicaldispatching(EMD)services–a

bestpracticefor911centers.However,nosimilarprogramisutilizedforeitherfire

orlawenforcementcalltypes.

Radio Operations

Differentfire,lawenforcementandEMSagenciesworkoffseparateassignmentandtacticalchannels,oftenrequiringmultipledispatchersforthesameemergencyincident.Asnotedelsewhereinthisreport,thereisahighlevelofradiousageforverbalcommunicationsbetweenfieldpersonnelandradiooperators.Thisraisesquestionsregardingtheutilizationofmobiledataterminals(MDTs)andtheefficiencyofthecurrentoperations.Dispatchstaffingcanbeutilizedmoreefficientlyiffieldagenciesagreetoutilizeoneassignmentand/ortacticalchannel.Regardingfireradiooperations,therearemultiplefiredispatchchannelsoperatingindependentlyofoneanother.Whilelikelyaremnantofpre-consolidation’sindependent911centers,thisisnotthemostefficientoreffectivewaytohandleradiooperations.Manylargersystemslimitthenumberofassignmentradiochannels,andthenquicklymoveunitsofftoanoperatingortacticalchannelbasedonthetypeofincident.Fire/EMSapparatushavemobiledatacomputers(MDCs)withaircardsinstalledintheunits.ItappearsthatrespondersdonotusetheMDCsasameanstoupdateunitstatuschangesorcommunicateroutineinformation.Thisinformationexchangeisbestexecutedviathemobiledatacomputers.UsingMDCscanreduceerrors,isamoreefficientmethodtocommunicate,andcanfreeupradiochannelsformorecriticalcommunications.

FINDING:Radiotrafficutilization,bybothfire/EMSandlawenforcementunits,is

comparativelyhigh.MDTsandMDCsarenoteffectivelyutilizedtoreduceradio

traffic.

Dispatch Facilities

FITCHconsultantsspentsignificanttimeinthethreePSAPS,North,CentralandSouth.WhileCentralhasthelargestfootprintoffloorspace,NorthandSouthdispatchfacilitiesmustcopewiththelimitedavailablesquarefootageattheirlocations.ItshouldbenotedthatstakeholdersundertookasignificantevaluationofpotentialdispatchsitespriortoselectingthecurrentPSAPlocations.Thisincluded


evaluatingcharacteristicssuchashurricaneratings,back-uppowergeneratorsandredundantpowerfeeds.WhileitisrecommendedthatasystemthesizeandcomplexityofBroward’shaveatleasttwogeographicallydisparatesites,stakeholderswererequiredtoselectexistingfacilitiesthatcouldbemodifiedtominimallyachievetheexistingneeds.

TheSouthdispatchcenterisnotapurposebuilt-facilitydesignedforhighvolumedispatchoperations.Thebuildingisasharedfacilitycombiningafirestationanddispatchcenteroperations.Currentdispatchcenterstructuralchallengesincludefluctuatingheating,ventilation,andairconditioning,inadequatetrainingroomsizeanddesign,andlimitedrestroomfacilitiesandquietrooms.Southdispatchcenterisdesignatedasoneofthe“fleeto”orbackupcommunicationfacilitiesintheeventaplannedorspontaneousevacuationoccursatoneoftheothertwocenters.Southdispatchisnotdesignedforsustainedlongtermdispatchoperationsasa“fleeto”center.Theformer911centerintheBrowardSheriff’sheadquartersbuildingonW.BrowardBoulevardwaseliminatedasthe“fleeto”siteinlargepartbecausethestructurehasalowerhurricaneprotectionrating.Notingrecentincidentswhereevacuationofanexistingsitewasrequired,theremaybeareasontoreconsiderthatplan.WithregardtotheSouthdispatchfacility,consultantsobservedthecloseproximityofpersonnelansweringcallsanddispatchingresources.Thedispatchroomisnotconduciveforeffectivecalltakinganddispatchoperations.Theroomisdesignedwithverylittlesoundabsorbingconstruction.Wallsrequiresoundingabsorbingelements.Dispatchandcalltakingpersonnelarealmostinarmsreachofeachother.Consolesrequiresoundabsorbingpanelsthatshieldtheindividualworkstationsfromeachother.Thecombinationoflimitedacousticabsorbingconstruction,personnelincloseproximitytoeachother,differentindividualspeakingvolumesandthelackofeffectivenoisecancellingheadsetsforthetelephoneconversationsresultsinexcessivebackgroundnoisethathampersoperations.

FINDING:CurrentPSAPs,trainingfacilityand“fleeto”planshavefacility

limitations,especiallyrelatedtoadequatespace.

Financial Structure

TheOperatorAgreementbetweenBrowardCountyandBSOclearlyspellsoutthemeansbywhichBSO,asthecontractor,istobecompensatedforservicesrendered.Article4.Compensation,Section4.2,oftheOperatorAgreement,statesthattheCounty“shallfundtheCapitalandOperationalExpensesoftheSystem.”Thissamesectionnotesthatthe“Countyshallprovideformanagement,administration,andoversight”oftheSystem.


AspartoftheCounty’sannualbudgetprocess,BSOdevelopsadetailedlineitembudgetinconcertwiththeCounty.Annualbudgetdocumentsprovidesufficientdetailtodeterminetheintentofexpenditures.TheCountymaintainsapprovalauthorityofthefinalbudgetamountandpositioncount.TheCountyandBSOmaintainstrictcontrolsonthebudgetedfunds.Atthebeginningofeachmonth,theCountyadvancestoBSOanannualizedmonthlypaymentbasedontheapprovedoramendedbudget.BSO’smonthlyreportsofactualexpendituresarereconciledagainstthemonthlybudgetandonaquarterlybasisanyexcessfundsarerecoupedbytheCountyinthefollowingmonth’sadvance.Pertheagreement,BSOmaintainsaseparatespecialfundexclusivelyforrevenueandexpensesassociatedwiththeE911System.The figurebelow indicatesa three fiscalyearhistoryofactualexpendituresandbudgetedexpenditures.12Figure6.BSOExpenditureandBudgetHistory

AppropriationLineItem FY14/15Actual FY15/16Budget FY16/17Budget

PersonalServices $37,878,111 $38,791,420 $40,865,470

OperatingExpenses $214,711 $453,940 $506,680

CapitalOutlay $160,200 $3,230 $0

Total $38,253,022 $39,248,590 $41,372,150

Positions 443 447 447

TheCounty’sFY16/17budgetfortheOfficeofRegionalCommunicationsandTechnologyincludesatotalof33positionsandthetransferoffundingtoBSOfordispatchservices.TheFY2017AdoptedOperatingBudgetindicatesthatORCAT’spositioncountincreasedbysixpositionssinceFY15/16asfollows:

§ Two positions were added in mid-year FY16 to provide “on-site county management andadministrationofthePublicSafetyNetwork.”

§ OneAssistantDirectorpositionisaddedtosupportongoingcapitalprojectandcustomerrelations.§ TwoInformationTechnologySpecialistpositionsareaddedtosupportPublicSafetyApplications.§ OneCommunicationsSystemTechnicianisaddedtosupporttheLocalGovernmentRadioSystem.

RevenuesupportfortheRegionalE911Systemisderivedprimarilyfrom911communicationsfees,advaloremtaxes(propertytaxes),andintergovernmentalrevenues.MajorcapitalexpendituresfortheRegionalE911systemareaCountyresponsibilityandareincludedintheCounty’sCapitalImprovementPlan.TheCounty’sFiscalYear2017-2021CapitalProgramincludesadditionalfundingforthereplacementoftheCADsystems,$48.7millionforthereplacementofthepublicsafetyradiosystem,and$350,000forplanningandanalysisstudiesregardingPSAPfacilities.

12BrowardCounty,FL.FiscalYear2017AdoptedOperatingBudget,BSOConsolidatedDispatchContract,p.2-34.


STAKEHOLDER INPUT

Methodology

Thisreportsectionprovidesfindingsandanalysesbasedonqualitativedatafromstakeholderinterviewsandsurveys.Alongwithdetaileddataanalysesoutlinedlater,thesetwodatasourceswereanalyzed,andspecificattentionwaspaidtointersectionsbetweenqualitativeandquantitativedata.

Stakeholder Input

Earlyintheproject,theconsultantsconductednumerousinterviewswithCountyandBrowardSheriff’sOfficeofficialsandotherkeymanagementpersonnelthatincluded:

§ BrowardCountyAdministratorandseniorexecutives§ BrowardSheriff’sOfficeofficials§ OfficeofRegionalCommunicationsDirector§ BrowardCountyChiefsofPoliceAssociation§ FireChiefsAssociationofBrowardCountry§ BrowardCountyLeagueofCities§ MembersofBrowardCityCountyManagersAssociation

Inaddition,consultantsinterviewedRegionalE911management,communicationsoperatorsandCountystaff.Consultantsobserveddispatchprocessesandoveralloperations.AttheendofMarch2016,FITCHlaunchedasurveytoolspecificallyforRegionalE911CommunicationsOperatorsandaseparatesurveyforRegionalE911managementpersonnel.Surveyinvitationsweresentto377dispatchpersonnelandobtaineda34.5%response.Fifty-onesurveyinvitationsweresenttodispatchmanagementpersonnelanda47%responsewasobtained.Surveyresponseswereanonymous.

Stakeholder Perceptions – Level 1 Interviews

Attheinceptionofthisproject,andthroughoutitsinitialphase,FITCHmetwithseniorlevelstakeholdersfromBrowardCounty,BrowardSheriff’sOffice,lawenforcementagencies,firerescueagencies,andmunicipalleaders.Thesediscussionsincludedbothelectedofficialsandseniormanagementpersonnel.Thefocusofthesediscussionswastounderstandperceptionsandkeyconcernsregardingtheinitialimplementationandcurrentoperationsoftheregionalcommunicationssystem.Theissuesraisedinthesediscussionshelptofocustheanalysisofquantitativedataandensuresalientitemsarecaptured.Fromaqualitativeperspective,thesediscussionsprovidedinsightintotheperceptionsamong,andworkingrelationshipsbetween,majorstakeholders.Therewasahighdegreeofconsensusonavarietyofissues-bothpositiveandnegative.Whiletherootcauseofsomeitemsmaybeperceiveddifferentlybysomestakeholders,theconsistencyofthefollowingitemsindicatesthatfutureattentioniswarrantedtoaddresstheissuesraisedherein.


Positive Issues Related to the Regional E911 System

DuringdiscussionsleadingtoconsolidationofthedisparatePublicSafetyAccessPointsintoanintegratedregionalcommunicationssystem,stakeholdersidentifiedeightgoalsfortheirnewsystemasnotedbelow.Figure7.GoalsfortheBrowardRegionalE911System

1.Improveservice2.Employthebesttechnologyavailabletoexpediteemergencyresponse3.Establishconsistentperformancemetrics4.Reducedelayintransferofemergencycalls5.Fasteremergencyresponsetimes6.Enhanceinteroperabilityandcoordinationamongstrespondingagencies7.Fewererrorsduetostandardizedcallhandlinganddispatchprotocols8.Savesignificantamountoftaxpayers'dollars

Whilesomeofthegoalswerenotmetinthefirst20monthsofoperation,itisimportanttoacknowledgethatothershavebeenrealized.Thesearesummarizedbelow.§ Stakeholdersgenerallyagreedthatthenewsystemhasalreadyestablishedsomestringentperformance

measures,andthatthesemeasuresarebeingreportedinaconsistentmanneranddisseminatedwidely.Theseattributeswereabsentpriortoconsolidation.

§ Thegoaltoreducedelaysrelatedtotransferringmisdirected911callersfromonePSAPtoanotherhasbeendramaticallyreducedsinceinceptionoftheSystem.Quantitatively,therehasbeenasignificantreductioninthenumberoftransfersforemergencycallers,therebyreducinganydelaysingettingassistancetopersonsinneed.Table5belowreflectsthesignificantreductionthathasoccurredbetweenOctober2013toJanuary2016.ItisgenerallyassumedthatcalltransfersbetweenPSAPsincura30-secondimpactontotalcallprocessingtimes.13

Figure8.HistoryofCallTransfersBetween911Centers14

2013StandAlonePSAPs

(October2013)2014ConsolidatedPSAP

(October2014)2016ConsolidatedPSAP

(January2016)PercentChange2013–2016

Countof911Transfers 12,291 7,581 1,690 (86.25%)

13SeeSection7.4.4fromNFPA1221(2016).14DerivedfrompresentationtoFloridaE911Coordinatorsfoundat(http://www.dms.myflorida.com/content/download/111575/622381/broward)andJanuary2016ConsolidatedCommunicationsMonthlyReport.


FINDING:Thenumberof911callersrequiredtobetransferredhasbeenessentially

eliminatedundertheconsolidatedregionalsystem,andreducedtotalcall

processingtimesbyapproximately30seconds.

§ End-usersacknowledgethatcollaborationamongfirstresponderagencies,bothlawenforcementandfirerescueagencies,hasincreasedsincetheregionalsystembegan.Thiscollaborationincludesgreaterconsistencyindispatchpolicyandproceduresandmorecommonnomenclatureamongfirstresponders,therebyenhancingcoordinationandcontrolinthefield.ParticipantsalsoacknowledgetheleveloftransparencyintheRegionalE911Systemissignificantlygreaterthanagenciesexperiencedundertheirformermodel.

§ ItwasfurthernotedbyallstakeholdersthatBrowardCountyisintheprocessofupgradingmajortechnologicalcomponentsoftheRegionalE911System.TheCountyismakingsignificantinvestmentswhichwilladdressthecomputer-aideddispatch(CAD)system,publicsafetyradiosystem,andfirestationalertingsystem.TheseupgradesrepresentsignificantcapitalexpendituresfromBrowardCounty’sCapitalImprovementPlanBudget.WhiletheCountyhassoughttoensurestakeholderinputiswidespread,someexternalstakeholdersbelievethatoutreacheffortstotheend-usersneedtobefurtherstrengthened.

Issues of Concern Related to Regional Communications

ItisclearthatthemajorityofstakeholdersbelievetheSystemhasimproveditsoverallperformancesinceRegionalE911’sformallaunchinOctober2014.Nonetheless,thereremainconcernsthatexistingprocessesandgovernancestructureskeepthesystemfromachievingsignificantadditionalimprovements.Oneofthemajorconcernssharedbyallstakeholdersisthestateofrelationsamongthevariousparties,specificallytheCounty;BSOintheirroleascontractorinoperatingtheregionalcommunicationsystem;andend-users,namely,lawenforcementandfirerescuepersonnelwhoprovidedirectservicestoresidentsandvisitors.AllpartiesrelyontheRegionalE911System’stechnologyandoperationstosupporttheirrespectivemissions.Whilethesestakeholdergroupsareclearlyengagedandmotivatedtoachievethesamegoals,thereisaconsensusamongthepartiesthat“somethingisbroken”.Everygroupindicatedthat“thereisalackoftrust”betweensystemparticipants.StakeholdersotherthanBrowardCountyattributemuchofthistotheCounty’sroleinsystemoversight.FromtheCounty’sperspective,theyremainresponsiblefortheoverallsystem.Thisresponsibilityisparamountduetotwofactors.First,theBrowardCountyCharter,ArticleV.–PublicSafety,Section5.02.–Fireprotection,notesthattheCounty“shallprovidefundingforthecommunicationsinfrastructure...[that]shallfacilitateclosestunitresponseforlife-threateningemergencies...”TheCounty’sresponsibilitiescanonlyberealisticallyachievedthroughcoordinationamongvariousproviderstoensurearegionalapproach.Thisprocessmustincludeutilizationofcommon


technologyandapplicationofconsistentpoliciesamonglawenforcementandfireagencies.Aregionalsystemisthemosteffectiveandefficientmethodtoaccomplishthismandate.Withtherecentconsolidation,BrowardCountyisabletoprovideforclosestunitresponse.However,fire-rescueagencieshavenotyetadoptedthenecessaryprotocols,andthereforetheCountyandBSOareunabletoimplementthissystem.

FINDING:Theconsolidatedsystemiscapableofclosestunitresponsetolife-

threateningemergencies,butprotocolsarenotyetinplacetoimplementthis

capability.

ThesecondfactoristhatBrowardCountyfundstheregionalbackboneandcarriesthefinancialburdenforthetechnologyandinfrastructuretoachieveregionalcommunications.TheCountyalsofundsthecontracttostaffandoperatethethreeRegionalE911sites.Duringinitialdiscussionscontemplatingconsolidationtoaregionalcommunicationssystem,participants,largelymunicipalandend-userrepresentatives,draftedaseriesofstringentperformancemeasures,mostlypatternedaftervariousnationalrecommendationsincludingfromtheNationalEmergencyNumberAssociation(NENA)andNationalFireProtectionAssociation(NFPA)and.Byalmostalladmissions,thesemetricsarerelativelyaggressiveandweredesignedtoreflectthedesireofthecommunitythatpublicsafety,andspecifically911services,shouldmeethighstandardsofperformance.Forsome,adoptionofthesemeasures,asoriginallydraftedandtheirsubsequentadoptionasbenchmarksforRegionalE911’sperformance,canarguablybeconsidered“stretchgoals.”ThesemeasuresareembodiedintheBrowardCounty/BrowardSheriff’sOfficeoperatoragreementtitled,“TheOperationofCall-Taking,Teletype(QueriesOnly)andDispatchServicesfortheConsolidatedRegionalE911CommunicationsSystem.”AttachmentB,PerformanceMeasures,providesthedetailofperformancemeasuresasoutlinedinExhibitDoftheOperatorAgreementandwithintheagreementsbetweenBrowardCountyandparticipatingmunicipalities.Additionaldocuments,includingtheinterlocalagreementsrelatedtotheregional911systemcanbefoundonlineatwww.broward.org,RegionalCommunicationsandTechnology,RegionalAgreements.Stakeholdersclearlydesirestrongperformancemeasures,andtheCountyhastheresponsibilitytomonitorandreportonthatperformance.Yet,stakeholdersexternaltoBrowardCountygovernment,namelyBSOandtheparticipatingcities,believethattheCounty’sapplicationoftheseperformancemeasureshas,insomeways,beenunreasonableandpunitive.Countystaffbelievestheyhaveappliedthestandardsconsistentlyandwithinthelanguageoftheapplicableinterlocalagreements,whichcanbemodifiedwithstakeholderconsensus.Thisissueofrelevantandmeaningfulperformancemeasuresisanareaofsignificantfrictionbetweentheparties.FITCHhasidentifiedanumberofproblemsinthecurrentassessmentofSystemperformance.ThisissueisdiscussedinmoredetailintheDataAnalysis


sectionofthisreport,withspecificrecommendationsregardingappropriateperformancemeasuresprovided.StakeholdershaveexpressedconcernwiththequalityofservicesbeingprovidedbytheBrowardSheriff’sOfficeastheSystemOperator.Someconcernsrevolvearounddispatchercompetency(largelyseenasanoutcomeofthecurrenttrainingreceivedbyRegionalE911personnel)andtheapplicationofpoliciesandprocedurescurrentlyusedbycalltakerandradiooperatorpersonnel.Thereisalsoaperceptionthatcollectivebargaining/laborissueswithintheSheriff’sOfficehaveimpededtheSheriff’sabilitytoeffectivelymanagetheworkforce.Theselimitationsarebelievedtohaveledtopoorerserviceandsupportforfieldpersonnel.StakeholdersoutsideBrowardCountyperceivethattheCounty’sintenseprocess-drivenoversightofthesystemischaracteristicoftheseissuesoutlinedabove.ComplaintstowardsCountystaffincludethattheCountystaffisessentiallyattemptingto“runoperations”ofthelawenforcementandfirerescueagencies.ThesestakeholdersciteexamplesoftheCountydefiningandmanagingprocessesforsystemchanges.Uponexaminingthisissuemoreclosely,FITCHhasidentifiedexamplesoftheCounty’sworkintrudingintoareasthatareclearlyoperationalinnature.Whilethereissomevaliditytotheseconcerns,itmustbefurthernotedthatend-usersofthesystem,namely,lawenforcementandfirerescueagencies,haveinmanywaysacquiescedcontroltotheCountybyagreeingtoasomewhatlimitedandambiguousroleforinputintothesystem’soperations.Most,ifnotall,protocolchangesandguidanceofthesystemoccursafterfireandpolicechiefassociationshaveapprovedofthesechanges.Therealsoexistsanincidentmanagementsystemdesignedforend-userstoidentifysystemissues.Thisissueofoversightandroles/responsibilities,inessencetherulesofengagement,isonethatFITCHhasidentifiedasrequiringmoredialogamongtheparties.Thisreportincludesrecommendationsonwhatrulesshouldbeadopted.Finally,anumberofmunicipalelectedandchiefexecutiveleaderslargelyvoicedsimilarconcernstothoseabove.Oneadditionalconcernbymanycommunitieswasthattoomuchemphasisisplacedonperformancemetricsinlieuofensuringthequalityofservices.However,leadersinPembrokePinesvoicedtheopinionthatcall-processingtimes(referredtoasP2/P3)neededgreaterattentiontomorecloselymirrortheirexperiencepriortoconsolidation.Whilemunicipalleadersdonotfullyshareacommonperspective,theprevalentconcernbymanylocalleadersofanoveremphasisonmetricsoverqualitycanbebettercharacterizedasgoaldisplacement.TheCounty’ssignificantfocusonperformancemetricsandmanagingchangeprocessesincludingsomeofanoperationalnature,doesnotlenditselftofullyallowanimblesystembedevelopedthatcanadjustandensurequalityservicesandmeetend-userexpectations.Theimpactofthisgoaldisplacementhas


ledtotheSheriff’sOfficeexpendingextraordinaryefforttoaddressprocessissuesratherthandealingwithmoresubstantiveissuesofstaffing,training,andstrongerRegionalE911oversight.

FINDING:TheCountyhasinappropriatelymade,andpublicsafetyofficialsallowed,

someoperationaldecisionstobehandledbytheCountythatshould,instead,be

determinedbypublicsafetyofficials.

Stakeholder Perceptions - Level 2 & 3 Interviews

Interviewsofmid-levelandsupervisorypersonnelwereconductedacrossallthreemajorstakeholdergroups:TheCounty,BSO,andend-users.Level3interviewswithCommunicationsOperators/DispatchersatNorth,Central,andSouthRegionalE911facilitieswerealsoconductedinthefirsthalfofMarch2016.One-on-oneinterviewswereconductedonavoluntarybasisusingopen-endedquestions.Positiveattributesnotedconsistentlythroughouttheinterviews,werethatRegionalE911personnelarededicated,wanttosucceed,wanttodoagoodjob,andtheyfeelthatfailureisnotanoption.Intheinterviewprocess,concernswererepeatedlyexpressedaboutthefollowing:

§ Teamwork§ Personnelintegration§ Inefficientprocedures/processes§ Ongoingtrainingandaccountability§ Qualityimprovement/assurance§ Equipmentfailuresandemergencyprocedures§ Staffingandworkschedules§ Workenvironment/respect

Belowaresummariesofcommentsregardingeachoftheaboveitems.

Teamwork

OnethemethatemergedthroughouttheLevel2and3interviewscaneasilybedescribedassilosorthelackofteamwork.Mid-levelmanagersandsupervisorypersonnelwerequestionedonhowtheirroleintegrates,orotherwiseassistsothersinresolvingsystemissues.Theexpression,“Idon’tlookatthat,”or“someoneelsedealswiththat,”wasacommonresponse.Therewaslittleevidencethatsupervisoryandmid-levelmanagershaveachievedamoreglobalperspectiveoftheSystem’sfundamentalgoals.Therewasalsolittleevidenceofasenseofteamworkbetweenvariousoperatingunits,evenwithinthesameemployer.


Personnel Integration

Jobclassificationtitlesandskillsofpersonnelfromsmallerdispatchcentersdidnotequatetorequiredcorecompetenciesneededtoachievesuccessinaregionaldispatchcenter.Initialtrainingwasconductedmonthspriortotheregionalimplementationanditappearsthatsomepersonnelinthesmallercentersreceivedlimitedinitialtrainingthatprovedinadequate.Thesefactorsmadeintegrationofpersonneldifficultatbestand,insomecases,continuestoimpactoperationalefficiencies.

Inefficient Procedures/Processes

Variousproceduresaretime-consumingandineffective.Thetrainingprocessof“readandsign”isineffectivedueprimarilytothevolumeofdocumentscirculated.Twotothreenewreadandsigndocumentsarepublishedperweekthatrequiredispatchpersonnelacknowledgementandunderstanding.Itwasreportedthatmanyofthesedocumentseithercancel,modifiedperviouspolicies,orarenotapplicabletodispatchoperations.TheTrainingSupervisorisrequiredtomanuallygradetrainingtestsandassessments.Thisisatime-consumingandinefficientprocessthatcanbealleviatedthroughthepurchaseofarelativelyinexpensivegradingdevice.Thiswouldprovidethetrainingsupervisorwithmoreactualtrainingtime.Policiesaffectingfire,law,andEMSagenciesarenotcommunicatedtofieldpersonnelinatimelymannercausingconflictsbetweenthefieldandBSOdispatchers.DutyOfficersaremireddowninadministrativedutiesandarenotfocusedonsupervisingdispatchpersonnelormaintainingsituationalawareness.

On-going Training and Accountability

Dispatchersexpressedasaprimaryconcernwhattheyperceiveasalackofqualityon-goingtraining.PersonnelreportedthattheTrainingOfficershavenotformallymetwiththeTrainingSectionintwoyears.Thiscancreategapsinknowledgeexchangeandnewtrainingtechniques,anddoesnotallowfordiscussionofthestrengthsandopportunitiestoimprovenewpersonnel.Itwasalsoreportedthatdispatchpersonnelareoftenheldaccountablefortrainingtheydidnotreceive.

Quality Improvement/Assurance

PersonnelexpressedthattheirperceptionoftheBSOQualityAssuranceunitisthatitisfocusedmoreonpunitivemeasuresthanskillsenhancement.

Equipment Failures and Emergency Procedures

CADoperationalissues,lock-ups,slowdowns,andrebootsareadailypartofBSOoperations.Whilethereremainsareportingsysteminplaceforthesetypesofissues,end-usersadmittheydon’treportproblemsbasedontheirexperienceof“noresponse”topriorefforts.DispatchpersonnelexpressedlimitedknowledgeortrainingonmanualmodeproceduresintheeventofaCADfailureforanextended


duration.CalltakersanddispatchersreportednorealaccesstothecriticalsupportivedocumentationsuchasmapbooksorruncardsintheeventofaCADfailure.Likewise,whendispatcherpersonnelwerequeriedabouthurricaneoperationsandpreparednesstheyexpressedlittletonoknowledge.TheonecommonprocedurementionedisthattheyaretoreporttotheE911Center.Personnelcouldnotidentifywhetheron-sitesuppliesorsleepingarrangementsareaccountedfor,nordidtheyexpressknowledgeofschedulingorspecialoperationalexpectations.

Staffing and Work Schedules

Inadequatestaffingwasarecurringthemevoicedbydispatchpersonnelatalllevels.Dispatchersreportthatmandatoryovertimeisassignedmultipletimeseachweek.Personnelvoicedthatthecurrentworkschedulecompoundedwiththefrequencyofmandatoryovertimeiscreatingburnoutandhighstresslevels.ItwasnotedthatBSOcurrentlyutilizesonly8-hourshiftschedulesforpersonnel.Thispracticeisnottypicallyseeninlargedispatchcenterswhere8,10and/or12hourshiftsinvariouscombinationsareemployedtomoreeffectivelyalignstaffingwithsystemdemands.Asnotedlater,FITCHfoundevidencethatBSOadjustsstaffingpatternsveryeffectivelytoaddressvarianceindemand.Yet,alternateshiftschedulesmayalsoprovidegreatersatisfactiontoemployeesandhelpaddresscurrent‘burn-out’perceivedbymanyworkinginthe911centers.FITCHwillprovidespecificrecommendationsregardingalternateschedulingpracticesinsubsequentreports.

Work Environment/Respect

RegionalE911personnelwhoareco-locatedinfacilitieswithotheragenciesandorganizationsnotethattheyhavelimitedaccesstobasicbuildingfacilitiessuchasrestrooms,elevators,parking,andentrancesites.Communicationsoperatorsnotedthatsomeagenciesdonottolerateanytypeofdisrespecttowardsdispatchpersonnelwhileothersseemtoignorethenegativebehavior.PersonnelperceiveexcessiveinvolvementbytheCountyinoperationalissuesandmentionthataCountysupervisoroccupiesanofficeonthedispatchfloorwhiletheNorthCentersitemanagerislocatedonadifferentfloor.

Dispatcher and Management Surveys

Inanefforttoexpandoutreachtostakeholders,FITCHlaunchedtwosurveytools,onefordispatchpersonnelandonefordispatchcentermanagement.Thepurposeofthesurveyswastoobtainabroaderrangeofimpressionsandopinionsfromthepersonnelbymeansofananonymoustool.OnMarch29,2016,surveyinvitationsweresentdirectlyfromtheFITCHofficesto377dispatcherpersonneland51managementpersonnel.ThesurveytoolwasavailablefortwoweeksandclosedonApril12.Thesurveyaddressedservicelevels,workloads,equipment,attitudesandmanagement.Participantswereprovidedwithstatementsandaskedtoindicatetheiragreementordisagreementwiththestatementusingthefollowingchoices:


1. Stronglyagree2. Agree3. Neutral4. Disagree5. Stronglydisagree

Ofthe377invitationstodispatchpersonnel,130personnelsubstantiallycompletedthesurveyresultinginaparticipationrateof34.5%.Therewere15additionalincompletesurveysandthoseanswerswereincorporatedintotheresults.Ofthe51invitationssenttomanagementpersonnel,therewere24completedsurveysresultinginaparticipationrateof47%.Therewerealso9incompletesurveysandwhereapplicable,thoseanswerswereincorporatedintothesurveyresults.

Survey Participant Demographics

RespondentsarefairlywelldistributedacrossthreeofthefourworklocationsasnotedinTable6below.

Figure9.WorkLocationsofDispatcherandManagementSurveyRespondents

WorkLocation %ofDispatcherRespondents %ofManagementRespondents

PublicSafetyBuilding 8% 20%NorthDispatch 31% 32%CentralDispatch 26% 20%SouthDispatch 35% 28%

Ofthedispatchsurveyrespondents,thelargestpercentage(47%)workedforBSOpriortoconsolidationoftheRegionalCommunicationsCenter,andthenextlargestcontingent(14%)previouslyworkedfortheCityofFortLauderdale.Approximately15%ofdispatchsurveyrespondentsreportedthattheyhadnotpreviouslyworkedforanyoftheparticipatingcitiesorforBSO.Ofthemanagementsurveyrespondents,thelargestpercentage(46%)workedforBSOpriortoconsolidationandthenextlargestcontingent(25%)previouslyworkedfortheCityofFortLauderdale.Approximately4%ofmanagementsurveyrespondentsreportedthattheyhadnotpreviouslyworkedforanyoftheparticipatingcitiesorforBSO.Thefiguresbelowindicatethepercentageofdispatchandmanagementsurveyrespondentsandthenumberofyearsofexperienceworkingina911environment.


Figure10.DispatcherSurveyRespondents’911WorkExperience

Seventy-fivepercent(75%)ofdispatchsurveyrespondentsreportedthattheyhadworkedina911environmentforatleastsixandupto15ormoreyears.Eighteenpercent(18%)indicatedtheyhadtwoyearsorlessexperienceworkingina911environment.Figure11.ManagementSurveyRespondents’911WorkExperience

Noneofthemanagementsurveyrespondentsreportedfewerthansixyearsworkexperienceina911environmentandtheoverwhelmingmajority,75%,reportedatleast15yearsofexperience.

18%

7%

23%

18%

34%

0% 5% 10% 15% 20% 25% 30% 35% 40%

0*2+years

3*5+years

6*10+years

11*15+years

15++years

9*1*1+Work+Experience+of+Dispatch+Survey+Respondents

0%

0%

12%

12%

76%

0% 10% 20% 30% 40% 50% 60% 70% 80%

0+2,years

3+5,years

6+10,years

11+15,years

15+,years

9+1+1,Work,Experience,of,Management,Survey,Respondents


Summary of Survey Results

Foreaseofreading,thefiverangesofagreement/disagreementwithsurveystatementsaresummarizedintothreegroupsasfollows:

1. agree/stronglyagree,2. neutral,and3. disagree/stronglydisagree.

ThefiguresthatfollowprovidethesummarizedpercentagesforboththeDispatchandtheManagementsurveysandresultsaregroupedintothreecategoriesasnotedabove.

Service Level Statements

Figure12.Ibelieveweprovideagoodlevelofservicetocitizenswhocall911.

DispatcherResults ManagerResults

§ 64%eitheragreed/stronglyagreed§ 13%wereneutral§ 23%eitherdisagreed/stronglydisagreed


Figure13.Pleaseratethefollowing:Ibelieveweprovideagoodlevelofsupporttopublicsafetyfieldpersonnel.




Figure14.CallersforemergencyservicesprovideaccurateinformationregardingtheADDRESSofthe

emergency.

