Apresentation of

Humanfactorsinship designandoperation:Challengesrelatedtodigitalizationandautomation

based onAPhDthesisby Vincentius RumawasDepartmentofMarineTechnology

@HFC– HumanFactorsinControlMeeting:Digitalizationandautonomy–ChallengesandopportunitiesrelatedtohumaninteractionsTrondheim,17– 18Oct2017

2

Backgroundofthepresentation

• Allmaterialsinthispresentationarebasedonaresearchcontractconductedin2008– 2012.

• Thetitle:HumanFactorsinShipDesignandOperation:ExperientialLearning

• TheresearchwasfullyfinancedbytheDepartmentofMarineTechnology,NTNU

• ThethesiswasdefendedinJan2016.• Fulltextisavailableat:

– http://hdl.handle.net/11250/2382315– http://hdl.handle.net/11250/2382316

• Selectedtopicsrelevanttodigitalization&automationarepresentedhere.

3

Introduction

Background of the research

• Fatalaccidents atsea

• Caused byhumanerrors and/orhuman‐related factors (mostly)

• Humanfactorswere barely aconsiderationwhen designingaship

Howshipsaredesignedandbuilt

Basic Ship Theory, Rawson & Tupper 2001

The Ship Design Process, Gale 2003 in Lamb (Ed)

The concept of design spiral represents the sequential and iterative aspects of the process that include: Conceptual design Preliminary design Contract design Detailed design

Ship design is a complex and multifaceted process, influenced by a number of actors. … 

A successfully designed ship is the result of close and good cooperation between the designer, the customer, the yard and the equipment suppliers (Vossen et al 2013)

5

Researchquestions

1. Arehumanfactors considered inship design?How…?

2. Isthere any effect of implementing HF?

a. towards the crew

b. towardsincidencesonboard

3. AretheexistingknowledgeofHFeffective/sufficient?

4. Whatfactorsinfluencesafety&crewperformanceatsea?

5. Whatare“HF”inshipdesignandoperation?

6. Howtotakeintoaccountthe“HF”inmajorriskassessment?

6

Researchdesignandoutline

RQ1: Are HF considered  in ship design? How?RQ1: Are HF considered  in ship design? How?

RQ2: Is there any effect of  …RQ2: Is there any effect of  …

RQ3: Are the existing 

knowledge of HF effective/sufficient?

RQ3: Are the existing 

knowledge of HF effective/sufficient?

RQ4: What factors 

influence safety & performance at sea?

RQ4: What factors 

influence safety & performance at sea?

RQ5: What are “HF” in ship design & operation?

RQ5: What are “HF” in ship design & operation?

RQ6: How to account HF in major risk assessment?

RQ6: How to account HF in major risk assessment?

7

Some definitions

• “humanfactors”:

– “Ergonomics(orhumanfactors)isthescientificdisciplineconcernedwiththeunderstandingofinteractions amonghumansandotherelementsofasystem,andtheprofessionthatappliestheory,principles,data,andothermethodstodesigninordertooptimizehumanwell‐beingandoverallsystemperformance”(IEA,2012)

– “humanfactors”isconcernedwiththetaskpeopleperformandtheenvironmenttheydoitin– fittingthejobtotheperson.Thetopicofhumanfactorsisdividedintoeightconsiderations:habitability,maintainability,workability,controllability,manoeuvrability,survivability,occupationalhealthandsafety(OHS)andsystemsafety(LR,2008).

NB: “humanfactors”≠“humanelement”(physiological,psychological),“humanerror”,“humanperformance”,“HSE”,“humanreliability”.

8

Theresearch

• RQ1:

AreHFconsidered inship design?How?

There are several ways toanswer this:

Askthe designer,the shipyard and/orthe shipowner,

orcheck the designspecification/contract

Check andreview the ship itself (see Study 2)

Askthe users (see Study 2andStudy 3)

Consult the existing rules,regulations andstandardsavailable (Study 1)

9

Study 1.Literature study

Tocheckifhumanfactorsissuesaretakenintoaccountintheexistingstandards,asurveyofliteraturewasperformed.

Twoquestionsaretobeanswered:

What/whichpublicationscontainHF?

WhichaspectsofHFareaddressed/considered?

Results:

“AContentAnalysisofHumanFactorsintheDesignofMarineSystems”.TheInternationalConferenceonShipandOffshoreTechnology,11‐12Nov2010,Surabaya

“AContentAnalysisofHumanFactorsinShipsDesign”TheInternationalJournalofMaritimeEngineering,RINATransactionsPartA3,Vol156,Jul– Sep2014

RQ1.AreHFconsidered inship design?How?

10

Study 3.Qualitative study

Humanfactorsframework

derived from:Lloyd’s Register,2008,2009anddevelopedinRumawas &Asjbjørnslett 2010

Study 1.…Literature study

Results

• There are abundantdocuments coverHF• HFissues havebeen sufficiently

addressed.

• AllHFDIMENSIONS are covered• Mostmentioned: SYSTEM SAFETY

(highestfreq)• Mostextensively:HABITABILITY (COMFORT)

– Noise,vibration,indoor climate &lighting/illumination

• CONTROLLABILITY– Alarms,control centres,

workstations,control &switches

• Theleast covered:MAINTAINABILITY

• HFissues isdeveloping very fast

• Thedocuments are optional.

RQ1.AreHFconsidered inship design?How?

12

Study 2.Exploratory field study

Tocheck if the facts inreality isinaccordancewith the facts onpaper

Exploratory field surveyswere performed using qualitative approach,incl:

• Goonboard• Jointhetrips• Observations• Dointerviews• Discussions,focusgroup

NB: Rapportisimportant Actionresearch,participatory,asa‘naïveobservant’

RQ1.AreHFconsidered inship design?How?