DispatcherResults ManagerResults§ 5%eitheragreed/stronglyagreed§ 17%wereneutral§ 78%eitherdisagreed/stronglydisagreed

Notaskedofmanagers

Figure15.WhenIbeganmycurrentjob,theinitialtrainingIreceivedpreparedmewellforthework.




Figure16.Theongoingtraining,Ireceivecontinuestoenhancemyskills.





Figure17.TheRegionalCommunicationsSystemisequippedandpreparedtohandlelargescaleemergencies

suchashurricanesormassshootingincidents.




Workload Statements

Figure18.Theworkmethodsweutilizehelpimprovetheefficiencyinourwork.




Figure19.Thetechnologiesweutilizeimproveourefficiencycarryingoutourwork.




Figure20.Policiesandproceduresareeasilyunderstoodandapplied.




Equipment Statements

Figure21.Icaneffectivelyusetechnologytolocatewirelesscallerswhodon’tknowtheirlocation.


Notaskedofmanagers

Figure22.Thetechnologyandinformationsystemsweusearereliableandareappropriatetothejob.




Figure23.CADhasthetoolsIneedtohandleincidentsefficiently.


Notaskedofmanagers


Figure24.EquipmentproblemsarehandledappropriatelyandIgetfeedbackonproblemsIreport.




Attitude Statements

Figure25.OtheroccupantsofthebuildingIworkattreatmewithrespect.




Figure26.Uppermanagementsupportsouroperations.


Notaskedofmanagers

Figure27.ThereiscleardivisionbetweentheCountyandBSOonwhomanagesthecommunicationscenter.




Figure28.Dutyofficersandsitemanagersareavailableandwillingtohelpmewithproblemsorconcerns.


Notaskedofmanagers

Figure29.Ireceivefeedbackonmyjobperformance,includingpositiveacknowledgement.




Figure30.Pleaseratethefollowing:Differentworkscheduleswillimproveourcurrentstaffingchallenges.





Figure31.Publicsafetyfieldpersonneltreatthedispatchcenterpersonnelprofessionally.




Management Statements

Figure32.ManagementgivesteammembersaclearpictureofthedirectionBSOCommunicationsisheaded.


Notaskedofdispatchers


Figure33.Managementunderstandsthedailyproblemswefacewithourjobs.




Figure34.Overall,Iamsatisfiedwiththejobbeingdonebymyimmediatesupervisor.




Figure35.Managementencouragesotherstoproposenewandinnovativeideas.




Figure36.Managementeffectivelydealswithmisconductorunsatisfactoryperformance.





911 Center Concerns Rankings

Figure37.Pleaserankthefollowingissuesinorderofimportance(1isyourtopconcernand5istheleast

concern.

DispatcherRankings ManagerRankings

1. Adequatestaffing2. Officersafety3. Increaseinworkload4. Lossofspecificcommunity5. Improvedaccountability

1. Adequatestaffing2. Officersafety3. Training4. Improvedaccountability5. Increaseinworkload6. Lossofspecificcommunity

Communication Center Equipment Satisfaction Rankings

Figure38.Pleaserankthefollowingissuesinorderofimportance(1isthemostsatisfiedtoyouand5istheleast

satisfied.

DispatcherRankings ManagerRankings

1. CAD2. 911telephonesystem3. Radiosystem4. Recordsmanagement5. Communicationcenterfacility

1. Radiosystem2. Communicationcenterfacility3. 911telephonesystem4. CAD5. Recordsmanagement

Inadditiontothespecificquestionssummarizedabove,anopen-endedquestionpermittedrespondentstovoiceissuestheyfeltmostimportant.Forlinepersonneltheissuesofmandatoryovertimeduetolimitedstaffingandtheneedforadditionaltrainingwerehighlightedmostoften.Supervisorypersonnelfeltmoststronglythattheinitialconsolidationwasrushedandthisresultedinamultitudeofproblemsthatremaintoday.Overall,theresultsabovehighlightanorganizationthathassignificantmoraleproblemsandfrustrationwithlingeringstaffing,trainingandmanagementissues.

FINDING:BSO’soperationofthePSAPsarechallengedwithsignificantmorale

problemsembeddedinissuesofstaffing,trainingandmanagement.


DATA ANALYSES

Sources of Data

Background

DispatchoperationsinBrowardCountyareconductedatthreelocations,theNorth,Central,andSouthdispatchcenters.Dispatchfunctionsinallthreecentersoccurat“intake”workstationsand“assignment”workstations.TheanalysesrequiredtocharacterizetheBrowarddispatchoperationsinvolvesquantitatingalltheworkloadsflowingacrosstheseworkstationsbytallyingalltheprocessingintervalsexperiencedattheseworkstations.Theprimarydatarequiredforthesecalculationsresideinthreerepositories:TheIntradoVIPERtelephonyserver,theComputerAidedDispatchSystem,andtheradiologs.Therealsoexistsalogofoutgoingtelephonecalls.FITCHenteredintothisprojectwiththeexpectationthatcompletedownloadsofrawdatafromthesethreesourceswouldbeavailable.Theactualavailabilityofrawdatawassignificantlyless.Thesingleexportofdatathatwentsmoothlywastheoutgoingtelephonelogs.Substantialdelayswereintroducedintotheproject’stimelineduetoexportsofincompleteandincorrectdataelementsprovidedfromCountystaff.Onceidentified,thesedataissueswerecorrectedorappropriateanalyticalapproachesweredevelopedbyFITCHtoaddressanylimitations.Thetelephonyserverandradiologspresentedmoresevereproblems.Inthesetwocases,Browarddidnothavethetechnologytodirectlyexportanydatafromthesesourcesinmachinereadableformats.Instead,FITCHwaspresentedwithhumanreadabletextdocuments.FITCHhadtoapplycumbersomeworkaroundstoconvertdatainhumanformattodatathatwasusefullymachinesearchable.

CAD Export

InterpretingthecontentsoftheCADexportwasnotasmoothprocess.TheprimaryproblemwasgettingCountystafftoprovidecleardefinitionsofwhicheventalonganincidentprocessingtimelinewasbeingloggedintowhichtimestampintheCAD.TheP1,P2,andP3timeintervalsarealldelimitedbystartandstoptimestamps.Initialdata,whenanalyzed,hadunusualcharacteristicsandwassubsequentlydeterminedtocontainincorrectdatafields.NewdatawasquicklyobtainedoncetheissuewasidentifiedtotheCounty,andFITCHwasabletoverifyitusefulnessfordataanalysis.Ultimately,CADdataforFIREandLAWincidentswasprovidedforJanuary2015throughDecember2015.

Telephony Export

BrowardCountystaffinformedFITCHthattheywereunabletooutputrawdatafromtheIntradoVIPERtelephonyserver.ThebesttheycouldprovidewastooutputhumanreadableCallDetailRecords(CDRs)astextdocuments.TheyoutputonereportperdispatchcenterperdayofyearfromJanuarythroughOctober2015inthetelephonysystem’sabbreviated“BasicFormat”.TheyoutputacombinedreportforallthreedispatchcentersperdayofyearforNovember,2015throughJanuary2016inthetelephonysystem’s“ExtendedFormat”.Morethanonethousandindividualreportdocumentswereprovidedto


FITCH.Thereportsthatwereprovidedwereintendedtobehumanreadable.Assuch,thetextfilesthatwereprovideddidnotconformtothestandardtextfileformatsroutinelyusedfordatatransfersbetweendatabases.FITCHhadtoconverteachoftheBrowardCDRreportstoamachinereadableformatsuitableforinputintoadatabase.Thisrequirededitingthereportdocumentsatthelevelofthehexadecimalbytescomprisingthefiles.Uponinspection,theBasicFormatCallDetailRecord(CDR)reportswerefoundtocontaininsufficientdetailsoftelephonyoperations,andwereunsuitablefortheanalysesrequiredfortheconductofFITCH’sstudies.BrowardinformedFITCHthatExtendedFormatreportswerenotavailablefortheperiodJanuary2015throughOctober2015duetoanupgradeofthetelephonysystem.Asaconsequence,theanalysesoftelephonydatainthisreportarelimitedtothethree-monthperiodofNovember2015throughJanuary2016forwhichtheExtendedFormatCDRswereavailable.GettingtheExtendedFormatCallDetailRecordsintomachinereadableformatwasonlythefirststep.Thereafter,theblockoftextdescribingeachsingleincidenthadtobeparsedintoindividualdatafields.FITCHreverseengineeredthetelephonyprimarydatatablefromthehumanreadablereportsthatweregeneratedbyBrowardfromthetelephonyserver.TheoverlapbetweenthetelephonydataandtheCADdataislimitedtoNovemberandDecember2015.Althoughnotcomplete,theconsultantsfeelthatthisisasufficientsampletocometomeaningfulconclusionsaboutthebehaviorofthesystemoverthewholeyear.ThisopinionisbolsteredbythelargenumberofincidentscapturedinthistimeperiodandthelimitedimpactofseasonalityhasonperformancedataintheBrowardsystem.

Radio Export

BrowardCountystaffinformedFITCHthattheywereunabletoexportrawdatafromtheradiologs.Theonlyinformationtheycouldprovidewasa611-pagePDFofayear-endsummaryreporttitled“TalkgroupsatZoneSummary150101–151231”.FITCHwaseventuallyprovidedacross-referencetableshowingacronymsfortheradiochannelsandtheagencybeingdispatched.Unfortunately,thecross-referencetable,asinitiallyprovided,wasinaccurate.Acronymsappearinginthecross-referencetabledidnotappearinthePDFoftheyearendsummary,andvice-versa.Multipleverbalinquirieswererequiredtofinallyachieveaconsistentpictureofacronymsfortheradiochannelsandtheagencybeingdispatched.Onlytwopiecesofrelevantdataperdispatchchannelweretobefoundinthedocument.Thefirstwasthetotalannualtransmit-receivetimeperdispatchchannel(air-time),andthesecondwastheaveragedurationpertalk-listencycle.Theannualair-timeperdispatchchannelwascombinedwiththeannualincidentcountperdispatchedagency,astakenfromtheCAD,toobtaintheaverageair-timeperincidentforeachspecificagency.ThesebroadaveragesaresufficientforthecalculationsofworkloadsneededintheErlangmodelingforthisreport.


CAD and CDR Timelines

Relationship

AccesstodataintheBrowardsystemiscomplicatedbecausethereisnosinglesourceforallofthefactoidsrequiredtodescribetheoverallperformanceofthesystem.ThetelephonyserverandtheComputerAidedDispatchsystemcontaintheprimarydatatablesforthesystem.Therelationshipofthesetwodatatablesisdiagrammedbelow.Theuppertimelinebelow,showsthenamesandrelativesequenceofthetimestampsthatcompriseaCallDetailRecord,CDR,foranincidentinthetelephonyserver.ThelowertimelineshowsthenamesandrelativesequenceofthetimestampsthatcomprisetherecordofanincidentintheCADserver.

Figure39.RelationshipBetweenCDRandCADTimelines

Theprocessingofanincidentbeginswhenacallringsintothe911trunklineat[CDRBegin].TheanswerdelayintervalP1,asdefinedtoFITCHbyBrowardCounty,extendsfromwhenthecallingphonenumberisvalidatedat[ANIValid]untilacalltakerisidentifiedasavailableat[CallConnected].Theintakecalltakerpicksupthecallat[AGENTCONNECTED].Thespilloverofdatafromthetelephonyserver(theCDRtimeline)totheCADserveroccursatthispoint.Thebeginningofthespilloverprocessisloggedinthetelephonydatatablesasthe[AGENTCONNECTED]timestamp.TheendofthespilloverprocessisloggedintotheCADdatatablesasthe[Received]timestamp.

CDR Timeline

CAD Timeline

CDR Begin

CDR End

Agent ConnectedCall Connected

ANI Valid

Agent Disconnected

ReceivedTransmitted

ArrivedEnroute

Closed

Dispatched

P3P2

P1

VIPER Spillover


Itisimportanttograspthatthereisnochangeincalltaker,thatis,thesamecalltakerremainsonthelineinthespilloverfromtheCDRtimelinetotheCADtimeline.Togetacompletepictureofwhatanintakecalltakeractuallydoes,itisnecessarytolookattimestampsloggedintobothtimelineswhich,inturn,requiresalinkbetweenthetimelines.Theintakeprocessinginterval,P2,extendsfromthe[Received]timestamptothe[Transmit]timestampwhentheintakedispatcherreleasestheincidenttotheassignmentworkstations.Theassignmentprocessinginterval,P3,extendsfromthe[Transmit]timestampuntilthe[Dispatch]timestamp.ThecombinedP2/P3intervalextendsfromthe[Received]timestampuntilthe[Dispatched]timestamp.ThecriticalfailureofinformationtechnologyintheBrowardsystemisthatthe[Received]timestampisemptyforasignificantnumberofincidentrecordsintheCAD.

FINDING:Formorethanhalfoftheincidentrecords,theeventintheCADcannotbe

linkedtotheuniqueCallDetailRecord(CDR)thatinitiatedtheincident.

Validation of [Received] Timestamps

Almosthalfofthe[Received]timestampsaremissingintheCAD.Thosethataremissingareblatantlyobvious.However,therearecorruptionsapparentevenamong[Received]timestampsthatarepresent.Thereareexampleswherethe[Received]timestamphasthewrongdatecomparedtotheothertimestampsthatcomprisetheincidentrecord.Thereareexamplesinwhichthe[Received]timestampischronologicallyafterthe[Transmit]timestamp,inlargepartbecausetheCADwasoverwritingtimestampswhenacalltakerrebidtheANI/ALIinformation.Thesecorruptionsbecamedetectablebecausetheyaresoextreme.Theconsultants’concernwasthatlessextremecorruptionsremainedundetectedamongthe[Received]timestamps.Forthoserecordswherea[Received]timestampexists,theCountyusesallthoserecordsfortheircalculationofperformancemeasures.Wherearecordhasatimestampwithanobviouswrongdate,Motoroladevelopedacomputerscripttoextractonlythetimeofdayfromtherecordtouseinitscalculationandignorestheerroneousdate.FITCHdeterminedavalidationofdataonthereceivedtimestampwasnecessarytoincreasethestatisticalvalidityofreportedperformance.Thisvalidationprocessisexplainedinmoredetailbelow.Tovalidatesomesubsetoftheexisting[Received]timestamps,theconsultantsappliedthefollowingmethodology.A[Received]timestampintheCADdatatableswasconsideredtobevalidatedwhentwocriteriaweremet:

Thereexistsan[AGENT_CONNECTED]timestampinthetelephonedatatableswithinthepreceding5seconds.

AND


ThetelephonenumberintheCADdatatablematchesthetelephonenumberinthetelephonedatatables.

The[Received]timestampintheCADdatatablesistakentologtheendoftheVIPERspilloverprocessfromthetelephonedatatables.The[AGENT_CONNECTED]timestampinthetelephonedatatablesistakentologthebeginningoftheVIPERspilloverprocess.TheVIPERspilloverprocessitselfrequires2–3secondstocomplete.A5-secondwindowwasappliedtoaccommodateanyslightoffsetsinclocktimebetweenthedatatables.

Statistics for Received Timestamps

Thefigurebelowprovidesstatisticsfortheavailabilityofvalidated[Received]timestampsaswellasthenumbersofincidentrecordsintheCADthatcanbeclearlylinkedtothetelephonerecordthatinitiatedtheincident.Figure40.Validated[Received]Timestamps11/1/2015through12/31/2015

Parameter Count Percentage

LAWRecords 136,595 With[Received]timestamps 36,417 26.7% With[Received]timestampsvalidated 24,131 17.7% With[Received]timestampsOut-of-Range 890 0.7%FIRERecords 43,722 With[Received]timestamps 29,369 67.2%

With[Received]timestampsvalidated 22,067 50.5% With[Received]timestampsOut-of-Range 235 0.5%

The[Received]datafieldcontainsthreecategories:NULLS,validatedtimestampsandnon-validatedtimestamps.Onlythevalidated[Received]timestampsshouldbeusedtocalculateP2/P3intervals.Evenwhenavalidated[Received]timestampisused,thereisstillnoguaranteethattheP2/P3intervalwillbefreeofreversebias.15

15Fromstrictapplicationofindustrialengineeringandstatisticalstandards,theCountycanmakenoassurancethattheP2/P3dataprovidedtoFITCHandtostakeholdersisstatisticallyvalid.TheinescapableflawwithallcurrentP2/P3statisticsisthattheydependonthe[Received]timestamp–ofwhichthereareonlysamples.TheCountyisunabletofullyidentifywhy/how[Received]timestampsaremissingor‘outofsequence’;doesnotknowifthereisabiasforhow[Received]timestampsareselectedtogomissingorallowedtobecome‘outofsequence’;doesnotknowifthereisa“reverse”biasforthe[Received]timestampsthatarelefttorunstatisticson(describedaboveasvalidated)–andthereforecannotstatisticallyprovethatremaining[Received]timestampshavebeenrandomlyselected.Withoutproofofrandomness,thennoneoftheP2/P3statisticsarecredibleunderstrictstatisticalmethods.Thisisnotauniqueproblemencounteredwithcomplexdataanalysis–yetaproblemnonetheless.Notwithstandingthisdisclaimer,theresultsreportedherearemadeunderanassumptionthattheremainingsampleprovidedistheresultofrandomness.


FINDING:Employingtheproceduresabove,FITCHfoundonly25.6%ofCADrecords

validforuseinanalysisofP2/P3.

Suitability of Performance Targets

TheConsolidatedDispatchSystemwaslaunchedwithhighexpectationsandaconcurrentsetofaggressiveperformancetargets.TheSystemwasdesignedtoincludeQualityImprovementTeamsandqualityassuranceprocessestomonitorperformanceasjudgedbymeetingornotmeetingspecifictargets–essentiallya‘PASS/FAIL’or‘YES/NOanalysis.WhiletheCountydoesreporttrenddataforcertainmetricsintheirsupplementalsections,thefocusonpercentage‘PASS/FAIL’or‘YES/NO’againsttargetsdoestheCountyadisserviceinthatitmayfosteranexpectationthatthesystemcansomehowbemadeperfect.Therealityofemergencyservicesystemsisthattheyareexpectedtobeoverwhelmedatsometimeoranother.ConsidertheimpactrecentshootingsinOrlandohadontheiremergencyservices–orarecenttornadoinBrowardCounty.Theinitialchallengeuponconsolidationwaslearninghowtomakethesystemwork.Forexample,theCountyhasimplementedfairlycomprehensivequalityassurance/qualityimprovementprocessesaspartoftheconsolidatedSystem.TheIncidentManagementTrackingSystemtoidentifyissuesfromendusersandOperationalReviewTeamsmadeupofendusers,addvaluetotheSystem.ThesetypeofeffortsallowforamoreclinicalperspectiveonhowtheSystemcanimprove,andhasledtheconsultantstofeelthatthesystemhasturnedacorner.Thechallengeisnowhowtomakethesystemworkevenbetter.Performancetargetsshouldbeselectedsuchthattheycontributetomakingthesystemworkbetter.

FINDING:TheCountyhasimplementedasetofqualityassurance&improvement

processesthatassistinobjectivelymovingtheSystemforward

TheinterpretationofthecurrentperformancetargetsisfromtheperspectiveofaPASS/FAILcutoff.Thisconceptisborrowedfromtheindustrialengineeringcommunitywhereitisreferredtoas“InspectionbyAttributes”.Themostformalized,currentembodimentofPASS/FAILacceptancetestingis“SamplingProceduresandTablesforInspectionbyAttributes”,ANSI/ASQZ1.4-2008.ThemethodologyusedinBrowardisclassifiedasa“singlesamplingplan”whereinalotisacceptedorrejectedonthebasisofpullingasinglegroupofsamplesfromthelotforinspection.W.EdwardDemingwastheindustrialstatisticianwhoiscreditedwithbeingamajorcontributortotheJapaneseindustrialresurgenceafterWWIIthroughhisintroductionoftotalqualitymanagement(TQM).DemingheldtheuseofPASS/FAILtargetsinverylowregard.HenotedthatthemainuseofPASS/FAILtargetswastobeatthesupplieroverthehead.Thecorollarytothisishisadmonishmentto“managethe


causenottheresult.”16DonaldWheeler,anotherwell-knownqualitycontrolexpert,cautionsthatyoucannotimprovethequalityintheprocessstreamusingPASS/FAILtargetsbecausethemethodteachesnothingabouttheprocessthatproducedtheproduct.17FITCHseesbothDeming’sandWheeler’sdynamicsplayingoutinBrowardCounty.TheattractionofPASS/FAILtargetsisthattheyareeasytoimplement,and,atfirstglance,appeareasytointerpret.Theunderlyingrealityismuchmorecomplexandlessconvenient.

FINDING:TheCounty’suseofPASS/FAILtargetsprovideslittleinthewayof

informationforcontinuousqualityandperformanceimprovement.

P1 Intervals

ThetargetthathasreceivedaninordinateamountofattentionfromBrowardstakeholdersgoesbythemoniker“P1”.Inthefigureabove,(RelationshipBetweenCDRandCADTimelines),theP1intervalsextendsfromwhenthecaller’stelephonenumberhasbeenvalidatedatthe[ANIValid]timestampuntilanavailableintakedispatcherhasbeenidentifiedatthe[CallConnected]timestamp.TheP1intervalisalsoreferredtoastheanswerdelay.ThistimeintervalisthesubjectofrecommendationsfromboththeNationalFireProtectionAssociation(NFPA)andtheNationalEmergencyNumberAssociation(NENA).

Implementation of the P1 Target

AspartoftheCounty’scurrentimplementationoftheP1target,dispatchoperationsofthepriordayarereviewed.The“busyhour”ofthedayisidentified,andtheanswerdelayinthathouriscomparedtothetargetinordertoissuethePASS/FAILassessmentforthathour.Forinstance,the“busyhour”lastWednesdaymayhavebeen1900hours,whilethe“busyhour”lastThursdaywas0300hours.Undercurrentpractice,the“busyhour”isavariablethatisselectedretrospectively.ThisimplementationislooselymodeledontherecommendationinNENA56-005andiswellunderstoodbyallstakeholders.Thismetricalonefailstorepresenttheoverallperformanceofthedispatchintakeoperationbyfocusingexclusivelyonone-offeventsthatrandomlyimpactthesystem.TheoutcomeoftheCounty’smethodologyisthatBSOisdriventodeploymaximumstaffingatallhoursofthedayanddisregardtheincreasedannualcostincurredtofixaone-offproblemthathappenedat3AMlastThursdaymorning.theCounty’simplementationoftheP1targetdoesnotleadtoactionableteachingsaboutthefunctioningofBSOdispatchoperations.

16TheW.EdwardsDemingInstitute,http://www.blog.deming.org,accessedMay2016.17DonaldJ.Wheeler,“UnderstandingStatisticalProcessControl”,SPCPress,1992.ISBN978-0-945320-69-2


NENA 56-005

ThetextofNENA56-00518Section3.1isreproducedinFigure7below.Figure41.NENARecommendation

TheCounty’simplementationofthe‘busyhour’criteriainNENA56-005,Section3.1,focusessolelyonthe“busyhour”oftheday,thusignoringtheother23hoursoftheday.Bydefault,thesehoursaredealtwiththroughanimpliedsyllogismthatmaybeparaphrasedasfollows:IF AlliswellinthebusyhourofthedayTHEN Allwillbewellintheremaininghoursoftheday.Takenbyitself,thissoundsreasonable.However,forthistobevalidandforNENA56-005toapplytoBroward,thesamenumberofdispatchersmustbeondutyatthebusyhourandatallotherhoursoftheday.ConfoundingtheapplicationofNENA56-005toBroward,BSOadjustsitsintakestaffingonanhour-by-hourbasis.TheCounty’simplementationoftheNENArecommendationdoesnotaccommodatethisreality.ThesecondcriteriainNENA56-005,Section3.1isthat95%ofallcallsshouldbeansweredwithin20seconds.WhenexaminingtheCounty’sreportingofthesetwocriteria,onemustconsiderthedisconnectintheirrelativeperformance–“busyhour”performancehaslargely“FAILED”whilethe95%within20secondscriteriahasPASSEDbyastatisticallylargedegree.Thisshouldcauseonetopauseandcontemplatewhy.

Busy Hour

AnyattempttoimplementNENA56-005requiresthatthe“busyhour”bedeterminedandthentheanswerdelayinthathourbecalculated.EvenifNENA56-005wasthecorrectmetrictoevaluateBSOdispatch,theCounty’sdeterminationofthebusyhourassumestheCountyshouldretrospectivelydefinethepreviousday’sbusyhour.Bycontrast,the“busyhour”istobedeterminedbyexaminingthehistoricrecordandcalculatingstatisticsoncallcountsineachhourofthedayacrosssomesubstantialspanofdays.The“busyhour”istobefixedasnotedbyNENAtobeapracticeinotherPSAPs.Itisnotavariable.

18NENACallAnsweringStandard/ModelRecommendation,NENA56-005,06/10/2006.NationalEmergencyNumberAssociation(NENA)StandardOperatingProceduresCommittee,Call-TakingWorkingGroup.


InthecaseofBrowardCounty,FITCHdeterminedthe“BusyHour”ofthedaytobe1800hoursaveragedoverCY2015asshowninFigure8below.Figure42.AverageBusyHourBasedonTelephoneTraffic

FINDING:CertainperformancemeasuresareapoorrepresentationofSystem

performanceandinconsistentwithcurrentindustrybestpractices.

Discrepancies Regarding Workstations

Answerdelaysaretiedtospecificworkstations,eachwithitsuniqueidentification(ID)number.Then,activitiesamongworkstationsareaggregatedforaspecificPSAP.FortheanswerdelayataPSAPtobevalid,therosterofworkstationsascribedtothePSAPmustbecorrect.FITCHconductedadetailedaccountingofthenumbersandIDsoftheactiveworkstationsbyhourofthedayateachPSAPfromNovember2015throughJanuary2016.Asacrosscheck,FITCHconductedthesameaccountingacrossBSOdispatchwithoutregardtotheidentityofthePSAPs.Thetwoaccountingscouldnotbereconciled.ThesumofactiveworkstationsattheindividualPSAPoftenexceededthesumofactiveworkstationsobtainedwhenPSAPIDsweredisregarded.Manuallysteppingthroughtherecordsinquestionrevealedthesourceofthediscrepancy.Inmultipleinstances,numberedinthethousands,asingleworkstationIDappearedundertwoPSAPs.Theseinstancesweremanuallyresolvedbyconsensus:allPSAP-workstationpairingswereadjustedtothereflecttheobservedmajorityPSAP-workstationpairingforeachworkstationinquestion.SincetheconsolidationoftheseparatetelephonynetworksinFebruary2016,thisworkstationIDduplicationhasbeenrectified.

0 4 8 12 16 20 240

50

100

150

200

250

300

Hour-of-Day

Brow

ard

E911

[Cal

ls p

er H

our]


Answer Delays

ThelaststepintheimplementationofNENA56-005isthecalculationoftheanswerdelayinthebusyhour.Themajorityof911callsenteringthesystemproceedinthenormalmannerfromring-intopick-upbytheintakedispatcher.Onthesecalls,theCountycalculatesanswerdelayscorrectly.However,thereareasubstantialnumberofcallsinwhichthecallerdisconnectsbeforetheintakedispatcherhastheopportunitytoconnecttotheincomingline.Onthesecalls,theCountycalculatesanswerdelaysinawaythatcouldbeunfavorabletoBSO.Answerdelaysonthemajorityof911callsarecalculatedastheintervalfromwhenthecallisreadytobepresentedtowhentheintakedispatcherpicksup.Thisportionofthetallyofanswerdelaysiscorrect.Theproblemisencounteredonthesecondsetof911callsinwhichthecallerdisconnectsbeforetheintakedispatcherpicks-up.Inthesecases,theCountykeepstheanswerdelayclockrunninguntiltheintakedispatcherconnectstothedeadline.FITCHtakesthepositionthatitmakesnosensetoincrementtheanswerdelayclockpastthepointwherethereisnolongeranyoneontheincomingline.Regardlessofthetwomethodsofcalculationnotedabove,therequirementforhang-up911callstobeproperlyaddressedby911personnelismet.ItturnsoutthataFAILSturnintoPASSESwhentheanswerdelayclockisstoppeduponcallerhang-upasreflectedintheexamplebelow.DifferenceswerefoundbetweentheCounty’scalculationsandthosedonebyFITCHbecauseoftheduplicateworkstationIDsidentifiedabove.TheCountyhasindicatedthattheissuewassubsequentlyresolved,butafterthetimeperiodforwhichdatawasprovidedtoFITCH.Figure43.ComparisonofORCATandFITCHPass/FailDeterminationBasedonAnswerDelays

PSAP Date HourORCAT FITCH

<10sec calls % P/F P/F <10sec calls % Ranked90th%-tile[sec]

South 01/15/16 2000 60 69 87.0 FAIL PASS 63 69 91.3 9.42

Thetruthtablesrequiredtocalculateanswerdelayswithandwithoutcallerhang-upsaresurprisinglycomplex.Forillustrativepurposes,thetruthtableandcodingrequiredtoimplementthecalculationofallanswerdelaysinFITCH’sdatatablesarepresentedinAttachmentC,CalculationofAnswerDelays.

Utility of the P1 Target

TheapplicationoftheP1busyhourtargetasameasureoftheoverallintakeperformanceoftheBSOdispatchsystemisapoorrepresentationofSystemperformanceandlacksthestatisticalvaliditytomeaningfullyserveasaguideforbalancingcostsagainstperformance.Acertainlevelof“overstaffing”inadispatchcenterisrequiredtoabsorbtherandomsurgesthatareexpectedinanysystem.

FINDING:ThefailureofthecurrentPASS/FAILorYES/NOP1busyhourtargetis

thatitprovidesnoguidanceastothelevelofsurgecapacitythatisfiscally

responsibletobuildintothesystem.


FITCH Assessment of Historic P1

Toprovideperspective,itisinstructivetoexaminethedetailsofBSO’shistoricanswerdelaysonpickupattheintakeworkstations.FromtheCounty’sNon-ComplianceReports,BSO’sCentralPSAPreceivedaFAILratingfor1600hoursonNovember7,2015.FITCHwenttotheCallDetailRecordsfromthetelephoneserverandcompliedthehistoricanswerdelayshour-by-hourforCentralPSAPonthisdate.Theseresultsarepresentedbelow.Figure44.AnswerDelaysatCentralPSAPon11/07/2015

Broward E911 Consolidated Communications SystemHistoric Answer Delays

FAIL 1600 Hours

([CIM] Call Connected) timestasmp - ([CIM] ANI: timestamp ) = Answer DelayExcept: Answer_Delay clock stops running when caller disconnectsas indicated by ([CIM] Caller Disconnected before Supervision) timestasmp

Central PSAP 2

00:0001:0002:0003:0004:0005:0006:0007:0008:0009:0010:0011:0012:0013:0014:0015:0016:0017:0018:0019:0020:0021:0022:0023:00

Hourof Day Σ Erlangs

Obs'd HrsOnTask

11/07/2015

18.08 4.70433.13 129.25

AverageErlangs

Phone Traffic

Avg Intakes per Hour

Ans Delay @ActiveWkStn's

Observed Staffing & Performance

Weighted% Immed Ans

WeightedAns Delay

446 1.42

90 %-tile [sec]OUT ADM 911

ADM 911OUT

1.311.38

1.591.431.401.491.581.501.371.381.391.41

1.37

1.391.341.401.31

1.311.581.661.601.371.281.33

1517

19202119212525232321

15

22212217

12101012192017

3.4893.109

5.6126.0257.3016.8734.9326.0086.0676.4845.6555.393

2.486

4.7175.1295.3016.312

2.9132.5821.9732.8293.2273.3025.175

8190

183168166176149188188187166173

63

121133148135

5460415784

118173

1719

38505455365546424642

15

38283531

1188

11353441

1113

21252930103122301118

6

10261522

411

120122630

Date PSAP Location ORCAT Assessment


Fourpointsareimportantinthedatapresentedabove:1. BSOcontinuouslyadjuststhenumberofactiveworkstationsbyhour-of-day.BSOdoesNOTuse

constantintakestaffing.2. BSOadjustsitsintakestaffingwithgreatfinesseasdemonstratedbytheconsistencyoftheanswer

delaysinthefaceofwidelyvaryingdemandbyhour-of-day.3. BSOdidNOTFAILat1600hourswhenFITCHcalculatedanswerdelayssoastoproperlyaccountfor

callerhang-ups.4. Theanswerdelaysineachhour-ofdayaswellastheweightedanswerdelayacrossall24hoursof

thedayareallexemplary.ItisFITCH’sexperiencethatBSO’sanswerdelaysabovearemorethancomparabletootherhighperformingdispatchsystemsinNorthAmerica.BSO’sanswerdelayatthe90thpercentileis1.4seconds.ThismeansthatBSOintakedispatcherspickupthenextincomingcallbeforethesecondring,ninetimesoutoften.Toputthatinperspective,theacceptableanswerrateis90%at3ringsor10seconds

P2/P3 Intervals

ThesecondtargetthatisthesubjectofattentionfromBrowardstakeholdersgoesbythemoniker“P2/P3”.Inthefigureonpriorpagesabove,(RelationshipBetweenCDRandCADTimelines),theP2/P3intervalsextendsfromwhentheVIPERspilloverfromthetelephonyservertotheCADserveriscompletedatthe[Received]timestampuntiltheincidentisreleasedbytheintakedispatchertotheassignmentdispatcheratthe[Transmit]timestamp.TheP2/P3intervalcanalsobereferredtoastheprocessinginterval.ThistimeintervalisthesubjectofarecommendationfromtheNationalFireProtectionAssociation.19NFPA1221definestwodifferent‘buckets’ofcalltypes,andforeach‘bucket’hasdifferentperformancemeasures.ThespecificsareoutlinedintheAssociation’sdocument,butgenerallyallowsEMSandotherspecializedincidentswithalongercallprocessingtime.WithinthisanalysiswedefinethefirstgroupasEmergencyMedicalDispatch(EMD),intendedtoevaluateunderonesetofcriteriaeightspecificcalltypesincludingthosethatrequireEMD20.Thesecond‘bucket’isshownhereas“n-EMD”,andgenerallyincludesmoretypicalfirerelatedcalls.Unfortunately,theCADhasnomannerwithprecisionbywhichtoidentifywhichcallsbelonginwhich‘bucket’.Forthatreason,theCountyindicatestheymakeassumptionsandsimplyplaceallmedicallyrelatedcallsintotheEMDbucket,whileeverythingelsegoesintothen-EMDbucket.Thispracticeisnotcompletelyconsistentwiththeadoptedperformancemeasures,butFITCHwasunabletoidentifyabetterprocessforthisadditionaldatalimitation

19NFPA1221(2016).StandardfortheInstallation,MaintenanceandUseofEmergencyServicesCommunicationsSystems.20Theothercalltypesincludecallsrequiringlanguagetranslation;callsrequiringtheuseofaTTY/TDDdeviceoraudio/videorelayservices;callsofcriminalactivitythatrequireinformationvitaltoemergencyrespondersafetypriortodispatchingunits;hazardousmaterialincidents;technicalrescue;callsthatrequiredeterminingthelocationofthealarmduetoinsufficientinformation;andcallsreceivedbytextmessage.Inaggregate,theseothercalltypesrepresentasmallnumberofoccurrences.


Treatment of [Received] Timestamps

CalculationoftheP2/P3intervaldependsonthe[Received]timestamp.This,ofcourse,presentsaproblembecauseonlyhalfofthe[Received]timestampscanbevalidatedintheCADforEMScalltypes(whichrepresentthelargestcategoryofcallsintheFirequeue).theCounty’sreportsappeartobeextremelyprecise.TheyreportpercentfailsintheP2/P3intervalsto±0.01%regardlessofthenumberofvalid[Received]timestampstheyactuallyhaveavailable.Whennumbersofavailablemeasureddataarerestricted,questionsregardingprecisionandconfidencemustbefaced.“Precision”istheintervalthatwillbrackettherightanswer:±10%,±1%,±0.01%?“Confidence”istheprobabilitythatrandomnoiseinthesamplesethasnotskewedtheanswer.Whenalimitedsetofdatagoesintoanaverage,theprecisionandconfidencelevelofthecalculatedaveragearenotamatterofopinion.Rather,theyarethesubjectsofspecificcalculations,asformallydescribedinthedocument“StandardPracticeforCalculatingSampleSizetoEstimate,WithSpecifiedPrecision,theAverageforaCharacteristicofaLotorProcess”,ASTM122-09e1.TheCountydoesnotshowthespecificprecisionsandconfidencesassociatedwiththeirreportedperformancemetrics.

Implications of the Missing [Received] Timestamps

Counter-intuitively,themissing[Received]timestampsposemoreofaproblemthantheonespresent.Themissing[Received]timestampserodethecredibilityoftheP2/P3intervalsthatcanbecalculatedfromthe[Received]timestampsthatareavailable.TheconvenientassumptionabouttheP2/P3intervals,ascalculatedbytheCounty,isthatthenumbersautomaticallyserveasametricforthesystemasawhole,thatthevariabilityintheP2/P3intervalsthattheycalculate,properlyreflectsthevariabilityinalltheP2/P3intervals,eventheonesnotabletobecalculated.Unfortunately,thisassumptionisnotnecessarilytrueandshouldnotcurrentlybereliedonasabasisforpolicydecisions.AgainreferringtoASTM122-09e1,forthepulledsub-lotofsamplestocorrectlyreflectthepropertiesofthefulllot,theprocessmustbeinastateofstatisticalcontrolwhereinthesub-lotofsamplesisinfluencedbyasinglesourceofvariability(asimposedbytheproductionprocess).Thisprocedurecannottreatmulti-levelsourcesofvariability.Thislimitationtakesusbacktothemissing[Received]timestamps.Tostartwith,wedonotfullyunderstandwhy/howthesetimestampsaremissing.Wedonotknowwhethertherewasaspecificbiasoperatingtoselectwhichtimestampswentmissing.ThefirstconsequenceisthatareversebiaswouldthenbeimposedontheP2/P3intervalscalculatedfromtheremainingtimestamps.ThesecondconsequenceisthatthecalculatedP2/P3timestampswouldbestatisticallybiasedandmaynotrepresentthepropertiesofthesystemasawhole.


Preliminaryinvestigationofthewhy/howbehindthemissingtimestampsindicatesthatoperatorinterventionbytheintakedispatchersplaysamajorroleinmissingtimestamps.Thisisaproblem,ashumaninterventionisalmostguaranteedtobevariableandtherefore,statisticallybiased.Evenmoreconfounding,thedegreeofbiasisthenalmostguaranteedtobeoperatorspecific,therebyintroducingatimedependentvariabilitytothebias.

Credibility of P2/P3 Statistics

FITCHalsolearnedthattheCAD[Received]timestampsbecomecorruptedwhenevera911calltakerrebidsorasksthe911systemsoftwaretoverifyandupdateacaller’slocation–essentiallyoverwritingtheoriginaltimestamp.Fortunately,thevalidationprocessFITCHemployedessentiallymitigatesthatbias.Arguably,theoverwritingofthe[Received]timestampshouldbenefitBSOintheCounty’sreportedcompliance.Howeveroncerejectingtheserecordsthroughthevalidationprocess,FITCH’scalculationofperformanceisbetterthanthatcalculatedbytheCounty–indicatingthereislikelyotherunknownfactorsstillinfluencingthisperformancemetric.TheCountyindicatedtheyincludeallrecordswith[Received]timestampsinanabundanceofcaution,whileFITCHemployedavalidationprotocolthatexcludessomerecords.WhileFITCHisabletoreportsomeP2/P3performance–readersshouldremainmindfulofthestatisticallimitationsandproceduraldifferencesdiscussedabove.Thefiguresbelowsummarizetheanalysisoffire-rescueincidents–distinguishingthosethathavebeencharacterizedasEMDrelated,andthoselabeled“n-EMD”.Ofthetotal43,722recordsavailable,only21,292areconsideredvalidforuseinthisanalysis.Figure45:EMDP2/P3Statistics&Performance

Parameter Value

EMDCount 39,214[Rcvd]absent 11,198[Rcvd]present 28,016[Rcvd]notvalidated 7,013[Rcvd]validated 21,003[Rcvd]validated>165sec21 718[Rcvd]validated<166sec 20,285 50th%-tile 54.72sec Average 61.16sec StdDev ±27.47sec 90th%-tile 100.80sec 95th%-tile 121.33sec 99th%-tile 157.79secCompliance Count<91sec 17,496 %<91sec 86.30% Count<121sec 19,331 %<121sec 95.30%

21TheP2/P3intervalsforEMDCallTypesgreaterthan165secondswereassumedtobe“purposefullypending”andexcludedfromanalysis.


Figure46:n-EMDP2/P3Statistics&Performance

Parameter Value

n-EMDCount 4,508[Rcvd]absent 3,155[Rcvd]present 1,353[Rcvd]notvalidated 289[Rcvd]validated 1,064[Rcvd]validated>180sec22 57[Rcvd]validated<181sec 1,007 50th%-tile 68.70sec Average 76.54sec StdDev ±33.16sec 80th%-tile 103.70sec 95th%-tile 147.00secCompliance Count<91sec 388 %<91sec 38.50% Count<121sec 831 %<121sec 82.50%

ForEMDcalltypes,actualperformancefellshortofthestatedbenchmarkby4%atboththe90thand99thpercentilecompliancetargets.Theimportantquestionforpolicymakersiswhatdoesthatshortcomingrepresentfromapracticalperspective.Thefigurebelowhighlightstheanswerbynotingtheactualperformancewouldrequireagoalofalmost101secondsinordertoreacha90thpercentilecompliance.Therefore,atthe90thpercentilethevariancebetweenactualperformancecomparedtothetargetisjustunder11seconds.While11secondsatthe90thpercentilemayseemsignificantitshouldbenotedthatthismaynotbetrue.ThemeasurementusedbytheCountyincludesallcalltypes,hotandcoldresponses.Responsetimehasalimitedoutcomeeffectonaverysmallsubgroupofcalls,namely,whatinEMDvernacularisreferredtoas“EchoandDelta”calls.Thissubsetrepresentsfewerthanaquarterofthecallsthatarebeingmeasured.Further,thissubsethasfewerquestionsatthecalltakingpointandisusuallysignificantlyfastertoprocessthanthelargergrouppfcalls.

22TheP2/P3intervalsfornon-EMDCallTypesgreaterthan180secondswereassumedtobe“purposefullypending”andexcludedfromanalysis


Figure47.P2/P3PerformanceforEMDCalls–TargetversusActual

Notwithstandingthechallengesdescribedabovewiththe[Received]timestamps,theCountyhasindicatedtwoeffortsarealreadybeingundertaken.ThefirstisaCADsoftwarepatchthatwillpreventthe[Received]timestampfrombeingoverwrittenwhenacalltakerrebidsacall.ThisisexpectedtooccurwithinthecurrentCADsystem.ThesecondeffortwillarguablyfixtheoverallissuebyclearlyassociatingphonerecordswiththeproperCADrecord.ThatfixiscurrentlybeingworkedonaspartofthenewCADsystemdeployment.

Law Enforcement P2/P3

ItshouldbenotedthatstakeholderswereunabletoidentifyanynationalstandardsorresearchthatsupportstheBrowardlawenforcementgoalsofhandling90%ofPriority1and2callswithin45secondsand90%Priority3callswithin90seconds.Thesecallsrepresentthemostimportanttimesensitiveincidentsforlawenforcement,generallyconsidered‘in-progress’typeincidents.Atpresent,theCountyisnotreportingonlawenforcementP2/P3performance–andalsonoteworthy,nopolicechiefinterviewedbyFITCHraisedtheissueorcomplainedaboutthelackofreporting.


AttheCounty’srequest,FITCHundertookanexaminationofP2/P3callprocessingtimeswithintheCityofPembrokePines.AccordingtothedataanalyzedbyBrowardCountyforNovemberandDecemberof2013,whilePembrokePineswasoperatingtheirowncenteronBrowardCounty’sCAD,performancewaswellbelowthetarget.ForPriority1and2calls,the45secondtargetwasmetlessthan1%ofthetime(0.79%)whilethePriority3callsmetthe90secondtargetonly12.78%ofthetime.WhenexaminingPriority1and2callsinPembrokePinesforNovemberandDecember2015,theperformanceincreasedtojust1.47%complianceforthe45secondtarget.Asnotedbelow,theperformancefortheentiresystemisconsistentlydismalcomparedtothebenchmarktarget.Figure48.LawP2/P3Statistics&Performance

Parameter Value

LAWallPriorities 136,595LAWallPrioritieswValidatedRcvdTimestamp 24,131PercentwithvalidatedRcvdandvalidP2/P3intervals 17.7%LAWwPriority1&2 10,030LAWwPriority1&2withvalidRcvdandvalidP2/P3intervals 5,244PercentagewithvalidatedRcvdandvalidP2/P3intervals 52.3%P2/P3ProcessingInterval Ranked50th%-tile 119.5sec Average 143.8sec Ranked90th%-tile 254.1sec Ranked95th%-tile 322.2sec

WithregardtolawenforcementP2/P3timesonly,therearenonationalrecommendationsonthismetric.Stakeholdersutilizedbestjudgment,butlackingdataappearedtohavemisjudgedthecapacityoftheSystemtoprocessthehighestlevelofcalls(Priority1and2)within45seconds.ThehistoricallawenforcementP2/P3performancereportedbytheCounty,andsimilarlydismalperformancecurrentlyfound,questiontheuseofa45-secondtarget.Infactonlyasinglestudycanbeidentifiedthatexamineslawenforcementcallprocessingtimes,publishedinlate2014.23Thatresearch,utilizingasimilarrankingschemataasinemergencymedicaldispatch(EMD),didnotevenconsidera45-secondbenchmark.Accordingly,thereisnobasistoassumea45-secondtargetisanachievableperformancestandardforlawenforcementcalls.

P3 Interval

ThedatafortheP3intervalisintheCAD,butitisnotbeingconsideredseparatelybytheBrowardsystemeventhoughitisavaliddatasource.P3whichfocusesonthattimefromwhenacalltakersendsanemergencyrequesttotheradiopositionuntiltheradiooperatordispatchestheappropriateunits.ThevalueofexaminingP3separatelyisthatitcanassistsystemmanagersinfocusingonwheretolook

23Warner,etal.(2014).CharacterizationofCallPrioritizationTimeinaPolicePriorityDispatch.AnnalsofEmergencyDispatch&Response2(2):17-23.


forfurtherimprovements.Asshownbelow,firecallstake11secondsonaverageand22secondsatthe90th%-tiletoprocesstheP3timeinterval.Therefore,withupto90secondstoachieveP2andP3,thisinformationindicatesthegreateropportunityexistsinmorecloselyexaminingtheP2componentofcallprocessingtimes.Figure49.P3PerformanceStatistics

DisciplineAssignmentInterval,P3 PreviousNFPA

Seconds@50th%-tile Seconds@90th%-tile Second@90th%-tile

FIRE 11.0sec 22.2sec 60secLAW 45.6sec 68.4sec notapp

BSOgenerallyappearstodowellonLAWassignmentswhenevaluatingP3alone,althoughasnotedabove,therearenoformalrecommendationsforthisinterval.Aswithfirerescuecalls,theintervalsusedtocalculatethe90thpercentileforLAWassignmentshavebeenfilteredtomodifylongdurationoutliersinthisdataset.Mostofthetime,operationsinthedispatchcenterareinsulatedfromeventsinthefield,butnotalways.Eventsinthefieldcan“backup”intothedispatchoperations.ThisisknowntobeacommonoccurrenceinLAWassignments.Policedepartmentstypicallyhavemorecallsinthequeuethanavailableunits,thusoneoftheresponsibilitiesofthedispatchcenteristoholdlowerprioritycallswhilehigherprioritycallsareattendedto.Addingtothecomplexityisthatcallsinpoliceenvironmentsarefromtwosources:thefirstisincoming911callsandthesecondisself-initiatedcalls(trafficstops).ThisrealitymeansthatthedispatcherhastobemorefluidforLAWcallassignmentsthanforFire/EMS.So,forexample,thenextrequestforservicehitsthependingscreenattheassignmentworkstation.Theassignmentdispatcherhasalotgoingoninthefieldanddeterminesthatthenewcallisoflesserpriority.Thedispatcherthendecidestoputthisnextrequestonhold.Whenfieldactivitydecreases,thedispatcherreturnstotheholdingrequestandexecutesitsassignment.However,theP2/P3clockkeepsrunningthroughoutthisprocess,leadingtoaninflatedP2/P3processingintervalthatdoesnotreflecttheamountoftimeactuallyconsumedprocessingtheincident.

FINDING:TheP1andP3intervalscanbeaccuratelyevaluatedbasedoncurrent

dataintheCADandtelephonysystems.BSOperformswellforthesedispatch

intervals.TheP2intervalmustbecautiouslyevaluatedduetotechnologyanddata

limitations.

P4 Interval

ThedatafortheP4interval–thattimefromwhentheradiodispatcheralertsemergencyresponders,typicallyinthefirestation,untiltheyareresponding–wasreportedintheCADdataprovidedtoFITCH.Summaryinformationonthismetricisprovidedbelow.


Figure50:ComparisonofP4Averagesand90thPercentiles(dataforNov-Dec2015)

Target

PerformanceCount Validated

RawCompliancesAverage

[sec]

90th%-tile

[sec]Actual

Count%

ChuteFIREResponse90%@80sec

6,620 6,620 3,051 46% 138.0sec 197.6sec

ChuteEMSResponse90%@60sec

37,102 37,102 13,787 43% 111.3sec 174.6sec

WhilenotpartoftheRegionalE911Systemperformance,itdoesimpactthecaller’sexperienceforpublicsafetyservice.Itisreportedheretoallowfordiscussionbystakeholders.


MODELING CURRENT DISPATCH OPERATIONS

Rationale

ThebenefitofamodelofdispatchoperationsisthatitpermitsFITCH,aswellasstakeholders,toposequestionsthatotherwisecouldnotbeaddressedintherealworld.Computertimeisinexpensivecomparedtoconductingthesameexperimentsusingtherealstreamofincomingcalls,actualdispatchersandrealPSAPs.Themodelbecomesacost-effectiveandtimelytoolforpredictingthebehavioroftherealsystem.Ofcourse,thelimitationtothisapproachisthevalidityofthemodel.Itmustbeemphasizedthattheperformanceofadispatchsystemhastwodistinctcomponentsthataresotightlyintertwinedthatitiseasytoconfusethedifference.Thefirstcomponentisthelengthoftimeittakestoexecuteeachfunctionofthedispatchprocess.WithinBSOdispatch,theP3intervalisanexampleofthiskindofcomponent.Thesecondcomponenttoperformanceishowlongittakesbeforeadispatchercanbeginexecutingthenextrequestinthequeue.WithinBSOdispatch,theP1intervalisanexampleofthiskindofcomponent.

Models of Dispatch Operations

APCO RETAINS

APCORETAINSisastaffingestimatorandretentionratecalculatorproducedbytheAssociationofPublic-SafetyCommunicationsOfficials(APCO).TheRETAINStitlestandforResponsiveEffortstoAssureIntegralNeedsinStaffing.TheestimatorisarespectedtoolforjudgingstaffingneedsTheoutstandingbenefitofAPCORETAINSisthatincanbeimplementedusingdatathatisreasonablyaccessible,oftenfromhardcopyreportsthatalreadyexistinthesystem.Fromthisstartingpoint,APCORETAINScanbeusedtocreatealowlevelmodelofdispatchoperations.Inputsaresuppliedtothemodelasbroadaveragesandestimates.Forinstance,countsofeventsareusedassurrogatesfortheactualdurationsrequiredtoprocesseachspecificevent.Theoutputsprovideaballparkestimateofthestaffingrequiredtomakethedispatchprocesswork.Thelimitationtotheutilityoftheseoutputsisthattheyaresilentontheperformancetobeexpectedfromthesystem.APCORETAINSprovidesnoguidancetodecisionmakersfacingfinancialandpolicyquestionsregardinghowmuchperformancewillchangewhenthestaffingbeingcommittedtothesystemchanges.


Erlang Analyses

AsnotedintheAPCORETAINSWorkbook:24Erlangformulasareconsideredthestandardforanyprocessthatrequiresanapplicationofqueuingtheory,suchasthenonlineararrivaltimesofincomingcallsinacallcenter.TheErlangformulasuseastatisticalsolutionthataddressestherandomnessofcallarrivaltime.

AgnerKrarupErlangwasaDanishmathematician,statistician,andengineerwhoinventedthefieldoftelephonenetworksanalysiswhileworkingfortheCopenhagenTelephoneCompanyfrom1908through1929.ThegoalofErlang’squeuinganalysesistodeterminehowmanyserviceprovidersshouldbemadeavailabletosatisfyusers,withoutoverprovisioning.Tomeetthisgoal,Mr.Erlanghadtoquantifythethree-corneredrelationshipbetweenrequestsforservice,numberofagents,andlatencyinthefigurebelow.Figure51.QueuingTheoryTriangle

TheconceptsandmathematicsintroducedbyMr.Erlanghavestoodthetestoftime.Inthemodernworld,thesemethodsareusedtoanalyzequeuingprocessesinsystemsasdiverseasshoppersusinggrocerystorecheckoutcashierstodatapacketswitchingthroughInternetroutersatmegahertzfrequencies.

Theassumptions,mathematics,andlimitationsofErlangqueuingtheory,asappliedtodispatchoperations,aretreatedingreaterdetailinAttachmentD,ErlangMathematics&Assumptions.

Work Stations

ThefirststepintheconstructionofanErlangmodeloftheBSOdispatchoperationsistoidentifywhichtypesofworkstationtoincludeinthemodel.ThisprocessisstraightforwardforBSOdispatchoperations.TherearethreePSAPlocations.AteachPSAPthereisagroupofintakeworkstations.Thisisthefirsttypeofworkstation.TherearealsoagroupofLAWassignmentandFIREassignmentworkstations.25Thesearethesecondandthirdtypesofworkstations.Theorganizationofworkflowsisdiagrammedbelow.

24APCORETAINSstaffingandRetentioninPublicSafetyCommunicationsCenters:EffectivePracticesGuideandStaffingWorkbook,page5,August2005.APCOInternational.25TheuseofFIREinthisreportreferstofiresuppressionandemergencymedicalservicesdispatchprocesses.

Requests

AgentsLatency


Figure52.WorkflowsandWorkstationsintheBSODispatchSystem

Theroundedrectanglesabove,representtypesofworkstations.Thetextinsideeachroundedrectanglesrepresentsthefunctionsthatareexecutedatthattypeofworkstation.“Intake”referstothefunctionofdeterminingwhatistheemergencyandwhereisitlocated.“A&P”referstoassessmentofacuityandprioritizationofresponse.“PreAr”referstoprearrivalinstructionsonemergencymedicalincidentswithEcho-Deltadeterminants(lifethreatening).“FIREAssignment”isthefunctionsofidentifyingasuitableunitfortheresponseandnotificationofthatunit.“RadioSupport”referstoradiocommunicationwithunitsinthefieldonincidentsin-progress.“LAWAssignment”isthefunctionsofidentifyingasuitableunitfortheresponseandnotificationofthatunit.“RadioSupport”referstoradiocommunicationwithunitsinthefieldonincidentsin-progress.TheconsultantsrecognizethattheLAWandFIREAssignmentworkstationsineachPSAParefurthersubdividedbyagencydispatched,atleastonthearchitecturalplotplansofthePSAPs.ThislevelofgranularitycouldnotbeincorporatedintothemodelingbecausedataavailableintheCADcontainsnoidentifiersofwhichassignmentworkstationisphysicallyresponsibleforeachincidentrecordintheCAD.

Waiting Queues

Theconfigurationofworkstationsthatisusedintheregional911centersisreferredtoasa“TripleQueue”model.Thisnomenclaturereferstothethreewaitingqueuesthatoccurinthemodel.Awaitingqueueexistswhenaclientrequestsservicefromadispatcherataworkstation.Thelengthofthewaitingqueueisameasureofhowlongittakesbeforethedispatcherattheworkstationisabletobeginservicingthisnextrequestforservice.TherearetwointernalandoneexternalwaitingqueuesintheTripleQueueModel.Whenanexternalclient(911caller)requestsservicefromtheIntakeWorkstation,thewaitingqueueinfrontoftheIntake

Pre-Arrival Instructions only on EMS incidents with Echo-Delta determinants1

LAW RadioSupportAssign

FIRE RadioSupportAssign

ClearRequest Intake + A&P + PreAr1







North PSAP

Central PSAP

South PSAP


workstationisreferredtoasthe“P1”intervalorthe“answerdelay”.WhentheIntakeWorkstationtransmitsanincidentrecordtotheAssignmentWorkstationforprocessing,thisactionisalsoarequestforservice,whereintheclienthasnowbecometheIntakeWorkstationsendingthecallfromthecalltakertotheradiodispatcher.ThewaitingqueueinfrontofeitherAssignmentWorkstationisreferredtoas“assignmentlatency”.

Definition of “Erlangs”

ThenextstepintheconstructionofanErlangmodeloftheBSOdispatchoperationsistoidentifytheworkloadsflowingacrosseachtypeofworkstation.Eachfunctionataworkstationrequiresadurationforitsexecution.Thesumofallthedurationsforallofthefunctionsbeingexecutedatatypeofworkstationistheworkloadflowingacrossthattypeofworkstation.Inqueuingtheory,workloadsaremeasuredin“Erlangs”.AnErlangissimplytheratioofthesummeddurationsofalltheactivitiesatatypeofworkstationperonehourontheclock.Inthemodelingthatfollows,bothErlangsandworkloadswillbeexpressedasdecimalhours.Forexample,aworkloaddurationof15minutes(00:15:00hh:mm:ss)willappearas0.250.Accountingforworkloadsmaysoundsimple,butinpractice,itisanextensivebookkeepingexercise.Thedurationsofalloftheactivitiesateachtypeofworkstationshavetobesummedforeachhour-of-dayforeachday-of-year.TheenormityofalltheseaccountingsisthebarrierthatpreventscasualusersfromattemptingErlanganalysesofqueuingprocessesincomplexsituationssuchasadispatchoperation.

Quantitation of Workloads

Primary Sources

AdetaileddiscussionofworkloadsatworkstationsispresentedinAttachmentE,QuantitationofWorkloads.ThedurationsoffunctionsbeingexecutedattheIntakeWorkstationswereobtainedfromtwoprimarysources:theCallDetailRecordsforincomingcallsfromtheIntradoVIPERtelephonyserverandthelogofoutgoingcalls.DataexportsfromthesesourceswereincorporatedintotheFITCHTelephonedatatableandtheOutgoingPhoneLogdatatable.AsamplerecordfromtheFITCHtelephonedatatableandfromtheOutgoingPhoneLogarepresentedinAttachmentF,SamplePhoneRecord.ThedurationsoffunctionsbeingexecutedattheAssignmentWorkstationswereobtainedfromtwoprimarysources:theexportofdatafieldsfromtheComputerAidedDispatchsystemandthe611-pagePDFyear-endsummaryreporttitled“TalkGroupsatZoneSummary150101–151231”.DatafromthesesourceswasincorporatedintoFITCH’sLAWIncidentdatatable,theFIREIncidentdatatable,andtheRadioTrafficdatatable.SamplerecordsfromthesedatatablesarepresentedinAttachmentG,SampleRecordsfromFireCAD,LAWCADandRadioStatistics.


Summation Database by Hour-of-Year

ThenextstepinthemodelingprocesswastocreatetwoSummationdatabases,eachonewith8,760records,onerecordforeachhouroftheyear.ThepurposeoftheSummationdatabaseswastoserveasarepositoryfordatathathadbeenaggregatedbyhour-of-year.Specialtyalgorithmswerewrittenthatqueriedtheprimarydatatablesbyhourofyear,fetchedandsummedtheactivitiesoccurringinthathour,andwrotetheresultsintothecorrespondingrecordforthatparticularhouroftheyearintheSummationdatabase.ASummationdatabasewaspreparedthataggregateddurationsfromthetelephoneCallDetailRecordsaswellastheOutgoingPhoneLog.AsamplerecordfromthisSummationdatabaseispresentedbelow.AseparateSummationdatabasewaspreparedthataggregateddurationsfromtheFIRE,LAW,andRadioTrafficdatatables.SamplerecordsfromtheSummationdatabasesarepresentedbelow.Figure53.RecordfromSummationDatabaseofTelephoneTraffic

Broward E911 Consolidated Communications SystemPhone Records by Hour of Year

68911 Trunks 7,246.5159ADM / AIM 7,313.6416Outgoing 633.91

143Totals 15,194.06

12/28/2015

Date

12 28 Mon 2 9 8,674

Mo Day DayName

Dayof Wk

Hourof Yr

Hr ofDay

Count Processing Σ [sec]Central

22911 Trunks 1,531.4153ADM / AIM 6,380.67

9Outgoing 1,428.5084Totals 9,340.58

North

33911 Trunks 2,680.2761ADM / AIM 7,674.9217Outgoing 1,132.05

111Totals 11,487.24

South

338Totals 36,021.88

BrowardCounty

Count Processing Σ [sec]



123 11,458.19911 Trunks

173 21,369.2342 3,194.46

ADM / AIM

Outgoing

1,378

RecordNumber

1.18

1.271.42

0.92

1.081.13

1.16

1.451.62

1.13

1.331.62

16Active Wkstat'n

10Active Wkstat'n

13Active Wkstat'n

39Active Wkstat'n

0.201.44

0.331.35

0.331.59

0.281.49

[sec]

Ranked 95th %-tile

Average

± Std Dev

911 Ans Delay

Predicted 90th %-tile

[sec]

Ranked 95th %-tile

Average

Ranked 90th %-tile

± Std Dev

911 Ans Delay


[sec]

Ranked 95th %-tile

Average

Ranked 90th %-tile

± Std Dev

911 Ans Delay


[sec]

Ranked 95th %-tile

Average

Ranked 90th %-tile

± Std Dev

911 Ans Delay


Ranked 90th %-tile


Figure54.RecordfromSummationDatabaseofCADandRadioTraffic

Averaged Databases by Hour-of-Day

ThenextstepwastocreatetwoAverageddatabases,eachcontaining24records,onerecordforeachhour-day.TheSummationdatabaseswerequeriedbyhour-of-day.Eachactivityinthathour-of-daywasaveragedoverallthedaysoftheyear,andtheresultswrittenintothecorrespondingrecordintheAverageddatabase.SamplesrecordsfromtheAverageddatabasesarepresentedbelow.

Broward E911 Consolidated Communications SystemCAD Records by Hour of Year

40Law 3,415

Fire

432.63 3,880.87

01/01/2015

Date

1 1 Thu 5 23 24

Mo Day DayName

Dayof Wk

Hourof Yr

Hr ofDay

CAD Count Assgn Workld[sec]

CentralPSAP

Xmit/Rcv'sCount

Radio Workld[sec]

11 100 117.84 999.42

19Law 757

Fire

217.70 1,982.38

CAD CountNorthPSAP

6 74 70.86 579.65

20Law 1,044

Fire

223.48 2,023.87

CAD CountSouthPSAP

7 68 83.52 732.99

79Law 5,215

Fire

873.81 7,887.12

CAD CountBrowardCounty

24 242 272.22 2,312.06

24

RecordNumber

Assgn Workld[sec]

Assgn Workld[sec]

Assgn Workld[sec]

Xmit/Rcv'sCount

Xmit/Rcv'sCount

Xmit/Rcv'sCount

Radio Workld[sec]

Radio Workld[sec]

Radio Workld[sec]


Figure55.RecordfromAveragedDatabaseofTelephoneTraffic

Broward E911 Consolidated Communications SystemCall Details by Hour of Day

12/28/2015

Date

12 28 Mon 2 9 8,674

Mo Day DayName

Dayof Wk

Hourof Yr

Hr ofDay

90.86 45.52 9,064.63 4,232.1235.49 18.93 4,293.97 2,494.0918.63 10.35 872.81 607.96144.97 68.88 14,231.41 6,301.04

911 TrunksADM / AIMOutgoing

Totals

Count ProcessingCentral

avg ± std dev avg ± std dev

29.74 15.78 2,906.29 1,438.2136.25 19.76 4,840.51 2,671.118.75 5.96 431.27 495.8674.74 36.03 8,178.06 3,720.07


Totals

Count ProcessingNorth


32.81 22.02 3,567.31 2,336.7432.33 22.23 4,311.92 3,022.0011.50 8.13 501.05 440.8576.63 46.61 8,380.28 5,125.83


Totals

Count ProcessingSouth


296.34 134.66 30,789.75 12,983.64Totals

Count Processing


BrowardCounty

153.40 69.31 15,538.23 6,082.26104.07 54.81 13,446.39 7,255.7138.87 19.67 1,805.13 1,044.79


15.58 4.29Intake WrkStn





Figure56.RecordfromAveragedDatabaseofCADandRadioTraffic

Broward E911 Consolidated Communications SystemCAD Records by Hour of Day

37.00Law

2,677.75

Fire

353.00 3,159.42

01/01/2015

Date

1 1 Thu 5 23 24

Mo Day DayName

Dayof Wk

Hourof Yr

Hr ofDay

CAD CountCentralPSAP

11.96 147.51 130.68 1,100.30

avg

± sd 8.84 824.76 94.64 844.43

3.67 73.72 40.23 338.91

avg

± sd

18.02Law

1,030.33

Fire

204.19 1,859.55

CAD CountNorthPSAP

5.40 78.74 64.05 517.07

avg

± sd

avg

± sd

5.77 478.77 65.05 592.38

2.59 55.53 30.73 247.98

28.82Law

1,498.91

Fire

266.46 2,629.78

CAD CountSouthPSAP

7.10 91.21 86.25 767.82

avg

± sd

avg

± sd

8.36 639.70 77.70 761.26

2.71 48.78 33.03 294.21

83.85Law

5,206.99

Fire

823.65 7,648.75

CAD Count

24.47 317.45 280.99 2,385.19

avg

± sd

avg

± sd

19.21 1,565.22 193.43 1,793.53

5.82 116.30 66.95 567.58

BrowardCounty

Assgn Workld[sec]

Xmit/Rcv'sCount

Radio Workld[sec]

Assgn Workld[sec]

Xmit/Rcv'sCount

Radio Workld[sec]

Assgn Workld[sec]

Xmit/Rcv'sCount

Radio Workld[sec]

Assgn Workld[sec]

Xmit/Rcv'sCount

Radio Workld[sec]


Erlang Modeling

Erlang Tables

AfterthetwoAveragedbyhour-of-daydatabaseshavebeenprepared,itisnecessarytocreatethefinaldatabasetorecordtheresultsoftheErlangmodels.IntheErlangdatabase,eachrecordisreferredtoasanErlangTableandrepresentsthehour-by-hourperformanceofasingletypeofworkstation.AcompletemodelofdispatchoperationsrequiresoneErlangTableforeachtypeofworkstationthatcomprisesthemodel.FromthewayBSOdispatchoperationsarecurrentlyorganized,acompletemodelwillrequirenineErlangTablesbecausethereareninetypeofworkstationspresent.EachErlangTableisanhour-by-hourcompilationtheeventcountsimpingingonaworkstationaswellasthesumofthedurationsrequiredtoprocesstheseevents.Thus,anErlangTableiscomprisedof24linesofdatafields,onelineperhour-of-day.TheeventcountsanddurationsareretrievedfromtheAverageddatabasesandwrittenbacktothecorrecthourintheErlangTable.ThesesamedataarealsopassedtoanalgorithmthatexecutestheErlangprobabilitycalculation.Thestatisticsfromthiscalculationarethenwrittentothecorrespondinghour.

Current BSO Staffing

InordertoconductanErlangmodelofcurrentoperations,itisnecessarytoknowhowmanydispatcherswereactivelyondutyattheirworkstationineachhourofthedayfortheperiodbeingmodeled.FITCHwasnotprovideddutyrostersthatcontainedthisdetailedlevelofdetailfor2015.TheworkaroundwastorelyonBSO’scallanalysisreportsfor2015,inwhichstaffingattheworkstationsattheCentral,North,andSouthPSAPswerereportedbyhour-of-day,onaverage.26

Model of Central Intake Workstation

ThefigurebelowpresentstheErlangTableforthemodeloftheIntakeWorkstationattheCentralPSAPasstaffedusingthehourlydeploymentofdispatchersasindicatedinthedocumentPSAPCALLANALYSISNOVEMBER2015.xls.

26BrowardCountyDocument:PSAPCallAnalysis,January2015.xlsthroughPSAPCallAnalysisDecember2015.xls.


Figure57.ErlangModelofCentralIntake

Thereareeightcolumnsinthefigureabove.Thecontentsofthesecolumnsisasfollows:

§ Column1presentsthehourofday.§ Column2talliestheaveragecountofcallscominginonthe911trunks.§ Column3talliestheaveragecountofcallscominginontheADMandAIMtrunks.§ Column4talliestheaveragecountofoutgoingcalls.


§ Column5talliesthetotalErlangsofworkloadflowingacrosstheCentralIntakeWorkstation.TheErlangsarethetotalofthedurationsrequiredtoprocessthe911incidents,theADMandAIMincidents,andtheoutgoingcalls.

§ Column6presentsthenumberofdispatchersactivelyondutyattheirworkstations(OnTask).§ Column7presentstheprobabilitythatthenextrequestforservicewillbeimmediatelyansweredby

adispatcher.§ Column8presentsthemaximumanswerdelayatthe95thpercentileexperiencedinthathourofday.

The7.10secondsappearingatthebottomofthecolumnofanswerdelaysisthe“weightedaverage”answerdelayforthewhole24hours.Theanswerdelayineachhour-of-dayisweightedbytheincidentcountinthathour,andtheweightedaverageforthewholedaycalculated.TheresultsofthisErlangmodelpresentacuriousresult.Theweightedaverageanswerdelayof7.10secondsatthe90thpercentileisextremelydivergentfromtheanswerdelayof1.42secondatthe90thpercentilethatwasobtainedbyreferringtothecallrecordsintheprimaryCDRdatatableandpresentedinFigure9(AnswerDelaysatCentralPSAPon11/07/2015)inpriorsectionsofthisreport.ThisdiscrepancyissoseverethattheconsultantsconcludethatBSO’sactualdeploymentofintakedispatchersissubstantiallyhigherthanindicatedinthedocumentPSAPCALLANALYSISNOVEMBER2015.xls.TheconsultantsemployedareverseErlanganalysistohoneinontheactualdeploymentofintakedispatchersusedbyBSO.Theapproachwastorunthemodelagain,retainingalloftheworkloadfromthehistoricrecord,butadjustingthedeploymentofdispatchersOnTaskuntiltheweightedaverageanswerdelaycamedownintotherangeof1.42seconds.Theresultofthisapproachispresentedbelow.


Figure58.ErlangModelCentralIntakeAdjusted

Broward E911 Consolidated Communications SystemWorkstation Performance by Hour-of-Day

ImmediateAnswer [ % ]

766665689

101111121212

13131412111010

9

13

94.3593.6696.6098.0098.9896.0697.7397.0196.3896.5897.0196.0497.4197.4697.07

98.5497.9298.2797.8197.6696.9998.0197.00

97.97

Central Intake workstations staffed so as to reproduce a 1.42 second weighted average answer delay.

Central Intake 1

63.7650.9640.5034.9231.5431.4042.0568.0584.9792.39

101.73

117.60124.32130.46132.68132.88138.77145.08125.20110.58102.88

90.1575.87

2.9782.4982.1141.8761.6331.6791.9293.0763.7494.2684.728

5.1835.1665.2925.5595.3165.5796.0125.0474.5354.1663.8853.613

00:0001:0002:0003:0004:0005:0006:0007:0008:0009:0010:0011:0012:0013:0014:0015:0016:0017:0018:0019:0020:0021:0022:0023:00

111.80 4.996


21.3718.0914.1212.6711.8811.5317.2429.5737.0243.1550.92

52.0152.5953.9757.7550.1550.6750.7742.3533.7829.7128.5528.37

53.43

97.25

Req'd HrsOnTask

13.4810.82

8.507.476.886.828.87

15.6619.3822.0124.38

24.8426.2627.5526.0227.2425.0724.2222.2221.3619.5417.2115.95

25.27

Triple PSAP2015Year Dispatch Model Workstation Name

90.86 3.95335.49 18.63

AverageErlangs

Hourly Averages

4.43

Ans Delay @

5.933.071.730.794.181.651.721.911.691.331.811.060.991.140.730.500.760.620.881.041.480.971.75

OnTask

Workstation Staffing & Performance

Weighted% Immed Ans

WeightedAns Delay

232 1.43

95 %-tile [sec]911 ADM Out

ADM Out911

0.00Surge

σ+Surge


Thedifferenceinstaffingbetweentheabovefiguresisstriking.InFigure57,BSOdeployed198dispatcherhoursOnTask.Figure58containsthenumberofdispatcherhourOnTaskrequiredtoreproducetheanswerdelaystakenfromthehistoricrecord.ThenumberofdispatcherhoursOnTaskincreasedfromtheoriginal198hoursto232hours,a17%increase.Theconsultanthasnorationalizationforwhythisoccurred.BSO’sdeploymentofdispatchersaboveneedstobeputintoperspective,asjudgedagainstotherhighperformingdispatchsystems.Achieving7.10secondsatthe90thpercentileweightedover24hoursatanintakeworkstationisaveryrespectablelevelofperformance.Addingadditionaldispatcherstotaketheanswerdelaydownto1.42secondatthe90thpercentilemustbeviewedasoverstaffing.

FINDING:BSOcurrentperformanceindicatesoverstaffinginCallTakerpositions

basedonErlangmodeling.

Model of Central FIRE Assignment Workstation

ThedocumentPSAPCALLANALYSISNOVEMBER2015.xlsalsospecifiedthestaffingattheCentralFIREAssignmentworkstation.InthiscaseBSOspecifiedaconstantlevelofstaffingas5dispatchersineachhourofday.TheErlangmodelforthisdeploymentofdispatcherattheCentralFIREAssignmentworkstationispresentedbelow.Theanswerdelaysexhibitedbythisdeploymentofdispatchersare0.00secondsacrosstheboard.Theseanswerdelaysarecompletely“offthecharts”.Inordertoplacethisdeploymentofdispatchersintoperspective,theconsultantsranasecondErlangmodeloftheCentralFIREAssignmentworkstationinwhichdeploymentsofdispatcherswereadjusteddownwardtobringanswerdelayintoarespectableandrealisticrange.Theresultofthismodelispresentedbelow.


Figure59.ErlangModelCentralFIREAssignmentWorkstationBSO



555555555555555

55555555

5

100.00100.00100.00100.00100.00100.00100.00100.00

99.9999.9899.9799.9799.9799.9799.97

99.9799.9799.9799.9899.9899.9999.99

100.00

99.97

Central FIRE Assignment workstations staffed to BSO specs as documented inPSAP CALL ANALYSIS NOVEMBER 2015.xls

Central Assign FIRE 6

9.968.848.057.446.937.328.69

11.8515.0017.1418.79

19.0118.6818.5519.1418.3718.7218.5716.9616.3115.4513.9311.96

0.2870.2550.2320.2140.1990.2110.2520.3420.4360.4970.547

0.5550.5460.5440.5680.5430.5550.5500.5020.4790.4540.4080.347

00:0001:0002:0003:0004:0005:0006:0007:0008:0009:0010:0011:0012:0013:0014:0015:0016:0017:0018:0019:0020:0021:0022:0023:00

18.73 0.546


99.98

Req'd HrsOnTask

108.5796.5687.7481.1575.5179.9994.39

129.07163.64186.62204.97

207.54203.84202.36208.99200.30204.37202.47185.08178.08168.45152.07130.68

204.29


14.35 0.4190.00 156.53

AverageErlangs

Hourly Averages

Hourly Averages

0.00

Ans Delay @

0.000.000.000.000.000.000.000.000.000.000.000.000.000.000.000.000.000.000.000.000.000.000.00

OnTask


Weighted% Immed Ans

WeightedAns Delay

120 0.00

95 %-tile [sec]FIRE LAW Radio

LAW RadioFIRE

0.00Surge

σ+Surge


Figure60.ErlangModelCentralFIREAssignmentsWorkstation



111111122222222

22222221

2

71.2874.5576.8078.6080.1578.8674.7895.2392.7790.9989.5089.5289.2589.5189.59

89.6289.2589.4190.8591.5292.2693.5365.34

88.85

Central FIRE Assignment workstations staffed to FITCH specifications

Central Assign FIRE 6

9.968.848.057.446.937.328.69

11.8515.0017.1418.79

19.0118.6818.5519.1418.3718.7218.5716.9616.3115.4513.9311.96

0.2870.2550.2320.2140.1990.2110.2520.3420.4360.4970.547

0.5550.5460.5440.5680.5430.5550.5500.5020.4790.4540.4080.347

00:0001:0002:0003:0004:0005:0006:0007:0008:0009:0010:0011:0012:0013:0014:0015:0016:0017:0018:0019:0020:0021:0022:0023:00

18.73 0.546


87.24

Req'd HrsOnTask

108.5796.5687.7481.1575.5179.9994.39

129.07163.64186.62204.97

207.54203.84202.36208.99200.30204.37202.47185.08178.08168.45152.07130.68

204.29


14.35 0.4190.00 156.53

AverageErlangs

Hourly Averages

Hourly Averages

10.22

Ans Delay @

8.637.666.886.246.798.630.731.181.531.851.851.901.851.842.031.851.941.901.591.441.291.05

13.48

OnTask


Weighted% Immed Ans

WeightedAns Delay

40 3.12

95 %-tile [sec]FIRE LAW Radio

LAW RadioFIRE

0.00Surge

σ+Surge


Achievinga3.12secondlatencyatthe95%weightedover24hoursisarespectablelevelofperformanceforanassignmentworkstation.Thesignificanttake-awayfromthismodelisthatarespectableleveloflatencycanbeattainedusingonly40dispatcherhoursOnTask.ThisisonethirddispatcherhoursthatBSOallocatestothisworkstation.Again,BSO’sallocationmustbeviewedasoverstaffing.

FINDING:BSO’scurrentperformanceindicatedoverstaffinginFIREAssignment

positionsbasedonErlangmodeling.


DISPATCH OPERATIONS MODELS — OPTIONS Allcommunicationscentersthatdispatchemergencyservicesmustexecutefourfunctionsintheconductoftheiroperations.Thesefunctionsarediagrammedbelow.Figure61.FunctionsRequiredInAllEmergencyServicesDispatchOperations

Intheabovediagram,“Intake”collectsthebasicinformation:Whatisthechiefcomplaint?Whereistheincidentlocated?“Processing”comprisesanAssessmentofAcuityandaPrioritizationofresponse,“A&P”.Theincidentisthentransferredtoan“Assignment”queuewhereunit(s)arechosentomaketheresponseandtheunit(s)notified.Thelastfunctioninthesequenceis“RadioSupport”forunitsrunningincidentsin-progress.Thedifferencebetweenvariousmodelsofdispatchoperationsboilsdowntowhichofthedispatchpersonnelexecutewhichofthefourfunctions.Inmanymodels,asingledispatchermaybetaskedwithexecutingmorethanoneofthesefunctions.PartofFITCH’scharterinthisprojectwastoproposechangestotheconductofdispatchoperationsthatmayleadtoimprovedefficiency.FITCH’sproposedmodelsofdispatchoperationswillbereferredtoasOption0,Option1,andOption2.Option0issonamedbecauseitincludesnochangestowayworkloadsandworkstationsarecurrentlystructuredinthePSAPs.ThequantitationoftheworkloadsusedinthefollowingmodelsisdescribedinAttachmentE,QuantitationofWorkloads.

Performance Targets

BoththeNationalEmergencyNumberAssociation,NENA,andtheNationalFireProtectionAssociation,NFPA,makerecommendationsconcerningtheconductofoperationsattheIntakeworkstations.Asdiscussedearlierinthisreport,theNENArecommendationisnotappropriatefortheBrowardsystembecauseitusesvariablestaffingbyhour-of-day.NENAisapplicableonlytosystemsusingconstantstaffingatallhouroftheday.ThisleavesNFPAastheapplicablerecommendation.NFPA1221,Section7.4.1,recommendsthattheanswerdelayattheIntakeworkstationsshouldnotexceed15secondsatthe95thpercentile.

0:00 T1 T2 T3 T8

Pickup; Chief Complaint

Assess Acuity; Set Priority

To Assignment Que

Unit Notified

Call Closed

ProcessingIntake Assignment Radio Support


FITCHconsidersNFPA1221tobean“idealistic”recommendationbecauseitisincompleteontwoimportantpoints.

§ First,NFPA1221doesnotclearlyspecifywhattimeintervalistobesampledtodetermine

compliancewiththe15second@95th%-tilerecommendation.§ Second,NFPA1221issilentonwhatshouldhappenatthedownstreamworkstationsthatcomprise

theremainderofthedispatchfunctions.

Inordertomodeldispatchoperations,FITCHdeveloped“OperationalTargets”thataddressedthedeficienciesofNFPA1221andareapplicableintherealworld.BrowardisauniquecaseinFITCH’sexperienceinthattheexistingsystemalreadyperformstoveryhighstandards.Infact,theexistingsystemcanbesaidto“overperform”andcanbesaidtobeoverprovisionedwithstaff.Howthiscameaboutisdiscussedinearliersectionsofthisreport.Thecurrentstaffinginthesystemissuchthatveryexpensiveincrementsofstaffinghaveledtoverysmalladditionalincrementsindispatchperformance,andhaveledtonodiscernibleimprovementtooutcomesinthefield.Giventhisuniqueenvironment,FITCHselectedoperationtargetsfortheBrowardsystemthat“dialback”dispatchperformancetoalevelthatisinbetterbalancewithfieldoutcomes.Thecurrentanswerdelayattheintakeworkstationsaveragedoveralltwenty-fourhoursofthedayis1.42secondsatthe95thpercentile.Bywayofcomparison,acrossNorthAmericaanintakeanswerdelayof15secondsatthe95thpercentileistakenasadistinguishingcharacteristicofahighperformingdispatchsystem.Withthesedataasbackground,FITCHdecidedto“dialback”theanswerdelayattheintakeworkstationsto6secondsatthe95thpercentile.Whilethisisslowerthancurrentperformance,itisstilltwoandhalftimesfasterthanthenationaltargetforahighperformingdispatchsystem.

Performance Targets for Modelling

Measurementsofperformanceattheworkstationsagainstthenewoperationaltargetsrequiresthreecalculationstepsandapplicationoftwocriteria:

IntakeWorkstations:

Answerdelaysarecalculatedforeachhour-of-dayatthe95thpercentile.Hourlyanswerdelaysareweightedbytheeventcountineachhour.Theweightedaverageanswerdelayiscalculatedoverthewhole24hours.FirstPerformanceCriteria

Weightedaverageanswerdelay(over24hours)<6seconds@95thpercentile.


SecondPerformanceCriteria

Answerdelayinanysinglehour<2X(6seconds@95th%-tile)

AssignmentWorkstations:

Latenciesarecalculatedforeachhour-of-dayatthe95thpercentile.Hourlylatenciesareweightedbytheeventcountineachhour.Theweightedaveragelatencyiscalculatedoverthewhole24hours.FirstPerformanceCriteria

Weightedaveragelatency(over24hours)<[email protected]

Latencyinanysinglehour<2X(6seconds@95th%-tile)

FRmsrRadioWorkstation

LatencyattheFRmsrRadioworkstationinOption2isaspecialcase.Thisworkstationhandlesradiosupportforincidentsrequiringmulti-stationresponses.Thisworkstationmaybeotherwisereferredtoasatacticalsupportworkstation.Theconsultantassumedthatthecrewswouldneednotacticalsupportwhiletheywereloadingtotheirapparatus.ChutetimeintheBrowardsystemonmulti-stationresponseincidentsaverages02minutes15seconds(02:15).Consequently,latencyattheFRmsrRadioworkstationwasdesignedtoaverage02:15orless.27

Forsystemsthatexhibitrandomdistributionsofresponsetimes(normaldistributions),2.907isthefactorthatconverts95thpercentilesto50thpercentilesandviceversa.Therefore,latencyof(6.00sec@95th%-tile)=Latency(2.06sec@50th%-tile).

Application to Dispatch Operational Model/Options

In Option 0 and Option 1

ThefirststepintheconstructionofdispatchmodelsistocollecttheaveragesworkloadsflowingacrosseachworkstationasdescribedinAttachmentE,QuantitationofWorkloads.FITCHthenincrementedtheseaverageworkloadsineveryhourofdaybythesurgeinthatparticularhourthathitsthesystemonedayoutoften.Surgesaremeasuredinunitsofstandarddeviationsrepresentedbythesymbol“σ”.ThemethodsusedtotreatsurgesinthisreportarepresentedinAttachmentH,CalculationofSurges.

27FRradiosupportfunctionshavebeenspecializedbasedonwhethertheincidentrequireasinglestationresponse(ssr),oramulti-stationresponse(msr).


Withthe+1.28σsurgeaddedtoeveryhour-of-day,thenumbersofdispatchersOnDutywasempiricallyadjustedoverthewhole24hoursuntilthecalculatedanswerdelaysorlatenciesconformedtothenewoperationaltargets.Thisnumberofdispatcherswasmaintainedandtheworkloadswerereturnedtotheiraveragevalues.ThetablesinAttachmentI,ErlangTablesofWorkstations,reflectoperationsofthesystemunderaverageconditionsofworkload,butwiththeenhancednumberofdispatcherHours-OnTask.Thisapproachtosurgecapacitywasacompromise,inanattempttodesignarobustdispatchsystemwithoutexcessiveoverprovisioningofdispatchers.Itmustbeemphasizedthata+1.28σsurgeineveryhour-of-day,backtoback,isaveryrareevent.Theincreaseinanswerdelaysandlatenciesongoingfromaverageworkload,0.00σ,toa+1.28σsurgeispresentedbelow.Figure62.ChangesinLatenciesfor+1.28σSurgesinAllHours-of-Day

Workstation

AnswerDelay&Latency

Weightedover24-Hours

[sec]@90th%-tile

0.00σ +1.28σ1

CentralIntake 2.58 5.73NorthIntake 2.12 5.37SouthIntake 2.03 9.35ConsolidatedIntake 5.32 6.00CentralLAWAssign&[email protected] 3.22 6.48CentralFIREAssign&[email protected] 2.40 4.66NorthLAWAssign&[email protected] 2.69 6.09NorthFIREAssign&[email protected] 2.67 5.18SouthLAWAssign&[email protected] 2.43 5.55SouthFIREAssign&[email protected] 1.95 4.01ConsolidatedFIREGatekeeper 1.64 [email protected] 136.06 [email protected] 1.35 2.08

1Asurgeof+1.28σwasappliedatall24hoursoftheday.

In Option 2

TheanswerdelaysandlatenciesforallworkstationsinOption2conformtotherestrictionsdescribedforOption0andOption1,excepttheFIREtacRadioworkstation.ThespecializedfunctionoftheFIRETacRadioworkstationpermitslatencyatthisworkstationtobemultipleminuteswithoutanynegativeimpactondispatchoperations.ThefunctionoftheFIREtacdispatcheristoexclusivelymonitorafirerelatedincidentin-progressandtointerveneasneeded.TheFiretacdispatcherisalertedatthesametimetherespondingunitsarealerted.IftheFIREtacdispatcherisdelayedforseveralminutesbeforetheycanbeginpayingexclusiveattentiontotheincident,itdoesnotmatter.TheaverageintervalrequiredintheBrowardsystemforfirecrewstoloadtotheirapparatusis00:02:19[hh:mm:ss].Afirecrewdoesnotneedtacticalradiosupportwhentheyarestillinquartersandhavenotstartedrollingenroute.


ThedispatcherHours-OnTaskattheFIREtacRadioworkstationwereadjustedsothatthelatencyreflectedthisreality.

Impact of Surges

Thedatabelowwasassembledtoillustratetheimpactofsurgesinthesystem.Thistablereportshow95thpercentileanswerdelaysattheCentralPSAPIntakeworkstationsat1800hourschangeinresponsetochangesinsurgesaswellasdispatchersOnDuty.Threelevelsofsurgesareincludedinthetable:surgesthatappearonedayinten;onedayinthirty;andonedayinonehundred.Theincidentcountsassociatedwiththesesurgesareincludedinthetable.Figure63.DependenceofAnswerDelaysonSurgesandDispatchersOnDutyforCentralIntakeat1800Hours

Surge

0.00σ +1.28σ +1.84σ +2.33σ

Incidents/hr

145 198 222 242

7 40.53 ----- ----- -----8 17.20 ----- ----- -----9 9.79 36.62 ----- -----

10 6.18 15.09 30.55 245.6211 3.98 8.79 13.28 23.1612 2.44 5.82 7.92 11.1213 1.34 4.09 5.35 6.9014 0.62 2.92 3.84 4.7315 0.25 2.03 2.82 3.49

DispatchersOnDuty

AnswerDelays[sec@95th%-tile]

Thegreenshadedcellsareacceptableanswerdelays.Thedataaboveillustratesthenon-linearrelationshipbetweendispatchersandcallvolume.WhenaoneinonehundredsurgecomesthroughtheIntakeworkstations,theincidentcountjumpsfrom145incidents/hourto242incidents/hour,a67%increase.However,therequirednumberofdispatchersincreasefrom10to13,a30%increase.TherelationshipbetweendispatchersOnDutyandworkloadisexplainedindetailinAttachmentD,ErlangMathematics&Assumptions.Dependingontheoperatingpointofthespecificsystem,therelationshipbetweenchangesindispatcherandchangesinworkloadmaybemagnifiedorsuppressed.Browardisfortunateinthattheoperatingpointoftheintakefunctionisintheregimewherethesystemcanabsorblargechangesinworkloadwithmuchsmallerchangesintherequirednumberofdispatchers.

Dispatch Model Option 0

Thefigurebelow,presentsadiagramoftheflowofworkloadsthroughtheOption0dispatchmodel.


Figure64.DispatchModelOption0

Inthefigureabove,theredovalrepresentsarequestforservicefromthefieldafteracallerhasdetectedanemergencysituation.Asingleroundedrectanglerepresentsasingletypeofworkstation.Theremaybeseveralworkstationsofasingletype,allexecutingthesamefunctions,butinparallel.Thetextinsideeachroundedrectanglerepresentsthefunctionsbeingexecutedatthattypeofworkstation.Thegreenovalrepresentsthefieldrespondersreturningtoastateofnormalcy.Thefunctioningofeachtypeofworkstationischaracterizedbytwoparameters,workloadandlatency.Workloadisthefirstparameterthatcharacterizesaworkstation.Workloadistheamountoftimethedispatcherspendsactivelyexecutingthefunctionsofthatworkstation.ForpurposesofmodellingworkstationsintheBrowarddispatchcenter,workloadswereextractedfromthehistoricrecord.Latencyisthesecondparameterthatcharacterizesaworkstation.Everytimeaclientrequestsservicefromaworkstationthereisthepossibilitytheexecutionsofthatrequestwillbedelayedbyaneedtofinishprocessingapreviousrequestforservice.TherearetwotypesofclientsintheBrowardsystem.ThemosteasilyidentifiedistheexternalclientrequestingservicefromtheIntakeworkstation.LatencyaffectingthistransactionisspecificallyreferredtoastheAnswerDelayortheP1Interval.Therearealsointernalclients.WhenanincidentistransferredfromtheIntakeworkstationtotheAssignmentworkstation,theIntakeworkstationbecomestheinternalclientandtheAssignmentworkstationbecomestheagent.Latencyalsoaffectsthisinternaltransaction.











North PSAP

Central PSAP

South PSAP


ThestructureofworkflowsinDispatchModelOption0exactlyreflectscurrentoperationoftheBrowardE911ConsolidatedCommunicationsSystem.Thedurationsfortheexecutionofdispatchfunctionsweretakenfromthehistoricrecord.ThedistinctionbetweenOption0andcurrentoperationsisthatOption0ismodelledusingnumbersofdispatchersthatresultinlatenciesthatconformtonewOperationalTargets.TheErlangtablesthatcompriseOption0arepresentedinAttachmentI,ErlangTablesofWorkstations.DatafromthedetailedErlangTablesissummarizedbelow,representingNorth,CentralandSouthPSAPs.Figure65.North,CentralandSouthPerformanceandDispatcherHours-OnDutyforDispatchModelOption0

Workstations Functions

DispatchersImmed

Answer

WeightedAverage

AnswerDelay

[sec]@XXth%-tileMin/

Max

Hours

OnTask

NorthPSAPIntake

CallIntake,A&P1,&PreArv2

4/7 139 94.16%5.03sec@95th

1.73sec@50thNorthPSAPFIREAssign

FIREAssignFIRERadioSupport

1/2 34 90.39%2.67sec@95th

0.92sec@50thNorthPSAPLAWAssign

LAWAssignLAWRadioSupport

2/4 65 88.77%2.69sec@95th

0.92sec@50thNorthHours-OnTask 238


Dispatchers Immed

Answer

WeightedAverage

AnswerDelay

[sec]@XXth%-tile

Min/

Max

Hours

OnTask

CentralPSAPIntake


5/11 204 93.75%3.96sec@95th

1.26sec@50thCentralPSAPFIREAssign


1/2 42 89.00%2.40sec@95th1

0.83sec@50thCentralPSAPLAWAssign


2/5 92 85.65%3.37sec@95th1

1.16sec@50thCentralHours-OnTask 338


Dispatchers Immed

Answer

WeightedAverage

AnswerDelay

[sec]@XXth%-tile

Min/

Max

Hours

OnTask

SouthPSAPIntake


4/8 150 95.86%3.19sec@95th

1.10sec@50thSouthPSAPFIREAssign


1/2 40 91.76%1.95sec@95th

0.67sec@50thSouthPSAPLAWAssign


2/4 78 88.02%2.56sec@95th

0.88sec@50thSouthHours-OnTask 268 Option0Hours-OnTask 844

1Assessmentofacuity&Prioritizationofresponse.2Pre-arrivalinstructiononEcho,DeltaEMSincidents.


Dispatch Operations Model/Option 1

Thefigurebelow,presentsadiagramoftheflowofworkloadsthroughtheOption0dispatchmodel.Figure66.DispatchOperationsModel/Option1

ThedistinctionbetweenOption0andOption1isthatthereisacompleterolloverofincomingcallsbetweentheIntakeworkstationsregardlessofwhichPSAPtheintakeworkstationmaybephysicallylocatedin.TheorganizationofdispatchfunctionsinOption1maintainsthedistinctionsofNorth,Central,andSouthFIREandLAWAssignmentworkstations.TheBrowardE911CommunicationCenterisintheprocessofbeginningatransitiontowardsimplementingthismodelofdispatchoperations.TheErlangTablesthatcompriseOption1arepresentedinAttachmentI,ErlangTablesofWorkstations.DatafromthedetailedErlangTablesissummarizedinTable53,below,representingNorth,CentralandSouthPSAPs.




Clear


RadioSupportAssign

ClearIntake + A&P + PreAr1


RadioSupportAssign

Clear

North

Central

South

FIRE

FIRE

Request

Request

Request

North

Central

South


Figure67.PerformanceandDispatcherHours-OnDutyinDispatchModelOption1.


DispatchersImmed

Answer

WeightedAverage

AnswerDelay

[sec]@XXth%-tile

Min/

Max

Hours

OnTask

ConsolidatedIntake


7/17 301 93.18%5.32sec@95th

1.83sec@50thIntakeHours-OnTask 301


DispatchersImmed

Answer

WeightedAverage

AnswerDelay


Max

Hours

OnTask

NorthPSAPFIREAssign


1/2 34 90.39%2.67sec@95th

0.92sec@50thNorthPSAPLAWAssign


2/4 65 88.77%2.69sec@95th

0.92sec@50thNorthHours-OnTask 99


Dispatchers Immed

Answer

WeightedAverage

AnswerDelay


Max

Hours

OnTask

CentralPSAPFIREAssign


1/2 42 89.00%2.40sec@95th

0.83sec@50thCentralPSAPLAWAssign


2/5 92 85.65%3.37sec@95th1

1.16sec@50thCentralHours-OnTask 134


Dispatchers Immed

Answer

WeightedAverage

AnswerDelay


Max

Hours

OnTask

SouthPSAPFIREAssign


1/2 40 91.76%1.95sec@95th

0.67sec@50thSouthPSAPLAWAssign


2/4 78 88.02%2.56sec@95th1

0.88sec@50thSouthHours-OnTask 118 Option1Hours-OnTask 652

1Assessmentofacuity&Prioritizationofresponse.2Pre-arrivalinstructiononEcho,DeltaEMSincidents.ComparingOption1toOption0demonstratesthatconsolidatingadispatchfunctionusesmanpowermoreefficiently.UnderOption0,conductoftheIntakefunctionsrequired493dispatcherhours-OnTaskbetweenthethreeseparatePSAPlocations.UnderOption1,conductoftheIntakefunctionsrequiresonly301dispatcherHours-OnTaskwhenincomingcallsautomaticallyrolloverbetweenthePSAPlocations.UnderOption1,theperformanceoftheIntakefunctionsisindistinguishablefromOption0,yettheIntakefunctionsarebeingexecutedwith192fewerdispatcherHours-OnTask,a39%reduction.


Dispatch Operations Model/Option 2

Thefigurebelow,presentsadiagramoftheflowofworkloadsthroughtheOption2dispatchmodel.Figure68.DispatchModelOption2

ThedistinctionsbetweenOption1andOption2arethreefold.First,theCADsoftwareismorefullyutilizedtoautomaticallyrecommendunitstoFireRescueresponses.Beforeunitsarenotified,theseautomaticassignmentsaresubjecttoamanualreview.Ineffect,thisreviewisagatekeeperfunctionthatlimitsthenumberofassignmentpositionswithinthedispatchcenter.Thislargelyeliminatestheneedfordisparateassignmentdispatchersto‘shop’forresourcestoassign.Second,theuseofmobiledataterminals(MDTs)byFIREfieldresponderswasincreasedovercurrentpractice.TheconsequenceoftheincreaseduseofMDTsispresumedtobea40%reductioninthenumberofXmit/Rcveventsperincident.Third,theFRradiosupportfunctionshavebeenspecializedbasedonwhethertheincidentrequireasinglestationresponse(ssr),oramulti-stationresponse(msr).FRtacticalsupportandFRemssupportbothhavededicatedradiochannelsandspecializeddispatchers.ConductofoperationsalongtheLAWassignmentlegshasnotbeenaltered.TheErlangTablesthatcompriseOption2arepresentedinAttachmentI,ErlangTablesofWorkstations.DatafromthedetailedErlangTablesissummarizedbelow,representingNorth,CentralandSouthPSAPs.


Request

RadioSupport

LAWAssign

FR ssrRadio

FR msrRadio

FRAssgn


Request

North

Central

South

IDUnit


Figure69.PerformanceandDispatcherHours-OnDutyinDispatchModelOption2.


DispatchersImmed

Answer

WeightedAverage

AnswerDelay

[sec]@XXth%-tile

Min/

Max

Hours

OnTask

ConsolidatedIntake


7/17 301 93.18%5.32sec@95th

1.83sec@50thFRReview(Gatekeeper)

FRAssignReview 1/2 38 96.41%1.64sec@95th

0.57sec@50th

FRmsrRadio*FRTacticalRadioSupport

2/4 67 89.33%[email protected]@50th

FRssrRadio*FREMSRadioSupport

2/2 48 91.99%2.08sec@95th

0.46sec@50th

NorthLAWNorthLAWAssignNRadioSupport

2/4 65 90.95%2.69sec@95th

0.92sec@50th

CentralLAWCentralLAWAssignCentralRadioSupport

2/5 92 85.65%3.37sec@95th1

1.16sec@50th

SouthLAWSouthLAWAssignSouthRadioSupport

2/4 78 88.02%2.56sec@95th1

0.88sec@50thOption2Hours-OnTask 700

*FRmsrRadio=multi-stationresponse;FRssrRadio=singlestationresponse

ComparingOption1toOption2,thereisanetincreaseindispatcherHours-OnTaskfrom652to700.Thisisanetincreaseof48dispatcherHours-OnTask.TheincreaseisrestrictedtotheFIREAssignmentandRadioSupportlegbetweenthetwoOptions.ComparingtheFIREAssignmentandRadioSupportlegbetweenOption1andOption2,dispatcherHours-OnTaskincreasefrom116to164.ConsolidatingIntakefunctionsbetweenOption0andOption1resultsinareductionintherequireddispatcherHours-OnTask.TheinverseaffectisseenbetweenOption1andOption2,wheretheradiosupportfunctionsalongtheFIRElegweredividedintosub-specialties.


Comparison of Operations Models/Options

ThenumberofdispatcherHours-OnTaskcommittedtoeachOptionwasadjustedsoIntakeanswerdelays,P1,andassignmentlatenciesconformedtonewoperationaltargets.ThesetargetsprovideforacapacitytoabsorbsurgesindemandsizedtoreflectactualexperiencetakenfromBroward’shistoricrecord.ThefigurebelowsummarizestherequirementsfordispatcherOnDutyforOption0,Option1,andOption2.Figure70.ComparisonofOperationsModels/Options

DispatchModel

Dispatchers

Min/

Max

Hours

OnTask

Option0:CurrentOpswithFITCHperformancetargets 22/45 844Option1:ConsolidatedIntake 13/36 652Option2:Option1withFRGatekeeper,FRmsrRadio,andFRssrRadiochannels 14/38 700

Note:TheexactdistributionsofdispatcherHours-OnTaskbyhourofdayandworkstationarefoundinAttachmentI,ErlangTablesofWorkstations.FITCHbelievesthattheincreaseindispatcherHours-OnTaskinOption2isjustifiable.Thepurposeofadispatchcenteristofacilitatetheexecutionofemergencyserviceresponsesinthefield.HavingdedicatedcommunicationchannelsandspecializeddispatcherswillimprovetheperformanceandincreasethesafetyofFIREandEMSfieldresponders.FITCHbelievesthatthesebenefitsoutweightheadditionaldispatchcosts.


RECOMMENDATIONS

Technology

Findings —

TheCounty’sPSAPphonesystemandCADsystemsarenoteffectivelylinkedtoallowcomprehensiveevaluationofSystemperformance.Formorethanhalfoftheincidentrecords,theeventintheCADcannotbelinkedtotheuniqueCallDetailRecord(CDR)thatinitiatedtheincident.Technologylimitationsresultedinonly25.6%ofCADrecordsconsideredvalidforuseinanalysisofP2/P3intervals.Countystaffisunabletodirectlyaccessphoneandradiosystemdata–therebylimitingtheirabilitytoanalyzesystemperformancebeyondthatpermittedbypre-designed/cannedreports,whichmakessomeoftherequiredreportingtediousanderrorprone.TheSystemutilizesemergencymedicaldispatching(EMD)software–abestpracticefor911centers.However,nosimilarprogramisutilizedforeitherfireorlawenforcementcalltypes.TheCADnetworkisredundantintheeventofafailure.However,itisnottestedonaregularbasis.Thisisasignificantdeficiencyandisinconflictwithbestpractices.

Recommendations —

TheCountyneedstoinsurethemissionsoftechnologydevelopmentandtechnologysustainmenthavedifferentfocusesandroles.Therefore,thecountyshouldprovideforaTechnologyDevelopmentTeamandaTechnologySustainmentTeamoverthenextfewyearsasnewtechnologiesareimplementedandthesystemcontinuestostabilize.AnabsolutepriorityfortheCountyistodevelopalinkbetween911phonerecordsandtheassociateCADincidentrecords.BSOshouldmaintainEMDcertificationtrainingforallcalltakersthroughtheInternationalAcademiesofEmergencyDispatch(IAED).CalltakerpersonnelshouldalsobetrainedandcertifiedasEmergencyFireDispatchers(EFD)andinthesystemimplementEFDinthenearfuture.Thesecertificationsareconsideredindustrybestpractice.Finally,lawenforcementagenciesshouldconsiderandevaluatetheefficacyofEmergencyPoliceDispatch(EPD)beingutilizedinthefuture.Thissystemisemergingasanindustrybestpractice.


Discussion —

Challengesinobtainingdataduringthisproject,timelyrecognitionoflimitationsintechnology,andotherproblemsindicateresourceconstraintswithinORCATtosupportthesignificanttechnologyinfrastructurerequiredforsystemofthesize.Ataminimum,theCountyneedstoensureseparateorganizationalfocusonthosepersonneldedicatedtodeveloping/enhancingtechnologywithinthesystemagainstthoseresponsibleformaintaining/sustainingtheexistingtechnology.TheCountyshouldreorganizeORCATtoensuretheseresponsibilitiesarebifurcated.WenotethatintheFY15/16theCountyfundedtwoadditionalpositionstoprovidemaintenancesupporttotheirpublicsafetyapplications,andintheFY16/17budget,theCountyfundedsixadditionalpositionsintheOfficeofRegionalCommunicationsandTechnologytosupportthefollowing:thePublicSafetyNetwork,ongoingcapitalprojects,theRadioSystem,andPublicSafetyApplications.Theremaystillbearequirementformoreresourcesspecifictocreateaseparate911technologydevelopmentgroup.Theopportunitytorecaptureresourceselsewhereinthesystemallowsforthisrecommendationtobeaccomplishedwithoutanyadditionaloverallfunding.TheCountyhas,andis,expendingsignificantresourcestoupgradeRegionalE911Systemtechnologies.ThephonesystemwasrecentlycompletedandmajorupgradestotheradioandCADsystemsarecurrentlyunderway.However,anumberofchallengeswereencounteredintheharvestingofdata.Thefindingsregardingtechnologylimitationshighlighttheneedtoaddresssomefundamentaltechnologyissuesasthesesystemsarenowundergoingmajorupgrades.Generally,stakeholdersdonotappreciatehowtheseissuesimpacttheabilitytoeffectivelymanagetheSystem.AmajorflawofthecurrentsystemistheinabilitytolinkphonerecordstoCADrecordsandestablishaseamlessstart-to-finishtimelineforanincident.ThisresultsintheCountybasingoverallSystemperformancewithoutbenefitofallthefire/medicaldispatchrecords.ThistechnologydeficitsignificantlylimitstheabilitytocalculatetheP2/P3callprocessingintervals.Asnoted,theSystemcannotreliablyanswerthefundamentalquestionofhowlongittakesbetweenwhenacallismadeto911andwhenhelparrives.Withtheimplementationofthenewnextgenerationcomputeraideddispatch(CAD)systemanticipatedforearly2017,theCountyshouldensurethatthisflawisresolved.SincethereleaseofthePhase1report,Countystaffhasworkedtoidentifytechnologychanges/upgradesthatwillrectifytheproblem.TheCountyandMotorolaindicatetheyhaveaworkingsolutionthatwillbedeployedwiththenewCAD.Thegoalshouldbetolinkatleast90%ofCADrecordswith911recordswheretheyexist.Inaddition,theCountyisunabletocurrentlyaccessradioandphonedatadirectly.ToascribeperformanceevaluationstotheentireSystembasedonpartialandpotentiallystatisticallybiaseddataisquestionable.FITCHtookextraordinaryefforttoconstructdatatablesfromthesetwodatasourcesinordertoassessthesystem.Pass/failassessmentsshouldbecautiouslyweighedbydecision-makersuntilallplannedtechnologyimprovementsareinplace.Countystaffshouldcontinuetoreportonthetrend-datatoestablishbaselineperformance.Recently,theCountyhasindicatedthatthenewCADsystemwillcontainaccesstothenecessarydatafromthe911database.ThisinformationwillbereadilyaccessibletoCountystafftoallowformoredetailed/adhocreporting.TheCountyshouldadditionallyensurethesamecapabilityexistintheradiosystem.


BSOcurrentlyutilizestheMedicalPriorityDispatchSystem(MPDS)fordispatchingemergencymedicalcallsandprovidingpre-arrivalinstructionstocallers.AsanagencyaccreditedbyIAED,allofBSO’sdispatchersmaintainemergencymedicaldispatcher(EMD)certification.SimilartoEMD,TheCounty,BSOandothersystemparticipantsshouldadoptanddeployEmergencyFireDispatch(EFD)inthenearfuture.Utilizationofthesamevendor,employingsimilarinterfacesandprogramlogic,willallowforthisrecommendationtobeaccomplishedwithrelativeease.SincebothdispatchpersonnelandfieldpersonnelalreadyutilizeEMD,theadditionofEFDshouldbewellaccepted.Currently,thereissignificantlylessuniformityintheprocessingoflawenforcementcallsforservice.Thisisquitetypicalacrossthenation.However,increasingnumberoflawenforcementagenciesareemployingsimilarsoftwareproductsasidentifiedabove.Toprovideforgreateruniformityinhandlinglawenforcementcallsforservice,providedefensibleandobjectiveguidelinesforperformance,withqualityassuranceprocessesthatcanbeappliedandreviewedbymanagementpersonnel,lawenforcementagenciesalongwithBSOshouldundertakeanevaluationofproductssimilartoEmergencyPoliceDispatch(EPD).BecauseofthefundamentalwayinwhichEPDintegratesintothedispatchprocess,thedecisiontoutilizesuchatoolwouldrequirealllawenforcementagenciestoactivelyparticipateinthisevaluationanddecision.

Operational Oversight and System Governance

Findings —

BSO’soperationofthePublicSafetyAnsweringPoints(PSAPs)arechallengedwithsignificantmoraleproblemsembeddedinissuesofstaffing,trainingandmanagement.TheCountyhasinappropriatelymade,andpublicsafetyofficialsallowed,someoperationaldecisionstobehandledbytheCountythatshould,instead,bedeterminedbypublicsafetyofficials.Lowlevelsoftrustexistamongmajorstakeholders.Muchofthisisduetoroledefinitions.RelationshipsneedtoberedefinedinorderfortheSystemtomoveforwardeffectively.

Recommendations —

OperationalOversightandSystemGovernanceshouldberedefinedtostrengthentheroleofend-userswhilebalancingthelogisticalconcernsoftheOperator(BSO),andthefinancialandsystemgovernanceresponsibilitiesofBrowardCounty.Alternativeworkschedulesareavailableandshouldbeconsidered.AttachmentA,SchedulingMatrixSample,providessampleschedulesforconsideration.Fillingvacantpositionsinatimelymannerwiththegoalofmaintainingfullstaffingwillreduceexcessivemandatoryovertimeandtheassociatedstress.Thiswillallowresourcestoalignmorecloselytodemandpatterns,therebyimprovingefficiencyinthesystem.


SupervisiononthePSAPdispatchfloorsshouldbeataratioofsixtooneasopposedtothecurrentninetooneratio.GreaterqualityassuranceprocessesaretobehandledbyBSOdispatchfloorsupervisors.Resourcesfordispatchertrainingshouldbeincreasedthroughreallocationofcurrentfunding.A“baselevelof911services”fundedbytheCountyshouldbemoreclearlydefinedbyutilizingthecurrentinterlocalagreementsandFITCH’smodelingofperformancelevelsasnotedincalltakingandradiopositions.Individualagenciesdesiringhigherlevelsofserviceshouldbeabletofundadditionalstaffinghoursortechnologyinordertoreceiveservicesspecifictotheirjurisdictionalneeds.TheRegionalSystem’smanagementandtechnologyshouldfacilitatetheseadditionalservicesaslongastheydonotdisruptthebaseservices.

Discussion —

AsapprovedbytheCounty,BSOandmunicipalities,theSystem’sinitial,rapidimplementationtimeframerequiredamorecentralizedoversight/governanceprocess.Intheconsolidationprocess,somecommunitieswereabletoaddservicesthatwerenotprovidedindividuallybefore.Forexample,theconsolidatedSystemutilizesemergencymedicaldispatching(EMD)services–abestpracticefor911centers.ThefocusduringtheseinitialmonthswaswiththeCounty’sOfficeofRegionalCommunicationsandTechnology(ORCAT).Thisapproach,whilearguablyneededduringearlyimplementation,doesnotservetheongoingneedsofotherstakeholders.SomeexamplesoftheCounty’sassumptionofoperationalissuesresultedfromroleambiguity.Andwhilecurrentperceptionsindicatethereisalackoftrustamongstakeholders,thereisalsoevidencethatinotherregardstheSystemhasnow“turnedthecorner”.Futuresystemimprovementswillbenefitfromaredefined,collaborative,andsimplifiedgovernancestructure.Thechallengeformunicipalleaders–fireandpolicechiefs,alongwithlocallyelectedleaders–willbedefiningaclearsetofexpectationssharedbyall.Localofficials,especiallycitymanagers,policechiefsandfirechiefs,mustensuretheircommunitiesinterestsarerepresentedbyactiveengagementwithinanewgovernancestructurethatadjusttheperceivedbalanceofcontrolthathasexistedtothispoint.Whatisrequiredisanoversightprocessthatbalancesend-userconcernsforfieldoperations,againstBSO’srequirementstomanagethe911SystemoperationsandBrowardCounty’sfiduciaryandlegislativeresponsibilities–allwhileensuringtransparencythatallstakeholdersrequire.Therearetwogeneralapproachestomanaging911systems.Thefirstfocusesoncontrol,whichistypicallyfoundwhenonlyasingleagencyistheend-user.Inthiscase,thefocusisoncommandandcontroloffieldresources,dueprimarilytolimitedresourcesrelevanttodemandforservices.Thesecondapproachfocusesonsupportanditservesbothsingleandmulti-agency911systemswell.Thegoalofthisapproachistosupportfieldpersonnel,remainresponsivetofieldneeds,andberegardedasacriticalcomponentofthepublicsafetysystem.BrowardCountyhas,andshouldremainwith,a‘supportapproach’intheirRegional911System.


TheCounty’sobligations,andresponsibilities,areclear.Astheaccountableentity,itultimatelyhasbothlegislativeandfiscalresponsibilitiestothesystem.However,consistentwithasupportphilosophy,theCountydoesnotneedtonecessarily‘drive’thesystem,butrathershouldadoptaperspectiveoffacilitatingandultimatelymanagingtheSystem’sperformance.Thefollowingdiscussionprovidesspecificrecommendationsforfutureoperationaloversightandsystemgovernance.TheRegional911Systemmustservelawenforcementandfirerescueprovidersandtoachievethatgoal,fieldoperationaloversightmustfocusonthosestakeholders.However,stakeholder’sneedsmustalsobebalancedwiththelogisticalconstraintsofBSO,asOperator,andthefinancialandgovernanceconcernsofBrowardCounty.Duringthefirst18monthsofoperation,theSystemwasseenbystakeholdersastoocentralizedwithintheCountyOfficeofRegionalCommunicationsandTechnology(ORCAT)andtoofocusedoncontractualperformancemetrics.Therequirementsofend-userswerenotbeingmet.Operationaloversight,asusedherein,referstowatchful,responsibleandaccountablesupervisionofthefieldoperationalaspectsoftheRegional911System.Thisincludespolicies,proceduresandprocessesthatimpacttheend-user’sabilitytoprovideservicetoresidentsandvisitors.Forexample,thedeterminationofthenumberofresourcesassignedtoaspecificincident,orthenameanddefinitionofcomplainttypesarepurelyoperationalinnature.However,wheretheneedsoftheenduserbegintoimpactorinfluencethetechnologyneedsofthesystem,theoperationswithina911center,itsrequiredfundingoroverallsystemintegrity,thenothersystemstakeholdersmustproperlybeincluded.FITCHsuggeststhatexistingprocessesberedefinedtofocusmoreonend-userneedsattheinceptionofoperationalissues,andallowingBSOandORCATtoengagelaterintheprocess.Thisshiftprovidesanimblerenvironmentforend-usersandensuresoperationalissuesarefullyvettedbeforeconsideringlogistical,financialorgovernanceconcerns.Aspartofthechangeinfocus,ORCATCommunicationManagersrolesandresponsibilitiesshouldberevisedandclarified.Thesepositionsshouldfocusonqualityassurancereviewsofsignificantdispatchissuesraisedbyend-users,citizensorothersparties;ensuringPSAPfacilitiesandtechnologyareoperatingatmaximumeffectiveness;andcoordinationwithBSOSiteManagers.Toavoidanyconfusion,theexistingORCATpositiontitlesandjobdescriptionsshouldbeupdatedaccordingly.Thesystemhasutilizedapolicylevelcommittee,the4C,duringearlyimplementation.Forreasonsoutlinedpreviously,thefocusnowshouldproperlyshifttoamoreoperationalperspective.Theneedforthe4CnolongerexistsandtheCountyshouldsunsetthegroupasthisnewgovernancemodelisimplemented.Thefigurebelowsummarizesthestepsinasuggestedprocessfortheidentificationofanoperationalissueandthestepstowardaresolution.Thecountyshouldmaintainarepositoryofalldocumentationassociatedwiththisprocesstoensuretransparencyandarchivingofalldecisions.


Figure71.IdentifyingandResolvingOperationalIssues

Recommended Operating Guidelines for Governance Step 1: Issue Identification and Proposed Resolution 1) Aneedforthecreationand/orrevisiontoapolicy,procedureorprocessofthe911systemcanberaised

byanyofthestakeholders–lawenforcement/firerescueagencyasarepresentativeoftheirmunicipality;BrowardSheriff’sOffice;orBrowardCounty.

2) IssueswouldbeconsideredfirstbytheexistingOperationalReviewTeam(ORT).a) EachlawenforcementandfirerescueagencywouldhavearepresentativeontheORT.Typically,the

ORTwouldmeetbydisciplineasiscurrentpractice–lawenforcementorfirerescue.b) TheORTwoulddetermineifanissueinvolvedonlyasingleagency;asinglediscipline(lawvs.fire);or

aregionalconcern.i) Iftheissueonlyinvolvesasingleagency,theissuewouldmovetosub-paragraph4.

Step1:IssueIdentification

• IssueIdentification byanyStakeholder• ReviewbytheOperationsReviewTeam(ORT)•IfapprovedbyORT,summarizedandforwardedtoBSOandORCAT

•IfapprovedbyORT,BSOandORCAT-proceedtoStep2•Non-approved itemsmaybeescalatedtoStep3

Step2:End-UserApproval

•Chiefofdepartmentapproves,ordoesnotappove,issueandproposed resolutionasdefinedinStep1

•Simplemajorityofthosevotingthroughelectronicmeans•Appoved issuesmovetoStep4forimplementation•Non-approved itemsmaybeescaltedtoStep3

Step3:CAOEscalation

•IssuesnotapprovedatStep1orStep2maybeescalatedtorespectivechiefadministrativeofficer•Chiefadministrativeofficersreviewandconfer•TheymayreferbacktoORTforaddtionalconsideration,notapprovetheissue,ordirectapprovalandimplementationunderStep4

Step4:Implementation

•IssuesapprovedatStep2or3willbeimplemented•ORCATwillfacilitateandmonitor


c) EachORTgroupshouldbechairedbyauniformedmemberdeterminedbytherespectiveAssociation.BecauseofBSO’suniqueroleasOperatorinthesystem,thechairshouldpreferablybefromamunicipalagency.

3) TheORTwouldclarifytheissueandproposedresolution.Theitemwouldbesummarizedinwrittenformtoensuretheissueandproposedcourseofactionareclearlyidentified.

4) TheORTwouldrecommendapprovalordenialoftheitem5) UponapprovalbyORT,theissuewouldbeforwardtobothBSOandORCATforreview&comment.6) BSOandORCATwouldconsidertheitemunderthefollowingguidelines:

a) BSOwouldevaluateitsabilitytoprovidetherecommendedresolution.Theyshouldconsiderthelogisticalbenefitsandchallenges,aswellasiftheproposedresolutioncanbedonewithexistingresources.IfadditionalresourceswouldberequiredbyBSO,itmustidentifythefinancialimpact.ThefinaldeterminationoffiscalimpactwouldrestwithBrowardCounty’sbudgetoffice.

b) ORCATwouldconsideranylogisticalimpactsfromtheproposedissueresolutionasitrelatestotheCounty’stechnology(i.e.CAD,radio,911system,etc.).ORCATwouldalsoconsideranypotentialfiscalimpacts,thoughthefinalarbiteroffundingshouldresidewiththeCounty’sbudgetoffice.Finally,ORCATwouldevaluatetheissueandproposedresolutionagainsttheRegional911System’soverarchinggoalsandobjectives.

c) Iftheissueinvolvedonlyasingleagency(see2babove)ORCATandtheinvolvedagencywoulddiscussifthemunicipalitydesirestofundthechange/improvementifthesolutionisbeyondthebaseservicesprovidedbytheCounty.

d) IftheissueisdeniedbytheORT,theproposermayelecttoadvancetheissuetoStep3.7) Oneapprovedbyallthreestakeholdergroups(ORT,BSO,ORCAT),theissueadvancestoStep2.8) Iftheissueisnotapprovedbyanyofthestakeholdersin7)above,theproposermayelecttoadvancethe

issuetoStep3.Step 2: Approval by End-Users 1) Issuesapprovedbyallthreestakeholdergroupswillbeapprovedbyfirechiefsand/orpolicechiefs.

a) WhileORTwilloftenbecomprisedofoperationalmanagersfromlawenforcementandfirerescueagencies,approvalatStep2requiresthespecificreviewandapprovalfromthechiefofdepartmentforeachlawenforcementand/orfirerescueasapplicable.

b) Summaryoftheissueandproposedresolution,aspreparedandapprovedatStep1,willbesenttothechiefofdepartmentforlawenforcement,firerescueorbothastheissuemayrequire.

c) Items,assummarizedatStep1,willbeballotedtochiefsofdepartmentelectronically.ThechairoftherelevantORTshallensurethewrittensummaryoftheissueisforwardedtotherespectiveAssociationpresidentfordistributiontoeachchiefofDepartment.

d) Eachchiefofdepartmentisexpectedtoreviewandapproveornotapprovetheitem.e) ConsistentwiththeexistingpracticesforboththeFireChiefsAssociationandPoliceChiefs

Association,asimplemajorityissufficientfortheitemtobeapproved.UltimatelythedeterminationofeachprofessionalAssociation,andofthemethodsbywhichtheyapproveitems,isleftwithinthepurviewofthatAssociation.

2) UponapprovalatStep2,theitemwillmovetoStep4forimplementation.


Step 3: Escalation to Chief Administrative Officer 1) ShouldanissuenotgarnersupportforapprovalatStep2,theproposerorotherstakeholdermay

electtoescalatetheissuetotheirrespectivechiefadministrativeofficer(citymanager,countyadministrator,orSheriff)asmaybeapplicable.

2) Thechiefadministrativeofficer(ordesignee)oftheagencyseekingtoescalatetheissuemayelecttoupholdthedeterminationmadeatStep1orStep2,orconferwiththechiefadministrativeofficers(ordesignees)ofotherstakeholders.

3) Thechiefadministrativeofficersofallthreestakeholders,shouldtheyagree,mayalsodirectaspecificresolutiontotheissue;requestORTreconsidertheissue;ortakeotheractionsastheydetermineinthebestinterestfortheirlocalgovernment.

Step 4: Implementation 1) IssuesidentifiedatStep1,andapprovedateitherStep2orStep3,shallmovetoStep4for

implementation.2) ORCATwillfacilitatetheimplementationwiththeactivesupportofotherstakeholders.3) ORCATwillmonitorandreporttoallstakeholderstheprogressandissuesapprovedfor

implementation.4) ORTmembersareexpectedtokeeptheirrespectiveagenciesinformedofissuesundergoing

implementation.FITCHnotedthelevelsofstaffingappropriatedinthebudgetprocessfortheRegionalE911System,andthefocusofstaff’seffortsinoperatingtheSystem.Qualitatively,itwasfeltthatpersonnelinthe911centerssufferfromlowmoraleandaperceivedlackofleadership.Attentionhasbeendivertedfrommoremeaningfulactivitiesinordertoaddressissuesoflessimportance,andasensethatavailableresourcesarenotbeingusedeffectively.Quantitatively,applicationbyFITCHofmoredefinitivestaffingmodelsdemonstratesopportunitiestoachievemeaningfulperformanceinthe911centers–wellwithinexistingallocationsofpersonnel,andevenwithsomelevelofthoughtfulreductions.WebelievethiscanbeaccomplishedwhiletheBrowardSheriff’sOfficeremainsanAccreditedCenterofExcellenceasawardedbytheInternationalAcademiesofEmergencyDispatch.ThesectionofthisreporttitledDispatchOperationalModels–Options,proposesseveraloptionsthatincludeanumberofstaffingrealignmentstoaddresstheexistinginefficiencies.TheopportunityalsoexistsforBSOtoalteritscurrentstaffingscheduleinorderto1)alignresourcesnecessarytotheactualdemand,and2)provideopportunitiesfordispatchpersonneltoworkalternateshiftschedules,therebyprovidinggreateropportunitiesforreducedstressanddemandsofmandatoryovertime.AnexampleofonesuchscheduleiscontainedinAttachmentA,SchedulingMatrixSample,thatutilizesacombinationofeighthourand12hourshiftsinordertomorecloselyalignstaffinglevelswithactualdemand.Thesystemevaluatesdemandinfourhourblocksandbybalancingthenumberofpersonneloneithereightor12hourshifts,dispatchmanagersandsupervisorscandevisemoreefficientschedules.Ultimately,thisproposalallowsforindividualdispatchers,basedoncontractualseniorityandexistingmethods,toselectvariousworkschedulesthatwillmorelikelyalignwiththeirpersonaldesires.


911centeroperationsareextremelystressfulandfrequentlymustaddressunusualorintenseincidents.Therecommendationsregardingsupervisoryspan-of-controlarebasedonthedualrolesofdispatchfloorsupervisors.First,afloorsupervisorplaysanactivequalityassuranceroleinrealtime.Dispatchprocessesinvolvecomplicatedscenariosandinthisrole,afloorsupervisorprovidesimmediatesupportandadherencetopolicy/protocols.Thesecondroleisthatofanactiveliaisonwiththeirequivalentlevelfieldpersonnel.Whenthisroleisfulfilled,fieldsupervisorsanddispatchsupervisorsarewellinformedregardinganyissuesofthedayandthatinformationispassedseamlesslyfromshifttoshift.Characteristicssuchastheserequirehigherlevelsofsupervisionthenmaybefoundinotherdisciplines.Currently,BSOoperateswithasupervisortodispatcherratioofapproximately1:9.For911centers,thesupervisortodispatcherratioshouldbecloserto1:6.WhichtheproposedstaffingoutlinedineitherOption1orOption2,andkeepingexistingsupervisionlevelconstant,willcomeclosetoachievingthedesired1:6ratio.BothBSOandoftheCountyshouldmonitorandreallocateexistingresourcestoachievethistargetsupervisorlevel.Trainingwasalsoanareaidentifiedasrequiringmoreresources.Withtherecommendedadjustmentstostaffingoutlinedelsewhere,someresourcesshouldberedeployedtoenhanceBSO’sdispatchertrainingprograms.BSOhasindicatedtheyarepursuingnewapproachesandtechnologiestoongoingtrainingrequirements.Thisisapositiveeffort,butonethatshouldalsoincludeadditionalpersonnelinordertomorecloselysupporttheidentifiedtrainingdeficitsthatcomeoutofothersystemanalysisandreviews.AsaBSOimplementsanynewapproachestotraining,theyshouldidentifyanyadditionalstaffingresourcesrequiredandseekreallocationofresourcesfromtheCountyQualityassuranceeffortsfallwithinthepurviewofoversightandarerightfullyaresponsibilityoftheCounty.TheCountyisthebodythatshouldconductqualityassuranceassessmentsforcurrentEmergencyMedicalDispatch(EMD)functionsandforanyfuturedeploymentofEmergencyFireDispatchand/orEmergencyPoliceDispatchprotocols.IndependentqualityassuranceprocessessuchasPriorityDispatch’s“NationalQ”arereadilyavailableandcanprovideobjectivefeedbackoncompliancetoprotocols.TheCountyisalsoresponsibleforensuringsystemparticipant’sneedsarebeingmetandagreeduponresponsibilitiesbeingfullfilled.Whereappropriate,thisincludesperformingauditstoensurecontractualresponsibilitiesarebeingmet.Ataminimum,thecontractedqualityassuranceprogramsshouldbeabletoaccommodatethefollowingpoints:

§ Focusonevaluatingtelecommunicatorstrengthsandidentifyareasforimprovement,§ Provideastandardizedoperationthatprovidesconsistentandongoingimprovement,§ Developaclearlydefinedprocessforqualityassurancecasereviewsthatincludecalltakingand

dispatchactivities,§ Assurethatcasereviewsareperformedonaregularbasiswithfeedbackprovidedtothe

telecommunicator,assoonaspossible,§ AssurethatpersonnelresponsibleforperformingQAreviewsmeetminimumqualificationsfor

QualityAssuranceEvaluators,

Additionally,thefollowingpracticesaretobeincludedintheQAprogram:§ Casereviewcriteria


§ Evaluationguidelines§ OversightCommitteeasperprotocoloragencyguidelines§ Programmonitoring§ Recordkeeping§ Reportingandfeedback§ Writtendirectives.

Severalmunicipalitieshaveexpressedthedesiretoprovidedispatchpersonnelwhoarededicatedtohandlingagency-specificfunctionsuchashandlingpoliceandfirepolicyissues,dedicatedtacticalradiooperators,orhandlingnon-emergencyservicerequests.Beforetheserequestscanbeconsidered,theCountywillneedtoclearlydefinethebaseservicefunctions.ItisunderstoodthattheCountyisresponsibleforfundingthebase911services.Largely,thesebaseservicesarecurrentlydefinedintheInterlocalagreementsbetweenBrowardCountyandvariousmunicipalities.However,thiscurrentdefinitionshouldbefurtherclarifiedtoincorporatethelevelsofperformanceasidentifiedherein.Forexample,whileansweringof911callsisabaseservice,theperformancelevelrecommendedhereistoadjuststaffingtoalevelconsistenttoachievebetween3-5secondsatthe90thpercentile.Similarly,radiochannelstaffingforfirerescueroutineresponseshasbeenmodeledtoreflect2secondsatthe95thpercentile.Respectivemunicipalagenciesshouldbeabletopurchaseadditionalstaffing-hoursinordertoachievetheservicestheydesirethatareaboveandbeyondbaselevel911services.

Performance Metrics

Findings —

Certainperformancemeasureshavebeenmisinterpreted,incorrectlyapplied,orareinconsistentwithcurrentindustrybestpractices.TheCounty’suseofPASS/FAILtargetsprovideslittleinthewayofinformationforcontinuousqualityandperformanceimprovement.ThefailureofthecurrentPASS/FAILorYES/NO,P1busyhourtarget,isthatitprovidesnoguidanceastothelevelofsurgecapacitythatisfiscallyresponsibletobuildintothesystem.

Recommendations —

TheCountyshouldmodifythecurrentmonthlyperformancereportformatandreplaceitwithamonthlyreportthatfocusessolelyondataandprovidesnocommentary.The“busyhour”istoberedefinedinaprospectivemannerbasedonhistoricaldataandistobereassessedinnolessthan12-monthintervals.Thesechangesallowformeaningfulandactionableinformationexchangesandprovideuseragencieswithaneededlevelofoversight.


TheCountyshouldpurchaseaperformancemeasurementsoftwarepackagethatwillprovideagencieswithreadyaccesstotheactivitiesandperformanceoftheirrespectivefieldunits,andsimultaneouslyallowtheCountyandBSOtoevaluatesystemperformanceatthemicroandmacrolevels.Onlytheperformanceonemergency/911incidentsshouldbeincludedintheperformancereports.Thecurrentpracticeofevaluatingduplicate911callsonasingleincidentskewsmeasurement.Thetruestructureofthereportshouldbetopresentthenumbersinawaythathighlightsthecallswhereresponsetimeisimportant.Somethoughtshouldbegiventopresentresponsetimesstartingwiththecallreceipttoemergencyservicearrivalonscene.Thiswillgivetheproperpresentationofthecaller’sexperience.Specialattentionshouldbepaidtohighpriorityincidents.Regardingreportingperformanceforvariouscallprocessingtimeintervals,oncethetechnologyissuesareresolved,theP2andP3intervalsshouldbereportedseparatelyandasacombinedmetric.Thereasoningisthat,particularlyforfireandemergencymedicalDeltaandEcholife-threateningcalls,fastandeffectivedispatchperformancecontributestopositiveoutcomes.MonthlyreportsshouldalsoreportP4(turn-outtimes)forfirerescueincidentsandP5(traveltime)forbothfirerescueandlawhighpriorityincidents.Ingeneral,dispatchcenterperformancemetricsaretofocusonoptimizingdispatchprocessesasmuchaspossible,withtheendresultbeingtogethelpmovingtoemergenciesasquicklyaspossible.Theprimaryobjectiveistocontributetothepotentialforpositiveoutcomesforpatientsandproperties.

Discussion —

ThediscourseregardingsystemperformancebetweentheCounty,BSOanduseragencieshasbeendifficult.Multiplefactorsincludinglimitationsofsomeperformancemetrics;operationalgovernanceandoversight;andtechnologylimitations,contributetovariousproblemareas.However,FITCHalsofoundareaswheretherehavebeennoteworthysuccesses.Contrarytooftencitedperceptions,theSystemisperforming–quantitatively–betterthanconveyedbystakeholders.Awidelydiscussedmetricthatevaluates911call-answeringtimeswasfoundtobeextremelyrapid,someofthequickestFITCHhasidentifiedinotherlargesystems.Calltransfers,thathappenedwithsomeregularitypriortoconsolidationanddelayedeffectivesystemperformance,hasbeenvirtuallyeliminatedsinceconsolidation.TheCounty’seffortstoensurequalityandefficiency,supportedbyaqualityassuranceandimprovementprogram,shouldcontinue.Additionally,greateroperationalcoordinationandtransparencyamongSystemparticipantshasprovidedqualitativeimprovements.Asuggestedmonthlyperformancereport,focusedonobjectiveperformancedataandappropriatefordisseminationtopolicymakers,iscontainedinAttachmentJ,MonthlyPerformanceReportFormat.Asnotedintherecommendedformat,responsetimesforhighprioritylawandfirerescueincidentswillalsobereported–acknowledgingthepublic’sperceptionofservice.Stakeholdersshouldnotetheadditionofreportingaveragevaluesforperformancemeasures.Whilestakeholdersshoulddesign,assessandreportperformanceutilizingfractileorcomplianceperformance(e.g.90%within10seconds),membersofthepublicgenerallyareusedtoperformancebeingreportedasanaveragetime.Forthesereasons,


botharerecommendedinAttachmentJ,butstakeholdershouldbeutilizing“Target”andTargetCompliance”intheirassessmentofSystemperformance.Thisdoesnotprecludeadditionalormoredetailedsystemanalysis,butratherismeanttofocusonhigherlevelmetricofSystemperformance.Inaddition,whiletheintentforthemonthlyperformancereportistoremainrelativelyobjectivebysimplyreportingmetrics,theCountycontinuestohavetherightandobligationtoaddressperformanceissuesasoutlinedintheexistinginterlocalagreements.MeasuresoftheSystem’sperformance,asinitiallydraftedbylawenforcement,fireandmunicipalleaders,andimplementedbyCountystaff,donotprovideanappropriateassessmentoftheSystem’sperformance.ThemeasureoftheP1busyhourinterval–thetimefromwhenthe911phoneringsuntilanswered–isapoorrepresentationofSystemperformanceandinconsistentwithcurrentindustrybestpractices.Anotherexampleofgoaldisplacementisthefocusonthetimenecessarytoanswera911call,knownbythemonikerP1.Thismeasurehasreceivedsignificantscrutiny.WhilethereareseveralspecificmeasurestoevaluateP1,muchofthefocushasbeenonwhatisknownas“busyhour”performance.Thissinglemetrichasbeenthesourceoffrictionbetweenvariouspartiesandlikelyledtoabeliefthattheonlysolutionisincreasedstaffing.ThebusyhourmeasureisapoorrepresentationofperformanceintheBrowardsystem.WhenexaminingtheothermetricsassociatedwithP1,theBrowardSystemactuallyexhibitssomeofthebestperformanceseeninlarge911centersacrossthenation.Further,reportsoftheP2/P3interval–thetimefromansweringa911calluntilunitsaredispatched–thatappeartobeprecise,areinfactflawedduetodatalimitations.Interestinglythough,performancecalculatedbyFITCHdifferedfromthatcalculatedbytheCountybyonlyafewpercentagepoints.Oncethetechnologylimitationsareaddressed,thesystemshouldfocusonP1answertimesonlyforthoseincidentsthatgenerateacallforservice.Inthismanner,onlydatafromtheCADsystemwouldbeutilized.UtilizingaprospectivedefinitionofP1alsolimitsrandomsurgesinthesystemtoinappropriatelyleadtoconclusionsofpoorperformanceinthismetric.Basedondataelsewhereinthisreport,thesystem’scurrentbusyhouris1800hrs.Additionally,theCountyhaspreviouslyidentifiedthatcallintaketimesarewhatgenerallyleadtofailuresinmeetingP2/P3performancetargets.Byevaluatingeachcomponentseparately,P2forcallintaketimesandP3forunitnotificationtimes,thesystemandoperatorarebetterabletoidentifyrootcausesofperformanceissues.WhiletheSystemisseenasstrugglingtomeetsomeofitscurrentlydefinedperformancemeasures,thefocusoncertainspecificareashasresultedinalevelofgoaldisplacement.Theuseof‘PASS/FAIL’or‘YES/NO’againstpercentagecompliancetargetsdoestheCountyadisserviceinthatitfostersanexpectationthatthesystemcansomehowbemadeperfect.Therealityofemergencyservicesystemsisthattheywillbeoverwhelmedbysignificantunanticipatedeventsatsomepointintime,i.e.,therecentshootingsinOrlandooratornadoinSouthFlorida.Performancemeasuresshouldbeselectedsuchthattheycontributetoaknowledgebasetomakethesystembetter,ratherthanbeseenasavaluejudgement.Attentiontoperformancemetricsisabestpractice,butmustbeutilizedcarefullytoavoidemergenceofperversebehaviors.Thefocusofperformancemeasuresshouldbeonthetimelyandaccuratecollectionandtransmissionofinformationtofirstresponders.Metricsthatassessperformanceonlowpriorityincidents,whileinteresting,mustbeevaluatedseparatelyfrommetrics


designedtomeasureperformanceintrueemergencies.Forexample,EMSrelatedcallsarecategorizedthroughtheEMDprocess–withDeltaandEchodeterminantsrepresentingthoseemergenciesrequiringtheclosestandallvehiclestorespondwithlightsandsirens.Ingeneral,dispatchcenterperformancemetricsmustfocusonoptimizingdispatchprocessesasmuchaspossible,withtheendresultbeingtogethelpmovingtoemergenciesasquicklyaspossible.Theprimaryobjectiveistocontributetothepotentialforpositiveoutcomesforpatients,residentsandproperties.TheCounty,BSOandcitiesrequiretoolsthatmakethesedistinctions.TheCountyshouldprovideaweb-basedperformancemeasurementsystemthatprovidesfeedbacktokeystakeholdersinrealtime.Thesystemshouldallowindividualcommunitiestomonitorperformanceofthe911centerandtheirfieldresourcesinrealtime.BSO,asthesystemoperator,shouldbeabletoseeperformanceatboththesystemlevelandattheindividualdispatcherlevelinordertoidentifyandaddressperformanceissues.ThesesystemstypicallytieintotheCADsystemandtheallowevaluationofP1throughP4performancemetrics.Individualcommunitieswouldbelimitedtoseeingonlytheirindividualagencyresources.Boththedisplaysandtargetgoalsshouldbeuserdefinable,byagencyanddiscipline.Mostprovidersofthesoftwareallowforuserstoaccessinformationfromeitheradesktopormobileclient.TheCountyshouldconsiderthiscapabilitytobeabaseserviceforallsystemstakeholders.

Efficiency and Effectiveness

Findings —

CurrentPSAPs,trainingfacilityand“fleeto”planshavefacilitylimitations,especiallyrelatedtoadequatespace.Theconsolidatedsystemiscapableofclosestunitresponsetolife-threateningemergencies,butprotocolsarenotyetinplacetoimplementthiscapability.Radiotrafficutilization,bybothfire/EMSandlawenforcementunits,iscomparativelyhigh.MDTs(mobiledataterminals)andMCDs(mobilecomputingdevice)arenoteffectivelyutilizedtoreduceradiotraffic.BSOcurrentperformanceindicatesoverstaffingincalltakerpositionsbasedonErlangmodeling.BSOcurrentperformanceindicatesoverstaffinginFIREAssignmentpositionsbasedonErlangmodeling.

Recommendations —

CallprocessingstaffingshouldbeadjustedtoachieveP1/call-takingperformanceofbetweenthreetofivesecondsatthe90thpercentilebyadoptingtherecommendedworkstationfunctionalreorganizationasdetailedinthereportsectiontitled,DispatchOperationsModels–Options.Thisadjustment,in


conjunctionwiththealreadyimplementedsinglequeforcallintake,providessignificantefficienciesinthecalltakingprocesswhilemaintaininghighlevelsofperformance.Fire-rescueagenciesshoulddevelop,approveandimplementcountywidenearestunitresponseprotocolsthatapplyirrespectiveofjurisdictionalboundariesinthoseincidentsinvolvinghighpriorityincidents(e.g.Delta&EcholevelEMDcalls).RecommendedprocesschangestoradiochannelusageincluderequiringincreasedusageofMobileDataTerminals(MDTs)byfieldresponders.OncetheCADisupgradedtoallowautomaticcomputerassignment/recommendationofresponseunitsforfire/rescuecalls,asingle“gatekeeper”function/firerescuealertchannelcanbeimplementedtomanuallyapprovetheassignmentconsistentwithOption2.Upondispatch,pre-definedtacticalradiochannelswouldbeusedformoreroutineforfireincidentsandEMSincidents.Moresignificantincidents(structurefires,major/multipleunitresponses)wouldbeassignedadedicatedtacticalchannel.Thischangeinfirerescueradiooperationsprovidessignificantefficiencieswhilemaintaininghighlevelsofperformance.LawenforcementradiopositionsshouldbeconsolidatedtoincreaseefficiencyconsistentwithOption2.Long-termcapitalbudgetingprogramshouldbeconsideredassoonaspracticaltoincludetwonewpurpose-specific911facilities.

Discussion —

TheCounty’srecentCapitalBudgetincludes$350,000inFiscalYear2017forplanningandanalysisstudiestoidentifydevelopmentoptionsforneworreconstructedPSAPfacilities.TheultimategoalshouldbetohavetwogeographicallydisparatePSAPs28bothofwhicharestaffed24/7/365.Failureofonephysicallocationwillallowthealternatelocationtocontinuetoprovideessentialservices.Currentfacilities,ascurrentlyutilized,appeartobeforclientagenciestocohabitateratherthanconsolidate.Designeffortsthatfocusonconsolidationwillbethemostsuccessful.FITCHdoesnotunderstandfromarootcauseperspective,thefailurebyfieldpersonneltomakebetteruseofmobiledataterminals(MDTs).Thispracticeplacesalargerdemandontheradiosystemand911personnelandfurtherdecreasestheeffectivenessandefficiencyoftheSystem.RequiringfieldpersonneltousetheirunitMDTsinsteadofradiochannelswillallowformoreefficientuseofcurrentdedicatedradiochannelpersonnel.TheCountyshouldensurethatthenewmobileclient,implementedwiththenewCADsystemin2017,meetstheneedsofbothfirerescueandlawenforcementfirstresponders.Firechiefsandpolicechiefsshouldinternallyensuregreaterutilizationofavailabletechnologydesignedtolessentheexistingoverusageofverbalcommunications.

28NFPA1221-2016.Section4.1.5.2


Erlanganalysessummarizedelsewhereinthisreport,demonstratesinefficienciesinboththecalltakingandfirerescuedispatchoperations.Adjustingperformancetargets,andmodifyingradiooperationsforfirerescueincidents,bothprovidepotentialsignificantefficiencies.Forthepublic,thereshouldbenodiscernibleimpactonthecallansweringfunction.Forfirerescuepersonneltherewillbeaneedtoadjusttothenewradiochannelassignments-forexample,switchingtoaEMSworkingchannelimmediatelyuponassignment.Insomeregards,systemparticipantsdidnotconsolidatetheir911centers,butrathersimplygeographicallyrelocatedtheiroperations.Theinitialconfigurationoffireandlawradiooperationssimplyreflectedthethenexistingpractices.Toachievetheefficienciesidentifiedhere,fireandlawenforcementagenciesmustconsolidateradiochannelstomoresustainablelevels.Theproposedchangesallowforgreatercapabilityoftacticaloperatorstohandlesignificantfire-rescueincidents.Thisshouldaddressaconcernvoicedbybothfirepersonnelandcurrentradiooperators.Whilethemodelingdemonstratesthatsizableadjustmentsareavailable,implementationofchangesshouldoccurinamoredeliberateandmeasuredmanner.TheCountyshouldrequestBSOimplementanyrecommendedchangesintwoorthreephases–titratingstaffinglevelswhilemonitoringperformance.Ultimately,changesshouldbeabletobefullyimplementedwithin12months.AsnotedintheExecutiveSummary,itisimportantthattheCountyandBSOassurethateachchangephaseiscompletelyembeddedinoperationalproceduresandtheorganization’sculture,beforeseekingadditionalchange.TheCounty’sCharter,statesinpart,thatthe“Countyshallprovidefundingforthecommunicationsinfrastructureandallserviceproviderswillutilizetheelementsofthecommunicationsinfrastructure.Thecommunicationsinfrastructureshallfacilitateclosestunitresponseforlife-threateningemergencies...“.Whiletherehasbeensomepilotprojects,andincreasedautomaticaidforcertainincidents,therehasnotyetbeenaconcertedefforttofullyrealizetheintentoftheCounty’sCharter.Bylimitingtheuseofnearest-unitresponsetoDeltaandEchoincidents,asdeterminedthroughuseoftheexistingEMDprogram,thiswillmeettheCharterlanguageof“forlife-threateningemergencies”whichrepresentapproximately15%ofallEMScalls.

Broward County Attachment B - Page 1 © Fitch & Associates Assessment of Broward County’s Regional E911 December 2016

Attachment A

Scheduling Matrix Sample

Broward County Attachment A - Page 1 © Fitch & Associates Assessment of Broward County’s Regional E911 December 2016

AttachmentA.SchedulingMatrixSampleOneofthemorecommonworkschedulesfordispatchcenterpersonnelisthefour-dayworkweek.Thisscheduleprovidespersonnelwithalevelofwork-lifebalancethatisacceptableandappropriateforpersonnelworkinginahigh-stressenvironment.Thissectionprovidesthestepsforbuildingaschedulematrixbasedonacombinationof12/8-hourshiftschedulesforafour-dayworkweekandatraditional8-hourscheduleforafive-dayweek.• The4-hourincrementsstartatthetimeofdaytheshiftworkbegins.• Determinethestaffinglevelsneededusingthe4-hourincrements.• Determinethestaffinglevelsneededbydayofweekandhour(4-hourincrements)ofday

baseonnormalactivity.• Thescheduleforthisexampleisbasedonthefollowingsetsofdaysoff:

1. Saturday,Sunday,Wednesday2. Sunday,Monday,Thursday3. Monday,Tuesday,Friday4. Tuesday,Wednesday,Saturday5. Wednesday,Thursday,Sunday6. Thursday,Friday,Monday7. Friday,Saturday,Tuesday

Tables1and2belowprovidesampleschedulesastitled.


Table1.SampleCombination12/8HourShiftSchedule–4Days/Week

Sunday Monday Tuesday Wednesday Thursday Friday Saturday

1 X 06-18 06-14 X 06-18 06-14 X

2 X X 06-18 06-14 X 06-18 06-14

3 06-14 X X 06-18 06-14 X 06-18

4 06-18 06-14 X X 06-18 06-14 X

5 X 06-18 06-14 X X 06-18 06-14

6 06-14 X 06-18 06-14 X X 06-18

7 06-18 06-14 X 06-18 06-14 X X

8 06-14 X 06-18 06-14 X X 06-18

9 X 06-18 06-14 X 06-18 06-14 X

10 X X 06-18 06-14 X 06-18 06-14

11 X 10-18 06-18 X 10-18 06-18 X

12 X X 10-18 06-18 X 10-18 06-18

13 06-18 X X 10-18 06-18 X 10-18

14 10-18 06-18 X X 10-18 06-18 X

15 X 10-18 06-18 X X 10-18 06-18

16 X 18-06 18-02 X 18-06 18-02 X

17 X X 18-06 18-02 X 18-06 18-02

18 18-02 X X 18-06 18-02 X 18-06

19 18-06 18-02 X X 18-06 18-02 X

20 X 18-06 18-02 X X 18-06 18-02

21 18-02 X X 18-06 18-02 X 18-06

22 18-06 18-02 X X 18-06 18-02 X

23 X 18-06 18-02 X X 18-06 18-02

24 22-06 X X 18-06 22-06 X 18-06

25 18-06 22-06 X X 18-06 22-06 X

26 X 18-06 22-06 X X 18-06 22-06

Combination12/8HourShiftSchedule-4Days/Week


Table2.Sample8-HourShiftSchedule–5Days/Week

1 X 06-14 06-14 06-14 06-14 06-14 X

2 06-14 06-14 X X 06-14 06-14 06-14

3 06-14 06-14 06-14 06-14 06-14 X X

4 06-14 X X 06-14 06-14 06-14 06-14

5 X X 14-22 14-22 14-22 14-22 14-22

6 14-22 14-22 X X 14-22 14-22 14-22

7 X X 14-22 14-22 14-22 14-22 14-22

8 14-22 14-22 X X 14-22 14-22 14-22

9 14-22 14-22 14-22 X X 14-22 14-22

10 14-22 14-22 14-22 14-22 X X 14-22

11 14-22 14-22 X X 14-22 14-22 14-22

12 14-22 14-22 14-22 X X 14-22 14-22

13 14-22 14-22 14-22 14-22 X X 14-22

14 14-22 14-22 14-22 14-22 14-22 X X

15 14-22 14-22 14-22 14-22 X X 14-22

16 14-22 14-22 14-22 14-22 14-22 X X

17 22-06 22-06 22-06 22-06 22-06 X X

18 X X 22-06 22-06 22-06 22-06 22-06

19 X X 22-06 22-06 22-06 22-06 22-06

20 22-06 22-06 22-06 22-06 22-06 X X

8-HourShiftSchedule-5Days/Week


Ifthestaffinglevelsneedtobehigherfortheweekend,thenbuildthematrixinreverseorderstartingwithdaysoff(7through1).Additionalpeaktimescanbeaddedasneededtohandleanticipatedsurgesinthesystem.Figure1andTable3belowgiveexamplesofpeakstaffingbasedonfour-hourblocks.

Figure1.GraphicofStaffingChangesin4-HourBlocks

Table3.PersonnelOneDutyin4-HourBlocks

0

2

4

6

8

10

12

14

16

1 2 3 4 5 6

Staffingper4-HourBlocks

4-HourBlock AvgonDuty

9 9 11 10 10 11 10 06-10 10.0

10 11 12 11 12 13 11 10-14 11.4

14 16 16 11 13 14 16 14-18 14.3

15 16 13 11 13 14 16 18-22 14.0

8 9 9 8 11 10 9 22-02 9.1

6 7 6 7 9 7 6 02-06 6.9

Broward County Page 2 © Fitch & Associates Assessment and Recommendations of Broward County’s Regional E911 December 2016

Attachment B

Performance Measures

Broward County Attachment A - Page 1 © Fitch & Associates Assessment of Broward County’s Regional E911 July 2016

ATTACHMENTA:PerformanceMeasuresFROMOPERATORAGREEMENT 22

22FromExhibitDoftheAgreementbetweenBrowardCountyandSheriffofBrowardCountyforTheOperationofCall-taking,Teletype(QueriesOnly)andDispatchServicesfortheConsolidatedRegionalE911CommunicationsSystem

-14-

EXHIBIT "D"

PERFORMANCE STANDARDS

The performance of the Consolidated Regional E-911Communications System (System) will be based on the Lifecycle of an Emergency Call for calls received on the emergency lines (911 lines). As illustrated in the diagram below, operational performance indicators P1, P2, and P3 will be measured, reported and benchmarked against industry best practice standards. Efficiency (cost) measures will be utilized to evaluate the cost of the System.

Emergency Event

9-1-1 Call Initiated

Call Rings

at PSAP

Call Answer

Call Dispatch Unit Arrival

CAD Event

Closed

P1 9-1-1 Call Answer TimeP2 Time from Call Answered to Call Entered in CAD (and forwarded to Dispatcher)P3 Time from CAD Entry until a Unit is DispatchedP4 Time from Unit Dispatched until Unit Arrives on SceneP5 Time from Unit Arrives on Scene until Incident is Closed

Lifecycle of an Emergency Call

Prior to Scope of PSAP Operation

10 secsPSAP / Responder Dispatch Response TimePSAP 9-1-1 Call

Answer Time

Caller Interrogation and Call Entry into CAD

P1 P5P3P2 P4

To ensure the performance of the Consolidated Regional E-911Communications System is evaluated in a reasonable manner, performance standards have been separated based on a transition and post-transition period. COUNTY, OPERATOR and Operational Planning/Implementation Workgroup members will collaborate to provide recommendations to County Administrator on the appropriate operational measures to be used to evaluate the System and establish annual performance targets to ensure incremental progress is being achieved.

Performance Standards will become effective at such time the Participating Community is designated, in writing, by the County as having been migrated to the Consolidated System. Transition Period

The transition period shall begin upon the proper execution of this Agreement and continue through September 30, 2015, as it relates to those Participating Communities set forth on Exhibit "B" as of September 30, 2013.


-15-

The following Performance Standards ("Standards") will be utilized to track the efficiency and operational performance of the regional system on a monthly basis during transition phase:

Efficiency Measurements:

• Operational Cost per call for System • Operational Cost per E911 call received

Time to Answer Emergency (911) Lines Standard:

• Ninety percent (90%) of all 9-1-1 calls arriving at the Public Safety Answering Point (PSAP) during the busy hour shall be answered within ten (10) seconds (P1)

The busy hour is defined as the hour each day with the greatest call volume.

• Ninety-five (95%) of all 9-1-1 calls should be answered within twenty (20) seconds (P1)

Alarms (audible, silent, panic, fire, smoke, medical, etc.) Received on Alarm Lines Standard:

• Ninety-five percent (95%) of alarms received on alarm lines shall be answered within 15 seconds (P1)

• Ninety-nine percent (99%) of alarms shall be answered within 40 seconds (P1)

First Call Process Time Standard:

Emergency alarm processing for the following call types shall be completed within 90 seconds 90% of the time and within 120 seconds 99% of the time (P2 and P3):

• Calls requiring emergency medical dispatch questioning and pre-arrival instructions

• Calls requiring language translation

• Calls requiring the use of a TTY/TDD device or audio/video relay services

• Calls of criminal activity that require information vital to emergency responder

safety prior to dispatching units


-16-

• Hazardous material incidents

• Technical rescue

• With the exception of the above six call types, 80% of emergency alarm call processing shall be completed within 60 seconds, and 95% of alarm processing shall be completed within 106 seconds (P2 and P3)

• Where alarms are transferred from the primary public safety answering point

(PSAP) to a primary and secondary answering point, the transfer procedure shall not exceed 30 seconds for 95% of all alarms processed* (P2)

*Only applicable if non-participating municipalities operate their own primary and secondary PSAP

Law Enforcement Call Process Time Standard:

• Priority one and priority two law enforcement calls shall be processed within 45 seconds, 90% of the time ** (P2 and P3)

• Priority three law enforcement calls shall be processed within 90 seconds, 90%

of the time ** (P2 and P3) Note: Availability of police units shall be considered when reviewing performance.

Agencies must adopt standard signal codes to evaluate performance and the authority having jurisdiction shall determine time frames allowed to the completion of dispatch.

**Priority assignments based on current proposed standard

Emergency Medical Dispatch Standard:

• 95% case entry compliance rate

• 90% total compliance rate (case entry, chief complaint, key questions, and post-dispatch/pre-arrival instructions)

• 1% of all cases receive quality assurance case review*

*Based on NAED compliance standard for agencies with a call volume of over 500,000


-17-

Post-Transition Period

The post-transition period begins October 1, 2015. The performance targets of the Consolidated Regional E-911Communications System will be based on the Lifecycle of an Emergency Call for calls received on the emergency lines (911 lines). COUNTY, OPERATOR and Operational Planning/Implementation Workgroup members will collaborate to provide a recommendation to the County Administrator on the appropriate operational measures to be used to evaluate the System and establish annual performance targets to ensure incremental progress is being achieved.

The following Standards will be utilized to track the efficiency and operational performance of the regional system on a monthly basis during the post-transition phase:

Estimated Efficiency Measurements(Subject to Change):

• Operational Cost per call for System (Target: $9.83) • Operational Cost per E911 call received (Target: $14.85)

Efficiency Measurements shall be updated annually by COUNTY Time to Answer Emergency (911) Lines Standard:

• Ninety percent (90%) of all 9-1-1 calls arriving at the Public Safety Answering Point (PSAP) during the busy hour shall be answered within ten (10) seconds (P1)

The busy hour is defined as the hour each day with the greatest call volume.

• Ninety-five (95%) of all 9-1-1 calls should be answered within twenty (20) seconds (P1)

Alarms (audible, silent, panic, fire, smoke, medical, etc.) Received on Alarm Lines Standard:

• Ninety-five percent (95%) of alarms received on alarm lines shall be answered within 15 seconds (P1)

• Ninety-nine percent (99%) of alarms shall be answered within 40 seconds (P1)

First Call Process Time Standard:

Emergency alarm processing for the following call types shall be completed within 90 seconds 90% of the time and within 120 seconds 99% of the time (P2 and P3):


-18-

• Calls requiring emergency medical dispatch questioning and pre-arrival instructions

• Calls requiring language translation

• Calls requiring the use of a TTY/TDD device or audio/video relay services • Calls of criminal activity that require information vital to emergency responder

safety prior to dispatching units

• Hazardous material incidents

• Technical rescue • With the exception of the above six call types, 80% of emergency alarm call

processing shall be completed within 60 seconds, and 95% of alarm processing shall be completed within 106 seconds (P2 and P3)

• Where alarms are transferred from the primary public safety answering point

(PSAP) to a primary and secondary answering point, the transfer procedure shall not exceed 30 seconds for 95% of all alarms processed* (P2)

*Only applicable if non-participating municipalities operate their own primary and secondary PSAP

Law Enforcement Call Process Time Standard:

• Priority one and priority two law enforcement calls shall be processed within 45 seconds 90% of the time ** (P2 and P3)

• Priority three law enforcement calls shall be processed within 90 seconds 90% of

the time ** (P2 and P3) Note: Availability of police units shall be considered when reviewing performance.

Agencies must adopt standard signal codes to evaluate performance and the authority having jurisdiction shall determine time frames allowed to the completion of dispatch.

**Priority assignments based on current proposed standard

Emergency Medical Dispatch Standard:

• 95% case entry compliance rate


-19-

• 90% total compliance rate (case entry, chief complaint, key questions, and post-dispatch/pre-arrival instructions)

• 1% of all cases receive quality assurance case review*

*Based on NAED compliance standard for agencies with a call volume of over 500,000

For a municipality that elects to become a PARTICIPATING COMMUNITY subsequent to September 30, 2013, the development and implementation of the transition plan shall contain provisions to minimize adverse impacts on the System by the addition of such municipality.

REVIEW AND REPORTING OF PERFORMANCE STANDARDS (TRANSITION AND POST-TRANSITION)

Standards shall be evaluated monthly using data from the previous month. Each Participating Community, Police Chief’s Association, and Fire Chief’s Association shall be provided a report on OPERATOR’s performance utilizing this data no later than 30 days following the end of the previous month.

COUNTY shall provide an annual report on OPERATOR’s performance to each Participating Community, Police Chief’s Association and Fire Chief’s Association. A draft of the final version of the annual report shall be delivered to the OPERATOR fifteen (15) calendar days before the intended release date. COUNTY and OPERATOR shall meet within five (5) calendar days thereafter to discuss the annual report's content and attempt to amicably resolve any differences, if any, in the statements, findings, and conclusions, or any combination thereof. If no amicable resolution is reached, OPERATOR shall have five (5) calendar days from the meeting to respond to the annual report and contest the statements and findings therein by providing a written response to COUNTY which response shall be included as an exhibit to the final annual report.

OPERATOR will be evaluated on its ability to achieve the necessary operational and efficiency performance standards, adherence to established actions and overall performance of the Consolidated Regional E-911 Communications System.

FAILURE TO MEET PERFORMANCE STANDARDS (TRANSITION AND POST-

TRANSITION):

In the event a Standard is out of compliance in any month, the following shall occur:

(1) COUNTY shall issue a written Notice of Noncompliance to the OPERATOR.


-20-

(2) OPERATOR shall provide to the COUNTY, a written Notice of Mitigating Circumstance(s) if any, within two (2) business days of the issuance of the Notice of Noncompliance. The Notice of Mitigating Circumstances shall include detailed information and documentation to support OPERATOR's position. For the purpose of this Agreement, a Mitigating Circumstance shall be defined as a natural or man-made incident, accident, disaster, or other environmental or situational anomaly that is unpredictable and, in the reasonable opinion of COUNTY, its occurrence causes an overwhelming and unusual emergency response that greatly exceeds the resources of the SYSTEM. (3) COUNTY shall review any Notice of Mitigating Circumstance(s) that was timely submitted to determine whether the OPERATOR's failure to meet any Standard was due to a Mitigating Circumstance(s). The COUNTY review shall take into account all Mitigating Circumstance(s) that were submitted and their impact on the issue of noncompliance for each Performance Standard. COUNTY shall exercise its discretion to arrive at a reasonable determination that shall be final.

(4) In the event COUNTY determines that the OPERATOR has established, to COUNTY's satisfaction, Mitigating Circumstances related to its failure to achieve a Standard, COUNTY, in collaboration with OPERATOR, shall develop a written action plan to address the noncompliance. The Mitigating Circumstance(s) shall be a factor in the development of the action plan. The COUNTY shall have final approval of all action plans. The action plan may include changes to processes, practices, and procedures and shall include time frames in which the actions must be completed. OPERATOR shall comply with and immediately implement the action plan within the time frames established therein. In the event that OPERATOR shall timely implement all the elements of the action plan to COUNTY's satisfaction, the Notice of Noncompliance subject to Mitigating Circumstances shall be rescinded in writing.

(5) In the event that a Notice of Mitigating Circumstances was not timely submitted by OPERATOR, or following a determination by COUNTY that Mitigating Circumstance(s) were not established, COUNTY, in collaboration with the Operator, shall develop a written action plan to address the noncompliance. The COUNTY shall have final approval of all action plans. The action plan may include changes to processes, practices and procedures and shall include time frames in which the actions must be completed. OPERATOR shall comply with and immediately implement the action plan developed by COUNTY and comply with the time frames established therein.


-21-

(6) Compliance with an action plan shall not excuse OPERATOR from compliance with all Standards in a subsequent month.

(7) The written Notice of Noncompliance and the written Notice of Mitigating Circumstances shall be delivered by e-mail to the following e-mail addresses:

For County:

Rick Carpani [[email protected]], Director of Office of Communications Technology

For Operator:

Robert Pusins [[email protected]], Executive Director of Community Programs, and

Lisa Zarazinski, Lisa [[email protected]] , Director of Regional Communications

(8) In the event that the OPERATOR receives a Notice of Noncompliance for any three consecutive months (excluding any Notice of Noncompliance that was rescinded pursuant to the procedures in paragraph 4 above), the OPERATOR shall be deemed to be in breach and the Agreement shall be subject to termination as set forth in Article 7. In the event that COUNTY issues a notice of breach for noncompliance of the OPERATOR for any three consecutive months, the OPERATOR may cure the breach, if the breach is capable of cure, by performing any and all actions required to meet all Standards that were subject of the Notices of Noncompliance within thirty (30) calendar days from the date of notice of breach to COUNTY's satisfaction.

Notwithstanding the right to cure set forth in Article 7, in the event that the OPERATOR receives a Notice of Noncompliance for four months (excluding any Notice of Noncompliance that was rescinded pursuant to the procedures in paragraph 4 above), whether consecutive or not, in any twelve month period, this Agreement may be terminated upon not less than ten (10) days written notice for breach, without the right to cure.

Attachment C

Calculation of Answer Delays Mathematics & Assumptions

Broward County Attachment C - Page 1 © Fitch & Associates Assessment of Broward County’s Regional E911 December 2016

AttachmentC:CalculationofAnswerDelaysTR=911 IN=

IncomingIsEmpty

S0IsEmpty

E1IsEmpty

E2 E1–E2 AnsDel

0 0 – – – – blank1 0 – – – – blank0 1 – – – – blank1 1 1 – – – blank1 1 0 0 1 – E1-S01 1 0 1 0 – E2-S01 1 0 0 0 <0 E1-S01 1 0 0 0 >0 E2-S0

WithHang-UpsThelogicfordetermininganswerdelaysinvolvinghang-upsrequirestestingthecontentsoffivefields.Theseare:

TR: TrunkIN: IncomingS0: [CIM]ANIinterval,“Start”E1: [CIM]Disconnected,“End1”E2: CallConnected,“End2”

TheBooleanoutcomesofeachtestandthecorrespondinganswerdelaycalculationareindicatedinthefollowing6X8truthtable.Thecoderequiredtoimplementthistruthtableisasfollows:

Ans_Delay [hh:mm:ss.sss] = Let ([ var01 = If(Trunk = "911" and InComing = 1 ; 1 ; 0) ; var02 = not IsEmpty(CIM_ANI) ; var03 = not IsEmpty(CIM_Disconnected ) ; var04 = not IsEmpty(Call_Connected) ; var05 = CIM_Disconnected - CIM_ANI ; var06 = Call_Connected - CIM_ANI ; var07 = CIM_Disconnected - Call_Connected ; var08 = If ( var02 = 1 and var03 = 1 and var04 = 0 ; 1 ; 0 ) ; var09 = If ( var02 = 1 and var03 = 0 and var04 = 1 ; 1 ; 0 ) ; var10 = var02 * var03 * var04 ; var11 = If ( var10 = 1 and var07 < 0 ; 1 ; 0 ) ; var12 = If ( var10 = 1 and var07 > 0 ; 1 ; 0 ) ; var13 = Case ( var02 = 0 ; "" ; var08 = 1 ; var05 ; var09 = 1 ; var06 ; var11 = 1 ; var05 ; var12 = 1 ; var06 ; "" ) ] ; If ( var01 = 1 ; var13 ; "" ) ) /* If ( test=TRUE ; thenresultOne ; elseresultTwo )

Broward County Attachment C - Page 1 © Fitch & Associates Assessment of Broward County’s Regional E911 July 2016

Attachment D

Erlang Mathematics & Assumptions

Broward County Attachment D - Page 1 © Fitch & Associates Assessment of Broward County’s Regional E911 December 2016

AttachmentD:ErlangMathematicsandAssumptions

HistoryAgnerKrarupErlangwasaDanishmathematician,statistician,andengineerwhoinventedthefieldof

telephonenetworksanalysiswhileworkingfortheCopenhagenTelephoneCompanyfrom1908through

1929.ThegoalofErlang’squeuinganalysesistodeterminehowmanyserviceprovidersshouldbemade

availabletosatisfyusers,withoutoverprovisioning.Mr.Erlangquantifiedthethree-cornered

relationshipbetweenrequestsforservice,numberofagents,andlatencyasshowninFigure1,below.

TheconceptsandmathematicsintroducedbyMr.Erlanghave

stoodthetestoftime.Inthemodernworld,thesemethodsare

usedtoanalyzequeuingprocessesinsystemsasdiverseas

shoppersusinggrocerystorecheckoutcashierstodatapacket

switchingthroughInternetroutersatmegahertzfrequencies.

ThearticleauthoredbyChromy,Misuth,andKavackyisaconciseintroductiontotheapplicationofthe

ErlangCformulatoanalysesofemergencyservicescallcenters.1

MathematicsForErlang’sanalysestoapplytoasystem,twoconditionsmustbemet:

§ Usersarrivemoreorlessatrandomintervals;

§ Usersreceiveexclusiveservicefromanyoneofagroupofagentswithoutpriorreservations

TheflowofcallsthroughE911centers,includingBroward’s,conformtotheserequirements.

ThereareseveralversionsofErlanganalysesdependingontheexactmodelofthetrafficflowing

throughthesystem.ThespecificmodelapplicabletotheBSO’sdispatchoperationshasuserseither

beingservedimmediately,orwaitinginqueueuntilacalltakerbecomesavailable.Thespecific

mathematicalembodimentoftheanalysisapplicabletotheBSOsystemisreferredtoastheErlang-C

equation.

Erlanganalysesmustbeconductedoveraselectedintervaloftime.Inthecaseofemergencyservice

communicationscentersexperiencingthenumberofcallsseenatBSOthisintervalismostappropriately

onehour.Littleinsightwouldbegainedbyviewingeachhouroftheyearasaspecialcase.Theneedis

fortheanalysttoconsolidateindividualhoursintogroupsthatpresentavalidpictureofthewaythe

systemfunctions.TheconsolidationprocessappropriatetoBSOhasbeendescribedaboveinthis

Report.

1

E.Chromy,T.Misuth,andM.Kavacky,2011,AdvancesinElectricalandElectronicEngineering,ISSN1804-3119.

Requests

AgentsLatency

Figure1.QueueingTheoryTriangle


TheErlangCformulacalculatestheprobabilitythatanarrivingcallwillbedivertedtothewaitingqueue

ratherthanbeingservedimmediately.ThreecommonsenseparametersgointotheErlangCcalculation:

§ Theaveragearrivalrateofcallsduringthehoursbeingconsidered.

§ Theaveragelengthoftimethedispatcherspendsprocessingeachcall.

§ Thenumberofdispatchersonduty.

ForanErlanganalysis,theworkloadflowingthroughtheBSO’sdispatchoperationsmustbeexpressedin

unitsoferlangs,!.

! = $% Equation1 !:Workloadinunitsoferlangs

$:Averagecallarrivalrateincallsperhour %:Averagecall-processingtimeindecimalhourspercall

TheaveragecallarrivalrateandaveragecallprocessingtimesthatarerequiredtocalculateErlangsin

Equation1areextractedfromthehistoricComputerAidedDispatch(CAD)system,theIntradoVIPER

telephonyserver,andthewrittenreportsofradiousageperchannel.

Toavoidconfusion,thereadershouldbeadvisedthatmanyofthetimeparametersappearinginthe

tabulardatapresentedinthisreportwillbeformattedasdecimalhoursratherthanas

hours:minutes:seconds,hh:mm:ss.Forexample,15minutes,00:15:00,willappearas0.250hr.

Theprobabilitythatanarrivingcallwillbedivertedtothewaitingqueue,&',ratherthanbeingansweredimmediatelyiscalculatedfromtheexpansionoftheErlang-Cequation.

&' = ())!

))+(

(,,! -

())!

))+(

,.)+/,.0

Erlang-CEquation2

!:WorkloadinErlangsfromEqn1 1:Dispatchersondutyatworkstations

Discussionsofqueueingprocessesareoftentabulatedintermsofthreeadditionalparameters:

&2:Probabilitythatanincomingcallwillbeimmediatelyanswered.

3:Averageanswerdelay.Thetimeintervalthatacallinheldinqueue.

4:Averagenumberofcallswaitinginqueueforservice.


Oncetheprobabilitythatanarrivingcallwillbedivertedtothewaitingqueue,&',hasbeencalculatedusingEquation2,thenthesethreeadditionalparameterscanbecalculatedusingthealgebraic

transformationsinEquations3,4,and5.

&2 = 1 −&' Equation3

3 = 789:;< Equation4

4 = 78<:;< Equation5

VariablesPQ,N,andEaredefinedabove.

AbsolutelyrigorousapplicationofanErlang-Canalysisrequiresthatthreeadditionalconditionsbemet:

§ Thatcallersneverhangupwhilebeingheldinqueue.

§ Thatallcallsbeginandendwithinasingletimeinterval.

§ Thatcallersdonotcallbackafterhavinghungupwhileinqueue.

Whentheseconditionsarenotmet,aswillbethecaseintherealworld,thentheErlang-Cformula

predictsthatslightlymorecall-takersshouldbeusedthanarereallyneededtomaintainadesiredlevel

ofservice.Thus,theErlang-Canalysisisgenerallyviewedasprovidinganupperboundtotheneeded

numberofcall-takersrequiredtoserviceagivenflowofincomingtraffic.

WhilethislimitationofErlangCanalysisexists,inpractice,itresultsinanegligibleincreasetothe

numberofdispatcherspredictedforBSO’sdispatchoperations.Theflowofofferedtrafficthroughthe

BSOsystemismodestandthenumberofdispatchersrequiredismodest.Dispatcherscanbeaddedto

orsubtractedfromthesystemonlyinintegerincrements.Underthesecircumstances,incrementingthe

numberofdispatchersby+1willalwaysresultinsuchalargeincreaseinansweringprobabilitythatit

overwhelmsthepropensityofasimpleErlangCanalysistoslightlyincreasetherequirednumberof

dispatchers.

Workloads,StaffingandNon-LinearResponseAconcisepresentationofworkloadpatternsandnon-linearresponseofaqueueingsystemispresented

intheon-linePDFtitled,“CallCenterBasics”.2

Thefollowingisaparaphraseofportionsofthisarticle.

Anaïveapproachtocalculatingthenumberofagentsneededinacallcenteristodividethenumberof

callsexpectedperhourdividedbytheaveragelengthofacall.Forexample,if100callsarriveperhour

andtheaveragetimetoserviceacallis15minutes,thenitappearsthat25agentsshouldbeableto

servicetheworkload.Theflawinthismodelisthatcallsdonotarriveinanorderlyfashion,oneright

aftertheother.Callers,seekingservice,actindependentlyofeachother,andtheircallsarriveina

2

www.easyerlang.com/pdfs/call-center-basics.pdf(July15,2015)


randompatternsurroundingtheaveragespacingbetweencalls.Likewise,theintervalrequiredbythe

agentstoprocesseachrequestforservicedisplaysarandompatternsurroundingitsaveragevalue.

Forcallcenters,thearrivalrateisbestdescribedbyamathematicalfunctioncalledaPoisson

distribution.Thecallprocessingintervalisbestdescribedbyamathematicalfunctioncalledan

Exponentialdistribution.Figures22and23illustratetheshapesofthesedistributions.

Figure22.PoissonDistributionofCallArrivalRates

Figure23.ExponentialDistributionofCallProcessingIntervals

Thestatisticalbehaviorsofthecallarrivalsandcallserviceintervalsguaranteesthatchangesinthe

numberofagentswillhaveanon-lineareffectonperformanceofthesystem.Inthishypothetical

example,anincreaseof10%instaffingwillnotresultina10%decreaseintheaverageanswerdelay.

Rather,theaverageanswerdelayshowsthebehaviorshowninFigure24.

00.00

0.05

0.10

0.15

0.20

0.25

1 2 3 4 5 6 7 8 9 10 11 151412 13

Calls per Hour

Prob

abilit

y

00.00

0.05

0.10

0.15

0.20

0.25

5 10 15 20Call Processing [min / call]

Prob

abilit

y


Table1:AverageAnswerDelayVersusNumberofAgents

Thepurposeofthisexampleistoemphasizethattheperformanceofaqueueingsystemchangesina

verynon-intuitivemannerwithrespecttochangesinbothstaffingandworkload.

Thedependenceofaverageanswerdelayonthenumberofdispatcherisapproximatelyhyperbolic.At

constantworkload,anincrementordecrementof±1dispatchercanresultinverymagnifiedorvery

compressedchangesinaverageanswerdelaysdependingonwhichendofthecurveinFigure3contains

theoperatingpointofthesystem.Thereisnosubstituteforrunningdetailedcalculations,usingdata

specifictothesystemunderconsideration,inordertoaccuratelypredictitsqueueingbehavior.

Insystemswithlargenumbersofagents,therelationshipbetweenaverageanswerdelaysandthe

numberofagentsondutyisapproximatelyacontinuousfunction.Thisrelationshipisverydifferentfor

smallsystems(Figure25).

Table2:AnswerDelaysandAgentsinSmallSystems

Therelationshipremainsapproximatelyhyperbolic,buttheaccessibleanswerdelaysbecomeastep

function.Thenumberofagentsondutycanonlybechangedinintegerincrementsordecrements.

16 17 18 190

100

200

300

400

500

20 21 22 23 24 25

600

Agents On DutyAn

swer

Del

ay [s

ec]

0 1 2 3 4 5 60

5

10

15

20

25

30

35

40

45

50

Agents

Answ

er D

elay

[sec

]

Broward County Attachment D - Page 2 © Fitch & Associates Assessment of Broward County’s Regional E911 July 2016

Attachment E

Quantitation of Workloads

Broward County Attachment E - Page 1 © Fitch & Associates Assessment of Broward County’s Regional E911 December 2016

Attachment E. Quantitation of Workloads

Intake Workstations The data table Call Detail Records captures all incoming traffic to the Intake workstations. Counts of emergency and administrative calls were taken from these records summed by hour of year and averaged by hour of day. Hang-ups and nuisance calls are included in these records. Counts were also taken from the Outgoing Phone Log, summed by hour of year and averaged by hour of the day. Depending on the model being calculated, these counts may be segregated by Central, North, and South PSAPs. The processing time is the summation of the intervals from the [Agent Connected] timestamp to the [Agent Disconnect] timestamp. Additional processing time is the summation of the durations of outgoing calls taken from the Outgoing Phone Log. These processing times were summed by hour of year and averaged by hour of the day. Depending on the model being calculated, these processing durations may be segregated by Central, North, and South PSAPs. For calculation of workloads at the consolidated Intake workstations, counts and processing durations for all three PSAPs were combined. Assignment & Radio Support Workstations The count of incidents at the LAW and FIRE Assignment workstations were taken from the CAD, summed by hour of year and averaged by hour of day. The duration of assignment processing was taken as the interval from the [Transmit] timestamp until the [Dispatch] timestamp. These processing times were summed by hour of year and averaged by hour of day. Depending on the model being calculated, these processing durations may be segregated by Central, North, and South PSAPs. FIRE and LAW assignments were tabulated separately. All radio traffic in the Broward system is recorded. Broward did not make these records available to the consultant. Consequently, the exact radio traffic associated with each specific incident could not be determined. The workaround was based on a year-end written summary titled “Talkgroups at Zone Summary 150101 – 151231.” This document summarized total air time on each of the FIRE and LAW dispatch channels for the various jurisdictions. It was possible to obtain an incident count from the CAD for each LAW and FIRE jurisdiction. Combining total annual air time with total annual incident counts per jurisdiction permitted calculation of an average amount of air time per incident for all of the FIRE and LAW jurisdictions. These results were stored by agency in a dedicated data table titled “Talkgroup Stats.” Summing these average air times per incident by hour of year and averaging by hour of day provided the best estimate of the contribution of radio traffic workloads to total workload at the Assignment & Radio Support workstations. Depending on the model being calculated, these workloads may be segregated by Central, North, and South PSAPs.

Broward County Attachment E - Page 2 © Fitch & Associates Assessment of Broward County’s Regional E911 December 2016

ID Unit / FR Assign Workstation In Option 2, the CAD software has been upgraded with the ability to identify the unit(s) most appropriate for executing the response. The time interval for this identification process is assumed to be 1.5 seconds, based on FITCH’s experience with identification software in other large metropolitan systems. The recommended units are presented to the FR Assignment Dispatcher for review and notification. The manual review and notification is assumed to require 15 seconds, again, based on FI>TCH’s experience in other large metropolitan systems. FR msr Radio Workstation This workstation provides radio support for “multi-station response” incidents. These incidents comprise FIRE responses requiring the participation of units from more than one station. The average Time-on-Task for “msr” incidents in the Broward CAD is 00:24:52 [hh:mm:ss]. The radio support dispatcher at this workstation is assumed to be 100% dedicated to the incident for the first half of this interval ( 00:12:26 ), and 50% dedicated for the second half of the interval ( 00:06:13 ) for an average processing interval of 00:18:39 per msr incident. The function of this workstation is to provide tactical radio support on responses to complex FIRE incidents. Since crews do not require tactical radio support while loading to their apparatus, the latency used to model this workstation approximated the average chute interval experienced on these incidents.

FR ssr Radio Workstation This workstation provides radio support for “single station response” incidents. These incidents comprise emergency medical incidents plus the “simple” FIRE responses. The count of incidents was taken from the CAD, summed by hour of year, and averaged by hour of day. The workload was quantitated by multiplying the average count of incidents by the average duration of radio air time per incident as described in the section titled “Assignment & Radio Support Workstations”, above. In Option 2, the radio air time per incident was reduced to 60% of its current amount to reflect an increased use of MDTs compared to current practice.

Attachment F

Sample Phone Records

Broward County Attachment F - Page 1 © Fitch & Associates Assessment of Broward County’s Regional E911 December 2016

AttachmentF:SamplePhoneRecordandOutgoingPhoneLogTable1.SampleCDRPhoneRecord

Broward E911 Consolidated Communications SystemCall Detail Records Extended Format Begin Timestamp TimeDate

===== CDR BEGIN : 11/11/15 15:30:10.313 =====

0

===== CDR BEGIN : 11/11/15 15:30:10.313 ===== 00:00:00.000 [ TS] SYSTEM ID = broward 00:00:00.000 [ CIM] Incoming Call (ID: 911009-00291-20151111203010) Offered on Trunk 911009 00:00:02.269 [ CIM] ANI: (40)"9547295989" [VALID] PseudoANI: "" [NONE] 00:00:02.269 [ TS] Initial ALI Request for ANI : 9547295989 00:00:02.279 [ CIM] Call Presented 00:00:02.898 [VoIP] External Call-Identifier 911009-00291-20151111203010 00:00:03.100 [VoIP] Routing call QUEUE = 6023 00:00:03.309 [ PAS] Initial ALI Response received / ALI TYPE = 1 00:00:03.315 [VoIP] Routing call QUEUE = 6020 00:00:03.523 [ CIM] Call Connected 00:00:03.528 [VoIP] Routing call AGENT = 15002/2012 00:00:04.061 [ TCI] TRUNK = 911009 / LINE = 9 POS = 012 / STN = 2012 00:00:04.061 [ TCI] CALL CONNECTED BY AGENT = Adrian, Andrea/15002 ROLE = Central Call-Taker 00:00:04.061 [ TCI] From PSAP ID = 1 PSAP Name = Central 00:00:43.055 [ CIM] Tandem Transfer 00:00:44.552 [ TCI] Event Logged By POS = 012 / STN = 2012 KEY: TRANSFER SV: 77 LV:h,9547644357 00:00:44.552 [ TCI] agencyId: 471 agencyName: BROWARD COUNTY NON-EMERGENCYagencyTypeId: 9 agencyTypeName: Non-Emer 00:00:48.917 [ TCI] CALL RELEASED BY POS = 012 / STN = 2012 00:00:48.917 [ TCI] CALL DISCONNECTED BY AGENT = Adrian, Andrea/15002 ROLE = CentralCall-Taker 00:00:48.917 [ TCI] From PSAP ID = 1 PSAP Name = Central 00:00:50.743 [ CIM] Call Disconnected 00:00:50.753 [ CIM] Call Terminated 00:00:50.753 [ TS] Call Completed ===== Initial ALI ==== (954) 729-5989 15:30 11/11 8320 W SUNRISE BLVD PLANTATION FL 470 WPH2 SPRINT N SECTOR P# 729-5989 ALT# 954-816-8962 LEC:SPPCS WIRELESS CALL QUERY CALLER FOR LOCATION QUERY CALLER FOR PHONE # -080.256994 +26.162771 ===== CDR END =====

WIRELESSOrigin

11/11/2015 15:30:10.313

00:00:04.061 Agent Connected

15:30:10.31311/11/2015

CentralPSAP

911009Trunk

9548168962Phone_Number

-80.256994 Longitude26.162771 Latitude

00:00:48.917 Agent Disconnected

0Before

1,822

0 On-HoldOff-Hold0

0

0

ParkedUnparked

Char

00:00:03.523 Call_Connected

Caller Disconnects: After Supervision

Answer Delay

00:00:44.856 Agent Processsing

CDR Text Length

Mo Day DayName

Dayof Wk

Hourof Yr

Hr ofDay

11 11 Wed 4 15 7,5522015

Year

0TTY

44.856 sec

1.254

sec

15002Agent

2012 Station

Adrian, AndreaName

1.735

20.223

avg± sdsec

sec

11/11/2015 15:30:14.374 Agent_Connected_TS11/11/2015 15:30:59.23 Agent_Disconnected_TS

sec

00:00:02.269 [CIM] ANICaller Disconnects

11/11/2015 15:30:13.836 Call_Connected_TS

DNIS Alarm0

Call_Presented00:00:02.279

secsec

sec 90th95th

Broward County Attachment F - Page 2 © Fitch & Associates Assessment of Broward County’s Regional E911 December 2016

Table2.SampleofOutgoingPhoneLogs

Broward E911 Consolidated Communications SystemOutgoing Phone Logs 11/01/2015 - 01/31/2016

DateTimestampDialed Number ProcessingPSAP HoD

12/10/201512/10/2015 01:04:35Central 11.176(954) 279-0070 1 sec

12/10/201512/10/2015 01:06:25Central 10.188(954) 260-8290 1 sec

12/10/201512/10/2015 01:06:31South 3.436(954) 295-2251 1 sec

12/10/201512/10/2015 01:09:39North 756.561(800) 323-9949 1 sec

12/10/201512/10/2015 01:17:21South 2.206(954) 927-5287 1 sec

12/10/201512/10/2015 01:20:17Central 3.804(786) 487-7286 1 sec

12/10/201512/10/2015 01:24:00North 14.468(772) 626-7768 1 sec

12/10/201512/10/2015 01:25:02Central 258.527(786) 312-0238 1 sec

12/10/201512/10/2015 01:39:56Central 16.657(754) 423-5752 1 sec

12/10/201512/10/2015 01:45:40Central 38.065(954) 439-1070 1 sec

12/10/201512/10/2015 01:49:54South 7.559(718) 427-4308 1 sec

12/10/201512/10/2015 01:59:22South 6.804(754) 779-9183 1 sec

12/10/201512/10/2015 02:16:04Central 6.867(904) 236-2138 2 sec

12/10/201512/10/2015 02:16:31Central 32.045(954) 706-1753 2 sec

12/10/201512/10/2015 02:17:37South 480.740(786) 539-8293 2 sec

12/10/201512/10/2015 02:18:25North 57.829(772) 501-3443 2 sec

12/10/201512/10/2015 02:19:20Central 5.420(754) 322-8350 2 sec

12/10/201512/10/2015 02:19:34Central 46.076(754) 321-0161 2 sec

12/10/201512/10/2015 02:19:57South 4.558(786) 985-0380 2 sec

12/10/201512/10/2015 02:26:54North 8.784(772) 501-3443 2 sec

12/10/201512/10/2015 02:27:59South 2.232(954) 650-1660 2 sec

12/10/201512/10/2015 02:28:30North 33.352(954) 650-1660 2 sec

12/10/201512/10/2015 02:47:13South 93.104(954) 524-6991 2 sec

12/10/201512/10/2015 02:48:25Central 23.403(954) 235-9273 2 sec

12/10/201512/10/2015 02:48:28North 36.985(954) 971-7749 2 sec

12/10/201512/10/2015 02:57:32South 14.205(832) 335-7572 2 sec

12/10/201512/10/2015 03:18:51Central 306.468(954) 960-2463 3 sec

12/10/201512/10/2015 03:25:34South 4.157(954) 454-1472 3 sec

12/10/201512/10/2015 03:37:23Central 316.004(954) 268-4639 3 sec

12/10/201512/10/2015 03:42:15Central 2.145(912) 412-8662 3 sec

12/10/201512/10/2015 03:42:15Central 13.111(954) 245-2606 3 sec

12/10/201512/10/2015 03:52:41North 51.182(954) 609-4031 3 sec

12/10/201512/10/2015 03:55:45Central 397.942(561) 368-8462 3 sec

Attachment G

Sample Records from FIRE & LAW CADs & Radio Statistics

Broward County Attachment G - Page 1 © Fitch & Associates Assessment of Broward County’s Regional E911 December 2016

AttachmentG:SampleRecordsformFIRE&LAWCADs&RadioStatisticsTable1.SampleRecordsfromLawCAD

Broward E911 Consolidated Communications SystemLAW Incident Records

12/10/2015

Date

13:01:45

Time

12 10 Thu 5 13 8,246

Mo Day DayName

Dayof Wk

Hourof Yr

Hr ofDay

4900 W OAKLAND PARK BLVD

04

AOA76

Rcvd_timeCreate_time

Arrvd_timeClosed_time

L04151210000721

Event_number

2

Dispatch_time

12/10/2015 13:00:35

12/10/2015 13:01:45

12/10/2015 13:06:10

12/10/2015 13:09:46

12/10/2015 13:59:28

Transmit_time 12/10/2015 13:02:23

CentralBSO

Code Disp_LocResponse_Agency

Incident_IDPriority Incident_Description

Lauderdale Lakes

00:53:18Time-on-Task

BSO-07-DISP

Radio_Chn

00:03:47Assign_proc

Avg Talk/Xmit/Rcv

Avg Talk/Incident

9.55251.42 secsec

9546245921

Caller Phone

12/10/2015 13:00:28.768

Xmit/Rcv's/ Incident

5.38 41.871

SupportDuration

68 sec

sec

4B6

Unit

Enroute_time 12/10/2015 13:06:12.4

CDR_Begin

12/10/2015 13:00:32.973Agent Connected

12/10/2015 13:08:48.92Agent Disconnected

Portal to CDR Data Table

12/10/2015 13:00:32.265Call_Connected

DispatchDuration

9.552 sec

12/10/2015 13:00:32.97312/10/2015 13:00:28.768 12/10/2015 13:00:32.265 12/10/2015 13:08:48.92

0

Dup'd

CDR_Begin Agent_ConnectedCall_Connected Agent_Disconnected

P2

P3

00:01:50.027Intake_proc00:00:02.027VIPER_spillover

00:01:10Rcvd_offset

Assign_workloadP2 / P3335P2/P3 Interval

3Index

1

23

4

335

143.792avg

sec

sec


Table2.SampleRecordfromFIRECAD

Broward E911 Consolidated Communications SystemLAW Incident Records

12/10/2015

Date

13:01:45

Time

12 10 Thu 5 13 8,246

Mo Day DayName

Dayof Wk

Hourof Yr

Hr ofDay

4900 W OAKLAND PARK BLVD

04

AOA76

Rcvd_timeCreate_time

Arrvd_timeClosed_time

L04151210000721

Event_number

2

Dispatch_time

12/10/2015 13:00:35

12/10/2015 13:01:45

12/10/2015 13:06:10

12/10/2015 13:09:46

12/10/2015 13:59:28

Transmit_time 12/10/2015 13:02:23

CentralBSO

Code Disp_LocResponse_Agency

Incident_IDPriority Incident_Description

Lauderdale Lakes

00:53:18Time-on-Task

BSO-07-DISP

Radio_Chn

00:03:47Assign_proc

Avg Talk/Xmit/Rcv

Avg Talk/Incident

9.55251.42 secsec

9546245921

Caller Phone

12/10/2015 13:00:28.768

Xmit/Rcv's/ Incident

5.38 41.871

SupportDuration

68 sec

sec

4B6

Unit

Enroute_time 12/10/2015 13:06:12.4

CDR_Begin

12/10/2015 13:00:32.973Agent Connected

12/10/2015 13:08:48.92Agent Disconnected

Portal to CDR Data Table

12/10/2015 13:00:32.265Call_Connected

DispatchDuration

9.552 sec

12/10/2015 13:00:32.97312/10/2015 13:00:28.768 12/10/2015 13:00:32.265 12/10/2015 13:08:48.92

0

Dup'd

CDR_Begin Agent_ConnectedCall_Connected Agent_Disconnected

P2

P3

00:01:50.027Intake_proc00:00:02.027VIPER_spillover

00:01:10Rcvd_offset

Assign_workloadP2 / P3335P2/P3 Interval

3Index

1

23

4

335

143.792avg

sec

sec


Table3.SampleRecordsfromRadioStatistics

TotalRadio

Duration[sec/yr]

RadioChannelAbbrev

RadioTraffic

Xmit/Rcv's

AverageTalk

Xmit/Rcv[sec]

IncidentCount

TotalTalk perIncident

[sec]

Broward E911 Consolidated Communications SystemTalkgroups at ZoneStatistics CY2015

TalkCount

perIncident

4,161,542BCF-D1 480,434 8.662 36,910 112.75 13.02

4,671,511BCF-D2 529,742 8.818 48,609 96.10 10.90

2,008,058BCF-D3 225,762 8.895 20,198 99.42 11.18

1,539,092BCF-D4 178,202 8.637 14,023 109.75 12.71

3,160,509BCF-D5 399,290 7.915 30,872 102.37 12.93

3,456,751BCF-D6 419,797 8.234 32,836 105.27 12.78

5,252,965FLF-DISP1 656,110 8.006 54,316 96.71 12.08

4,070,692BCF-D8 447,385 9.099 30,929 131.61 14.46

2,525,851BCF-D9 288,696 8.749 21,580 117.05 13.38

541,606BCL-COMM 58,333 9.285 14,940 36.25 3.90

6,571,478BSO-02-DISP 717,158 9.163 32,778 200.48 21.88

7,212,269BSO-03-DISP 792,390 9.102 27,162 265.53 29.17

4,794,358BSO-04-DISP 497,838 9.630 42,827 111.95 11.62

7,145,394BSO-05-DISP 696,625 10.257 73,975 96.59 9.42

7,364,307BSO-06-DISP 762,379 9.660 117,985 62.42 6.46

10,128,828BSO-07-DISP 1,060,432 9.552 196,973 51.42 5.38

7,975,457BSO-08-DISP 872,305 9.143 75,711 105.34 11.52

8,527,021BSO-10-DISP 1,004,798 8.486 74,434 114.56 13.50

14,618,781BSO-11-A1A2 1,605,483 9.106 123,425 118.44 13.01

8,310,381CKP-MAIN 916,167 9.071 77,097 107.79 11.88

1,058,964CSF-MAIN 125,578 8.433 10,590 100.00 11.86

16,514,932HW-P-A1A2 1,377,654 11.988 189,878 86.98 7.26

9,362,099PPP-MAIN 988,952 9.467 86,900 107.73 11.38

7,195,082SNP-DISP 825,080 8.720 79,207 90.84 10.42

251,695WMP-TAC-1 30,847 8.159 2,517 100.00 12.26

Attachment H

Calculation of Surges

Broward County Attachment H - Page 1 © Fitch & Associates Assessment of Broward County’s Regional E911 December 2016

Attachment H. Calculation of Surges Theoretical Treatment Emergency services communications centers dispatch responses to defined geographic areas, the service jurisdiction. At a given hour of the day, and from day-to-day, the number of people in the service jurisdiction will be approximately the same. In turn, this condition leads to the historic observation that the number of requests for service will tend towards some daily average in that hour of the day. The next historic observation is that the number of requests in any particular day will vary above and below this long term average. As it turns out, the excursions to higher or lower numbers of requests really are random. The randomness of the excursions is very important because it makes the analyses of the flow of requests much simpler. Random processes are often characterized by statisticians using a “normal” distribution. A stylized example of a normal distribution is presented in Figure 1. Figure 1. Normal Distribution of Requests per Hour

The interpretation of this figure starts with the x-axis, which is the number of incidents per hour. Go to the vertical line at 65 incidents per hour, follow it up to the blue curve. The height of the curve at 65 incidents per hour gives the number of instances, the number of days in which exactly 65 incidents were experienced in the 1800 hour. The average number of incidents per hour is 100. There are exactly 40 days in which 100 incidents occurred in the 1800 hour. The distribution curve in Figure 1 has a width. The standard deviation, symbol σ, characterizes this width. In this example, the standard deviation is 25. The area under the normal curve from zero to the average is shaded green. The green area is one half the area under the curve. In the context of a dispatch center, the green area means


that one day out of two, there will be 100 incidents, or fewer, in the 1800 hour. Conversely, one day out of two, there will be 100 incidents, or more, in the 1800 hour. The valuable of the standard deviation, σ, is that it allows the extraction of the size and frequency of surges from the normal distribution. Consider Figure 2 where the green area has been extended to the right as far as [average + 1.28 σ] which happens to be 132 incidents per hour. Figure 2. Normal Distribution Showing a One Day in Ten Surge.

The green area now comprises 90% of the area under the normal curve. In the context of a dispatch center, the green area means that nine days in ten there will be 132 incidents, or fewer, in the 1800 hour. Conversely, one day in ten there will be 132 incidents, or more, in the 1800 hour. In Figure 3, below, the green area has been extended further right to [average + 2.33 σ] or 158 incidents per hour. The green area now comprises 99% of the area under the normal curve. In the context of a dispatch center, the green area now means the ninety-nine days out of one hundred there will be 158 incidents, or fewer, in the 1800 hour. On one day out of one hundred, there will be 158 incidents, or more, in the 1800 hour.


Figure 3. Normal Distribution Showing a One Day in One Hundred Surge.

The preceding discussion shows the usefulness of the standard deviation to answer questions of surges in dispatch systems. Once a collection of random incident counts has been converted to an average and a standard deviation, it becomes possible to conveniently extract the frequency and sizes of surges from the original set of data, at least in theory. Real Example Figure 4, below, presents the number of incidents per hour experienced at a large metropolitan dispatch center at 1600 hours. One year’s worth of data is included in the histogram. As can be seen, the day-to-day variability is substantial with a minimum of 12 incidents per hour to a maximum of 50 incidents per hour.


Figure 4. Incident per Hour at 1600 Hours

The data in Figure 4 was then consolidated into Figure 5. The process of this consolidation is referred to as “binning”. All of the instances where 12 or 13 incidents per hour occurred were counted and the total placed in a “bin” labelled 12-13, and so forth. The outcome of this binning process results in the distribution presented in Figure 5, below. As can be seen, the envelope, or shape, of the distribution of incidents per hour derived from the real data is not as smooth as the distributions treated in the theoretical section above. Figure 5. Distribution of Incidents per Hour

0 30 60 90 120 150 180 210 240 270 300 330 3600

5

10

15

20

25

30

35

40

45

50

Day of Year

Inci

dent

s pe

r Hou

r

0 5 10 15 20 25 30 35 40 45 50 55 600

5

10

15

20

25

30

35

40

45

50

55

Incidents/Hour

Instances


Numerical methods were next used to calculate the normal distribution curve that most closely follows the contour of the real distribution. The calculated normal distribution is presented in Figure 6, below. Three specific surge limits are specified in Figure 6. The values of these Figure 6. Normal Distribution Most Closely Conforming Figure 5.

surge limits are presented in Table 1, below. The surge limits may also be discussed in terms of the percentile contributions to the area under the normal curve. Table 1. Surge Limits Derived from Figure 6.

Frequency Offset [σ ]

Incidents per Hour %-tile Average Increment Total One Day in 2 0.00 σ 27.62 0.00 27.62 50th One Day in 10 +1.28 σ 27.62 8.36 35.98 90th One Day in 30 +1.84 σ 27.62 12.02 39.64 97th One Day in 100 +2.33 σ 27.62 15.21 42.83 99th

In Figure 7, the calculated normal distribution overlays the distribution of real data.

0 5 10 15 20 25 30 35 40 45 50 55 600

5

10

15

20

25

30

35

40

45

50

55

Incidents/Hour

Instances

Avg = 27.62 & σ = 6.53

1.28 σ

1.84 σ

2.33 σ


Figure 7. Comparison of the Real Distribution to a Normal Curve.

As ‘lumpy” as the real distribution may appear, it is a respectable approximation of a precisely calculated normal curve. Surges calculated using the theoretical methods described in this section are a good approximation of reality.

0 5 10 15 20 25 30 35 40 45 50 55 600

5

10

15

20

25

30

35

40

45

50

55

Incidents/Hour

Instances

Avg = 27.62 & σ = 6.53

1.28 σ

1.84 σ

2.33 σ

Attachment I

Erlang Tables of Workloads

Broward County Attachment I - Page 1 © Fitch & Associates Assessment of Broward County’s Regional E911 December 2016

Attachment I. Erlang Tables of Workstations Table 1, below, provides an index of Erlang Tables that describe the performance of the workstations comprising the three models of dispatch operations considered in this report. Table 1. Index to Erlang Tables and Models of Dispatch Operations

Figure Workstation Option 0 1 2

01# Central Intake X 02# North Intake X 03# South Intake X 04# Consolidated Intake X X 05# Central LAW Assign & Radio @ 1.00 Traffic X X X 06# Central FIRE Assign & Radio @ 1.00 Traffic X X 07# North LAW Assign & Radio @ 1.00 Traffic X X X 08# North FIRE Assign & Radio @ 1.00 Traffic X X 09# South LAW Assign & Radio @ 1.00 Traffic X X X 10# South FIRE Assign & Radio @ 1.00 Traffic X X 11# Consolidated LAW Assign & Radio @ 1.00 Traffic 12# Consolidated FIRE Assign & Radio @ 1.00 Traffic 13# Consolidated LAW Assign & Radio @ 0.60 Traffic 14# Consolidated FIRE Assign & Radio @ 0.60 Traffic 15# Consolidated FIRE Gatekeeper X 16# Consolidated FR msr Radio X 17# Consolidated FR ssr Radio @ 0.60 Traffic X

As indicated by the grey cells in Table 1, above, various workstations appear in more than one of the Options for the conduct of dispatch operations. In all of the Erlang tables below, the empirical adjustments of dispatcher Hours-OnTask by hour-of-day occurred in two stages. First, a workload surge experienced one day in ten, +1.28σ, was added to all hour-of-day. The number of dispatcher OnDuty by hour of day were adjusted so that the calculated latencies conformed to the FITCH operational targets. This number of dispatcher OnDuty by hour-of-day was retained, but the workloads were returned to their average values, +0.00σ. The statistics describing answer delays and probabilities of immediate answer reflect this design process in all models. The end result is a judicious “over-staffing” at all workstations. Stakeholder may decide that other surge levels and latencies should be built into the models. This certainly could be done. The value of these models is that they demonstrate a process that makes a quantitative connection between manpower deployed and the performance to be expected.


Table 2. Central Intake Workstation


87665567999

10101010

10101211

9988

10

97.4897.3396.6098.0096.4496.0697.7393.6996.3893.9392.1493.7193.1193.1692.80

92.7492.1194.8895.8692.8294.4092.6494.08

92.16

Surge tested to +1.28 std dev at all hours of day.

Central Intake 1

63.7650.9640.5034.9231.5431.4042.0568.0584.9792.39

101.73

117.60124.32130.46132.68132.88138.77145.08125.20110.58102.88

90.1575.87

2.9782.4982.1141.8761.6331.6791.9293.0763.7494.2684.728

5.1835.1665.2925.5595.3165.5796.0125.0474.5354.1663.8853.613

00:0001:0002:0003:0004:0005:0006:0007:0008:0009:0010:0011:0012:0013:0014:0015:0016:0017:0018:0019:0020:0021:0022:0023:00

111.80 4.996


21.3718.0914.1212.6711.8811.5317.2429.5737.0243.1550.92

52.0152.5953.9757.7550.1550.6750.7742.3533.7829.7128.5528.37

53.43

93.99

Req'd HrsOnTask

13.4810.82

8.507.476.886.828.87

15.6619.3822.0124.38

24.8426.2627.5526.0227.2425.0724.2222.2221.3619.5417.2115.95

25.27

Option 0 (Current Ops)2015Year Dispatch Model Workstation Name

90.86 3.95335.49 18.63

AverageErlangs

Hourly Averages

Hourly Averages

1.59

Ans Delay @

1.943.071.733.594.181.654.571.913.635.143.453.993.763.994.754.104.862.441.934.603.325.354.24

OnTask


Weighted% Immed Ans

WeightedAns Delay

204 3.67


ADM Out911

0.00Surge

σ+Surge


Table 3. North Intake Workstation


544444456667777

77777665

7

96.5895.7397.6697.9198.3297.9796.4295.8795.1292.7790.9294.7094.1294.1992.85

92.4793.3292.0294.5596.0194.5995.9094.19

92.91


North Intake 2

20.3415.2613.3611.57

9.7310.2914.3523.1828.2431.5735.55

42.5040.5942.4343.4243.0144.3847.7240.5134.4232.5428.2524.20

1.6161.2161.0080.9740.9130.9651.1481.7012.3162.6072.840

3.0123.0023.2033.1943.2623.1323.3342.9482.7262.3832.2131.882

00:0001:0002:0003:0004:0005:0006:0007:0008:0009:0010:0011:0012:0013:0014:0015:0016:0017:0018:0019:0020:0021:0022:0023:00

36.23 2.926


22.2415.6312.8813.5011.7211.8317.4026.5338.8945.1648.27

49.7049.9655.0254.4856.5853.4654.8946.6140.7536.6732.4925.55

49.82

94.16

Req'd HrsOnTask

5.654.673.593.142.863.875.088.08

10.5710.8010.78

11.6811.1112.1613.1612.9711.2111.8711.35

9.809.328.286.99

11.05


29.74 2.27236.25 8.75

AverageErlangs

Hourly Averages

Hourly Averages

3.54

Ans Delay @

5.492.772.492.142.594.103.854.136.649.024.104.474.495.755.616.115.196.634.213.144.743.636.46

OnTask


Weighted% Immed Ans

WeightedAns Delay

139 5.03


ADM Out911

0.00Surge

σ+Surge


Table 4. South Intake Workstation


544444456777888

88887765

8

96.3394.7797.0098.6898.9998.6497.4995.8595.4196.3395.4894.7496.2396.1896.34

94.9394.5394.1796.8695.9196.5595.9294.98

96.57


South Intake 3

22.3916.2513.9812.1510.6311.8913.6122.5829.6334.1838.71

47.8447.7147.9647.9349.2550.0851.1244.5039.4036.8630.1726.02

1.6471.3011.0870.8520.7880.8591.0281.7032.2782.6742.808

3.2273.2363.2083.1643.4613.5323.5973.1072.7412.6372.2101.800

00:0001:0002:0003:0004:0005:0006:0007:0008:0009:0010:0011:0012:0013:0014:0015:0016:0017:0018:0019:0020:0021:0022:0023:00

42.59 2.920


19.4016.4812.10

8.418.309.49

15.0327.2337.3043.7946.42

51.3750.6551.2544.3848.3646.9045.1837.2331.9829.7926.7722.99

45.10

95.86

Req'd HrsOnTask

8.256.575.294.174.174.084.388.36

10.1612.0213.61

15.5515.9616.1618.3818.2017.3816.1015.0414.2112.8410.33

9.75

14.93


32.81 2.32832.33 11.50

AverageErlangs

Hourly Averages

Hourly Averages

3.77

Ans Delay @

6.723.731.511.121.532.753.853.812.643.213.842.322.372.222.123.493.954.442.153.222.743.705.03

OnTask


Weighted% Immed Ans

WeightedAns Delay

150 3.19


ADM Out911

0.00Surge

σ+Surge


Table 5. Consolidated Intake


1110

98778

1012141415151516

1617161514131211

16

92.6393.6594.2093.5992.0291.0491.6390.8792.0693.2193.0493.5393.4693.4793.91

93.8894.2993.8293.5093.1092.6192.0791.55

93.89

Surge tested to +1.28 std dev at all hour of day.

Consolidated Intake 4

106.4982.4767.8458.6451.9053.5970.01

113.82142.84158.14175.99

207.93212.61220.85224.04225.14233.23243.91210.21184.40172.28148.58126.09

6.2405.0154.2083.7023.3353.5034.1056.4808.3439.549

10.375

11.42311.40511.70311.91712.04012.24312.94311.10210.002

9.1858.3087.295

00:0001:0002:0003:0004:0005:0006:0007:0008:0009:0010:0011:0012:0013:0014:0015:0016:0017:0018:0019:0020:0021:0022:0023:00

190.62 10.842


63.0150.2039.1034.5931.9032.8549.6783.33

113.22132.11145.62

153.08153.20160.24156.61155.09151.03150.85126.18106.51

96.1787.8276.91

148.35

93.18

Req'd HrsOnTask

27.3822.0517.3814.7813.9114.7618.3332.1040.1144.8448.77

52.0853.3355.8857.5758.4053.6552.1848.6145.3741.7035.8232.68

51.26

Option 1, & 22015Year Dispatch Model Workstation Name

153.40 8.553104.07 38.87

AverageErlangs

Hourly Averages

Hourly Averages

1.77

Ans Delay @

1.491.471.842.673.192.302.642.201.561.941.561.821.781.371.471.531.212.111.731.852.062.362.54

OnTask


Weighted% Immed Ans

WeightedAns Delay

301 1.83


ADM Out911

0.00Surge

σ+Surge


Table 6. Central LAW Assign & Radio


333322334444445

55554444

5

83.6188.7393.0694.4782.7784.6294.1382.2786.5783.5682.2881.1381.0780.6287.43

86.6686.3786.7388.4084.6186.2187.5890.64

86.31


Central LAW Assign & Radio 5

1.4111.1430.9090.8230.7580.7010.8451.4851.9302.1922.321

2.4612.5182.5762.7172.6712.7102.6622.4642.0951.9591.8511.621

00:0001:0002:0003:0004:0005:0006:0007:0008:0009:0010:0011:0012:0013:0014:0015:0016:0017:0018:0019:0020:0021:0022:0023:00

2.454


32.0126.5321.4719.1617.0215.5418.8533.6544.0449.9853.35

55.6657.0558.4160.3559.5659.2858.5254.5947.8745.4042.4337.00

55.76

85.65

Req'd HrsOnTask

314.60259.68203.74189.70174.25153.27181.63322.58416.34474.74508.71

530.95551.44560.91584.99575.07579.27563.94523.74451.34426.37398.92353.00

527.47

Option 0, 1, & 22015Year Dispatch Model Workstation Name

0.00 1.88742.64 409.44

AverageErlangs

Hourly Averages

Hourly Averages

4.39

Ans Delay @

2.531.401.055.755.141.205.102.843.984.565.375.405.662.262.642.522.642.542.043.542.932.541.71

OnTask


Weighted% Immed Ans

WeightedAns Delay

92 3.37

95 %-tile [sec]LAW Radio

LAW Radio

0.00Surge

σ+Surge


Table 7. Central FIRE Assign & Radio


211111122222222

22222222

2

96.5274.5576.8078.6080.1578.8674.7895.2392.7790.9989.5089.5289.2589.4889.59

89.6289.2589.4190.8591.5292.2693.5395.13

88.85

Surge tested to +1.28 std dev at all hour of day.

Central FIRE Assign & Radio 6

9.968.848.057.446.937.328.69

11.8515.0017.1418.79

19.0118.7218.5519.1418.3718.7218.5716.9616.3115.4513.9311.96

0.2870.2550.2320.2140.1990.2110.2520.3420.4360.4970.547

0.5550.5470.5440.5680.5430.5550.5500.5020.4790.4540.4080.347

00:0001:0002:0003:0004:0005:0006:0007:0008:0009:0010:0011:0012:0013:0014:0015:0016:0017:0018:0019:0020:0021:0022:0023:00

18.73 0.546


89.00

Req'd HrsOnTask

108.5796.5687.7481.1575.5179.9994.39

129.07163.64186.62204.97

207.54204.27202.36208.99200.30204.37202.47185.08178.08168.45152.07130.68

204.29

Option 0 & 12015Year Dispatch Model Workstation Name

14.35 0.4190.00 156.55

AverageErlangs

Hourly Averages

Hourly Averages

0.18

Ans Delay @

2.972.632.372.152.342.970.250.410.530.640.640.660.640.630.700.640.670.650.550.490.440.360.26

OnTask


Weighted% Immed Ans

WeightedAns Delay

42 0.83

5 %-tile [sec]E Radio

RadioE

0.00Surge

σ+Surge


Table 7. North LAW Assign & Radio


222222224333433

33333332

3

85.4490.4493.3995.3596.4795.9493.1884.3698.1590.7089.3488.0395.2787.2086.48

85.4084.1184.4587.6089.7991.4292.8181.31

85.53


North LAW Assign & Radio 7

0.6750.5160.4130.3380.2900.3130.4210.7090.9391.0391.111

1.2571.2221.2601.3091.3161.3841.3661.2021.0871.0000.9230.803

00:0001:0002:0003:0004:0005:0006:0007:0008:0009:0010:0011:0012:0013:0014:0015:0016:0017:0018:0019:0020:0021:0022:0023:00

1.179


15.3012.06

9.517.926.756.899.05

15.4420.1622.1623.55

26.0625.4626.1527.3727.2528.1127.9525.4923.3621.6420.1118.02

24.76

88.77

Req'd HrsOnTask

173.22136.58108.15

90.3376.7478.57

103.00174.54227.72250.42266.36

294.86287.74295.62309.37308.29318.32315.92288.02264.39245.19227.73204.19

280.08


0.00 0.92019.60 221.89

AverageErlangs

Hourly Averages

Hourly Averages

4.12

Ans Delay @

2.341.531.010.750.921.704.720.241.892.262.660.712.943.183.483.564.113.962.772.111.681.355.90

OnTask


Weighted% Immed Ans

WeightedAns Delay

65 2.69


LAW Radio

0.00Surge

σ+Surge


Table 8. North FIRE Assign & Radio


111111111122222

22221111

2

85.7887.4988.9989.2289.6889.1487.1382.1977.4976.0897.1597.1797.1796.8997.05

97.1797.2297.2097.6578.0879.7581.7683.45

97.21


North FIRE Assign & Radio 8

4.654.123.613.523.343.534.195.967.517.948.51

8.488.908.608.438.508.398.427.707.316.736.085.40

0.1420.1250.1100.1080.1030.1090.1290.1780.2250.2390.257

0.2560.2700.2620.2540.2560.2540.2550.2320.2190.2020.1820.166

00:0001:0002:0003:0004:0005:0006:0007:0008:0009:0010:0011:0012:0013:0014:0015:0016:0017:0018:0019:0020:0021:0022:0023:00

8.45 0.256


90.39

Req'd HrsOnTask

55.1448.8442.7741.6939.5541.8749.7070.6588.9894.08

100.92

100.51105.50101.99

99.97100.84

99.4699.8691.3386.7279.8472.1464.05

100.13


6.59 0.2000.00 78.19

AverageErlangs

Hourly Averages

Hourly Averages

4.12

Ans Delay @

3.533.073.022.893.053.695.277.097.710.400.400.400.440.420.390.400.390.390.336.846.215.444.94

OnTask


Weighted% Immed Ans

WeightedAns Delay

34 2.67


RadioE

0.00Surge

σ+Surge


Table 9. South LAW Assign & Radio


322222233444444

44444333

4

91.1781.8187.2190.1392.3892.8989.5591.2284.7192.2490.2488.4687.9087.6486.65

86.6086.4886.9589.2891.7884.1586.5088.65

85.80


South LAW Assign & Radio 9

1.0140.7870.6200.5260.4490.4310.5451.0111.3521.5021.651

1.8261.8461.9241.9931.9271.9371.9001.7221.5371.3821.2591.147

00:0001:0002:0003:0004:0005:0006:0007:0008:0009:0010:0011:0012:0013:0014:0015:0016:0017:0018:0019:0020:0021:0022:0023:00

1.783


26.3921.6317.3914.9412.7111.9414.2225.0433.4636.6939.88

42.6843.3544.3346.6545.3445.0844.3841.4137.5333.4730.7528.82

42.24

88.02

Req'd HrsOnTask

245.61201.45160.85138.95116.90111.15133.58240.12317.68345.86376.05

402.38409.80421.44438.81426.75422.01418.72393.42356.79315.73288.85266.46

399.05


0.00 1.33632.51 306.18

AverageErlangs

Hourly Averages

Hourly Averages

1.74

Ans Delay @

5.533.372.391.781.662.771.763.741.281.732.202.392.452.783.042.762.842.671.951.364.053.192.49

OnTask


Weighted% Immed Ans

WeightedAns Delay

78 2.56


LAW Radio

0.00Surge

σ+Surge


Table 10. South FIRE Assign & Radio


111111122222222

22222221

2

79.0881.8384.5284.7585.6284.3881.7697.3895.4494.1593.5493.3392.5293.1292.79

93.6793.7593.8894.6695.1996.0796.7576.14

93.19


South Assign FIRE & Radio 10

6.225.374.614.514.284.565.387.339.86

11.3111.95

13.1012.5212.7912.4711.9711.9011.7510.8510.27

9.148.267.10

0.2090.1820.1550.1530.1440.1560.1820.2460.3340.3850.408

0.4450.4230.4350.4200.4030.4000.3950.3650.3440.3070.2760.239

00:0001:0002:0003:0004:0005:0006:0007:0008:0009:0010:0011:0012:0013:0014:0015:0016:0017:0018:0019:0020:0021:0022:0023:00

12.32 0.415


91.76

Req'd HrsOnTask

75.5265.3155.8654.9351.7255.6465.6288.66

119.78137.54145.27

159.31151.67155.40151.53145.35144.45142.23131.82124.43111.09

99.9386.25

149.49


9.16 0.3090.00 111.20

AverageErlangs

Hourly Averages

Hourly Averages

7.08

Ans Delay @

5.974.914.834.515.026.000.400.740.981.101.131.301.181.251.161.061.051.020.870.780.620.508.38

OnTask


Weighted% Immed Ans

WeightedAns Delay

40 1.95


RadioE

0.00Surge

σ+Surge


Table 11. Consolidated LAW Assign & Radio @ 1.00 Traffic


554433456777788

88977666

8

83.9488.5786.4289.7582.0682.9988.1083.4584.8487.0586.6086.3186.2288.3488.20

88.0988.0289.8386.3387.0784.6385.2586.69

88.03


Consolidated LAW Assign & Radio 11

3.1012.4461.9421.6871.4961.4451.8113.2054.2214.7345.083

5.5445.5865.7606.0205.9156.0325.9295.3884.7194.3414.0323.571

00:0001:0002:0003:0004:0005:0006:0007:0008:0009:0010:0011:0012:0013:0014:0015:0016:0017:0018:0019:0020:0021:0022:0023:00

5.416


73.7060.2248.3742.0236.4734.3742.1274.1397.66

108.83116.78

124.40125.87128.89134.37132.15132.47130.85121.48108.77100.51

93.2883.85

122.76

86.82

Req'd HrsOnTask

733.43597.71472.74418.99367.89342.99418.20737.24961.75

1071.021151.12

1228.201248.981277.981333.181310.111319.601298.591205.181072.51

987.29915.50823.65

1206.60

2015Year Dispatch Model Workstation Name

0.00 4.14394.76 937.52

AverageErlangs

Hourly Averages

Hourly Averages

1.17

Ans Delay @

0.600.890.581.591.510.771.311.220.831.011.271.400.710.780.890.840.910.491.240.811.331.080.78

OnTask


Weighted% Immed Ans

WeightedAns Delay

148 0.98


LAW Radio

0.00Surge

σ+Surge


Table 12. Consolidated FIRE Assign & Radio @ 1.00 Traffic


222222223333333

33333322

3

86.6189.0591.0091.6792.5091.6188.9882.4991.5289.1487.4087.2986.5686.8586.83

87.5987.4687.6489.5790.6492.1079.2383.00

86.81


Consolidated FIRE Assign & Radio 12

20.8318.3316.2715.4614.5515.4218.2725.1432.3636.3839.25

40.5940.1439.9540.0438.8439.0138.7435.5133.9031.3328.2724.47

0.6390.5610.4970.4740.4450.4760.5630.7660.9951.1211.212

1.2551.2401.2411.2431.2021.2091.1991.0991.0420.9630.8670.751

00:0001:0002:0003:0004:0005:0006:0007:0008:0009:0010:0011:0012:0013:0014:0015:0016:0017:0018:0019:0020:0021:0022:0023:00

39.49 1.218


87.80

Req'd HrsOnTask

239.23210.71186.37177.77166.78177.50209.71288.38372.40418.24451.17

467.36461.44459.75460.49446.49448.27444.56408.23389.23359.38324.15280.99

453.91


30.10 0.9280.00 345.94

AverageErlangs

Hourly Averages

2.53

Ans Delay @

1.951.541.401.241.421.983.631.091.491.821.841.991.931.941.951.791.821.781.421.231.004.723.50

OnTask


Weighted% Immed Ans

WeightedAns Delay

62 1.94


RadioE

0.00Surge

σ+Surge


Table 13. Consolidated LAW Assign & Radio @ 0.60 Traffic


554433356666667

77766655

7

89.5394.7092.8795.3089.2389.7184.1888.5488.3186.4585.5884.9384.7384.6987.57

87.3387.1587.2984.9986.5187.8183.9986.29

87.19


Consolidated LAW Assign rRadio 13

2.3441.8291.4541.2551.1171.0911.3802.4493.2343.6333.900

4.2814.3024.4484.6484.5684.6754.5944.1493.6163.3263.0912.721

00:0001:0002:0003:0004:0005:0006:0007:0008:0009:0010:0011:0012:0013:0014:0015:0016:0017:0018:0019:0020:0021:0022:0023:00

4.176


73.7060.2248.3742.0236.4734.3742.1274.1397.66

108.83116.78

124.40125.87128.89134.37132.15132.47130.85121.48108.77100.51

93.2883.85

122.76

87.08

Req'd HrsOnTask

440.06358.62283.64251.39220.73205.79250.92442.34577.05642.61690.67

736.92749.39766.79799.91786.07791.76779.15723.11643.51592.37549.30494.19

723.96


0.00 3.17894.76 562.51

AverageErlangs

Hourly Averages

Hourly Averages

1.88

Ans Delay @

0.761.280.772.602.564.812.232.122.903.474.264.624.642.532.842.712.932.794.172.852.294.222.96

OnTask


Weighted% Immed Ans

WeightedAns Delay

131 3.05


LAW Radio

0.00Surge

σ+Surge


Table 14. Consolidated FIRE Assign & Radio @ 0.60 Traffic


222222222222222

22222222

2

93.4194.7395.7496.0896.5096.0494.6791.0986.4983.8181.8881.7880.9881.2881.20

82.0581.8882.0784.2185.4687.0889.0791.33

81.18


Consolidated FIRE Assign rRadio 14

20.8318.3316.2715.4614.5515.4218.2725.1432.3636.3839.25

40.5940.1439.9540.0438.8439.0138.7435.5133.9031.3328.2724.47

0.4130.3630.3210.3070.2880.3080.3650.4940.6430.7260.785

0.8130.8040.8060.8070.7800.7850.7790.7130.6750.6240.5610.486

00:0001:0002:0003:0004:0005:0006:0007:0008:0009:0010:0011:0012:0013:0014:0015:0016:0017:0018:0019:0020:0021:0022:0023:00

39.49 0.788


85.66

Req'd HrsOnTask

143.54126.43111.82106.66100.07106.50125.82173.03223.44250.95270.70

280.42276.87275.85276.29267.89268.96266.74244.94233.54215.63194.49168.59

272.35


30.10 0.6010.00 207.56

AverageErlangs

Hourly Averages

1.09

Ans Delay @

0.840.670.610.540.620.861.542.623.363.953.974.244.154.204.213.913.983.923.262.892.482.001.51

OnTask


Weighted% Immed Ans

WeightedAns Delay

48 3.00


RadioE

0.00Surge

σ+Surge


Table 15. Consolidated FIRE Gatekeeper


111111112222222

22222211

2

90.4591.6092.5492.9193.3392.9391.6388.4898.9998.7398.5498.5298.4498.4798.49

98.5798.5598.5798.7998.8999.0587.0488.79

98.48

Surge tested to +1.28 std dav at all hours of day.

Gatekeeper FIRE 15

20.8318.3316.2715.4614.5515.4218.2725.1432.3636.3839.25

40.5940.1439.9540.0438.8439.0138.7435.5133.9031.3328.2724.47

0.0950.0840.0750.0710.0670.0710.0840.1150.1480.1670.180

0.1860.1840.1830.1830.1780.1790.1780.1630.1550.1440.1300.112

00:0001:0002:0003:0004:0005:0006:0007:0008:0009:0010:0011:0012:0013:0014:0015:0016:0017:0018:0019:0020:0021:0022:0023:00

39.49 0.181


96.41

Req'd HrsOnTask

Option 22015Year Dispatch Model Workstation Name

30.10 0.1380.00 0.00

AverageErlangs

Hourly Averages

Hourly Averages

5.06

Ans Delay @

4.403.863.663.423.654.386.240.260.330.390.390.410.400.400.400.380.380.380.320.290.257.146.06

OnTask


Weighted% Immed Ans

WeightedAns Delay

38 1.64

95 %-tile [sec]E

E

0.00Surge

σ+Surge


Table 16. Consolidated FR msr Radio


222222233333333

34433333

3

91.3993.5894.9495.0796.0195.1591.7497.3393.6489.0188.8188.1489.2188.4987.73

90.6196.9096.6991.4292.8794.2496.2898.04

89.06


FR msr Radio 16

1.561.311.141.121.001.111.521.982.813.633.66

3.603.723.843.623.363.533.603.222.962.702.251.76

0.4840.4060.3540.3490.3100.3460.4720.6150.8751.1281.139

1.1181.1561.1951.1261.0441.0981.1201.0010.9200.8380.7000.546

00:0001:0002:0003:0004:0005:0006:0007:0008:0009:0010:0011:0012:0013:0014:0015:0016:0017:0018:0019:0020:0021:0022:0023:00

3.78 1.174


91.99

Req'd HrsOnTask


2.62 0.8130.00 0.00

AverageErlangs

Hourly Averages

Hourly Averages

184.62

Ans Delay @

130.93100.00

97.1776.8495.31

175.7736.4297.31

190.95195.42211.13186.52203.01221.02189.84156.09

34.7637.34

139.59111.49

86.6052.5725.91

OnTask


Weighted% Immed Ans

WeightedAns Delay

67 137.07

95 %-tile [sec] tac

tac

0.00Surge

σ+Surge


Table 17. Consolidated FR ssr Radio @ 0.60 Traffic


222222222222222

22222222

2

95.9496.7897.4297.6497.8997.6396.8394.3691.4489.8788.4888.3487.6588.0188.14

88.5088.5388.7390.1490.7891.9193.1494.57

88.02

Surge tested to +1.28 std dev at all hours of day n t is mo el air time on t e ra io c annels as een re ce to % of c rrent sage increase ata transfers ia mo ile ata terminals.

FR ssr Radio 17

0.3130.2750.2440.2320.2190.2330.2730.3770.4820.5350.579

0.6060.5950.5900.5940.5790.5780.5720.5260.5040.4660.4220.369

00:0001:0002:0003:0004:0005:0006:0007:0008:0009:0010:0011:0012:0013:0014:0015:0016:0017:0018:0019:0020:0021:0022:0023:00

0.584


90.95

Req'd HrsOnTask

132.58116.79103.45

98.5392.7998.52

115.20159.47203.91225.91245.07

255.93251.38249.36251.06244.88244.57241.85222.53213.52197.32178.87156.39

246.83


0.00 0.4480.00 189.45

AverageErlangs

Hourly Averages

Hourly Averages

0.20

Ans Delay @

0.160.120.110.100.110.160.300.480.590.690.700.760.730.720.730.690.690.670.570.520.450.370.28

OnTask


Weighted% Immed Ans

WeightedAns Delay

48 0.53

5 %-tile [sec]Radio

Radio

0.00Surge

σ+

++++++++++++++++++++++++

Surge

Monthly Performance Report Format

Attachment J

METRIC Count Average TARGETTARGET

COMPLIANCE

P1CallAnswerTime-BusyHourof1800hrs.(NENA56-005)

90%Within10Seconds

P1CallAnswerTime-NumberofDaysMeetingBusyHourPerformance

AllDaysinMonth

P1CallAnswerTime-All911Calls(NENA56-005)95%Within20

Seconds

P1CallAnswerTime-All911Calls(NFPA1221-2016)95%Within15

Seconds

P1CallAnswerTime-All911Calls(NFPA1221-2016)99%Within40

Seconds

P1CallAnswerTime-All911Calls(StateofFlorida)90%Within10

Seconds

P1CallAnswerTime-AlarmLines95%Within15

Seconds

P1CallAnswerTime-AlarmLines99%Within40

Seconds

TransfertoSecondaryPSAP(NFPA1221-2016)95%Within30

Seconds

P2EMSCallForServiceProcessingTime-Delta&EchoCallsOnly

90%Within70Seconds

P2LawEnforcementCallForServiceProcessingTime-Priority1&2CallsOnly

Report90th%-Nospecifictarget

P3EMSCallForServiceProcessingTime-Delta&EchoCallsOnly

90%Within20Seconds

P3LawEnforcementCallForServiceProcessingTime-Priority1&2CallsOnly


P2/P3EMS/SpecializedCallForServiceProcessingTime(NFPA1221-2016)

90%Within90Seconds

P2/P3EMS/SpecializedCallForServiceProcessingTime(NFPA1221-2016)

99%Within120Seconds

RECOMMENDEDMONTHLYPERFORMANCEMEASURES

P2/P3FireCallForServiceProcessingTime(NFPA1221-2013)

80%Within60Seconds


90%Within64Seconds


95%within106Seconds

P2/P3EMSCallForServiceProcessingTime-Delta&EchoCallsOnly

90%Within90Seconds

P2/P3LawEnforcementCallForServiceProcessingTime-Priority1&2CallsOnly


P4(newlydefined)EMSTurnoutTimes-Delta&EchoCallsOnly(NFPA1710-2016)


P5(newlydefined)EMS&FireTravelTimes-Delta&EchoCallsOnly(NFPA1710-2016)


P5(newlydefined)LawEnforcementTravelTimes-Priority1&2CallsOnly


EMDCaseEntryCompliance 95%

EMDTotalComplainceRate 90%

EMDQualityAssurance-CasesReviewed 1%

NOTES:

Busyhourdefinedas1800-1900hrs.

assessment of broward county’s … and recommendations of broward county’s regional e911...

Documents