13

Study 2.Exploratory …Before survey,some issues that were reported/found byother researchers onship designwere documented:

Accommodation facilities,crew expect adequate levels of privacy (Strong2000) Illumination problemson the bridge(Lutzhoft2005) Ergonomic issues;no legspace,incorrect height/orientation,must‐be‐fixed equipment

(Anderson&Lutzhoft2007,Grundevik2009) Problemwith access &personnel movement,incorrect control panel,console problem

(Dalpiaz etal2005).

Incidences&accidentsonOSVwerealsodocumented(Hansson2006,PSANorway2011):

• Personsqueezedbetweenmovingcontainers• Personhitintheheadbyamovinghook• Deckhandslip,fall,twistedafoot• Poorautopilotinterfacesystem*)• Collisionwithoffshoreinstallations*)

*)relatedtodigitalizationandautonomy

RQ1.AreHFconsidered inship design?How?

Collisioncasesrelatedtoautomation

07.032004*)FarSymphonyhadacoursetowardsthefacilityWestVenture.Enteringthesafetyzone,theautopilotwasengaged.Theofficeronthebridgedidnotrealizethattheautopilotwasengagedandcouldnotnavigatethevessel.Thisendedinacollision.

18.072007*)Grane wasidentifiedasatargetfortheautopilotonBourbonSurf.Themastermisjudgedtheship’sspeedanddistancetotheplatform.Hedidnotkeepaproperlookoutatthetime.itwastoolatetostopthevessel,buttheysucceededinreducingitsspeedfrom3m/sto1m/sbeforeithitGrane

06.06.2009*)WellstimulationvesselBigOrangeXVIIIwasapproachinginstallationEkofisk 2/4X.Thecaptainengagedtheautopilotandforgottoswitchitoff.Hecouldnotcontrolthevesselmanuallyasheintendedtodo.Insteadofslowingdown,thevesselstrucktheinstallationataspeedof9.5knots.

Analysis:Thecrewfailedtoseethattheautopilotwasengagedandmadeawrongdecisioninoperatingthevessel.

*)Petroleum Safety Authoritiy Norway. (2011). Risk of Collisions with Visiting Vessels  Retrieved 10 Oct 2011, from http://www.ptil.no/news/risk‐of‐collisions‐with‐visiting‐vessels‐article7524‐79.html

Offshoresupplyvesselsin

NorwegianContinentalShelf

• Carrygoodsto&fromoffshoreplatforms:containers,bulk,fluid(tanker),supportplatformsinvariousways.

• Hitech• Developingveryfast• 12‐ 16crewonboard• 2‐ 3tripsperweekserving2‐ 6platformspertrip

• Crewrotation:4weekson4weeksoff

15

16

Issuesfoundonboardrelatedtodigitalizationandautomation

• Thecrewknowlessoftheirvessel,andmoredependentofthemanufacturers

• Nomore‘wheel’andenginetelegraphonthebridge

Mosttraditionalcontrolsarereplacedbyjoystick,trackball,mouse,keyboardandtouchscreen

Givenemergencysituation,crewinterventionbecomeslessstraightforwardandlessintuitive

17

Problems:• Illumination• Systemreadiness,datavalidity• Operatingsystemrelatedproblems;updating,bugs

• Compatibilityissues• Softwareanddataexpirationdate

• Systemoverload,hang• Unresponsivesystem• Alargenumberofalarms• Limitedinternetbandwidth• Variationin‘electricityvoltage’

17

• Controllability

‘toomuchinformationonascreen’

‘overabundantcommunication’

Issuesrelatedtodigitalizationandautomation

Onavessel,thereareanumberofconningdisplaysinstalledindifferentlocations.Theyshouldprovideconsistentinformationatalltime.

Thesetwopicturesweretakenalmostatthesametimeonavessel,showingtwoconningdisplays.

Noticeanythingwrong?

Study 2.…Exploratory

• Ergonomics issue related tohumaninteractions• Discrepancyofknowledge:

• specialfamiliarizationortrainingprogram

Criticalincidentsrelatedtodigitalizationandautonomy

DPfailureOneOSVwaslyingbesideaninstallationonDP.Insteadofholdingsteadyonthe specifiedspot,thevesselbegantomovetowardtheinstallation.Theofficertookoverthecontrols,shutdowntheDPsystem,andbackedtheshipawayfromthestructure.Onehosewasstillconnectedtotheinstallationandsnappedoff.Thesystemascertainedthatthevesselwasmorethan100mawayfromtheinstallation,whileinrealityitwasapproximately20maway.

Explanationfromthemanufacturer:

“…Wehavefoundtherootcauseforthis,andimplementedasolutionforit.Thisfailurewillnothappenagain.”

Interpretedasnon‐random,systematicerror

Overlookedscenarioduringdesign/development

Further exploration regarding the incident, according to the seafarers experience, “DP2 fails all the time”

Criticalincidentsaffectingautonomy

Blackout

OneOSVwaspreparingtomaneuverawayfromaninstallation.Tocruisetothenextinstallation,thebridgeaskedforasuddenincreaseofpowerthatthesystemwasunabletoaccommodate.

Atthetime,thesystemwasrunningonLNGfuelandattemptedtoautomaticallyswitchovertodiesel,buttheswitchoverfailed,andthesystemblackout.

Explanation:

Apparently,LNGfuelhasacharacteristicofwhichtheoperatorhadnotbeenmadeaware:

itislessresponsivetovariationsinthepowerrequirements

thedesignerandthemanufacturerwerenotawareofthischaracteristicofLNGfuel.

13.11.2006 A collision caused by blackout: Navion Hispania (tanker) blacked out, due to polluted fuel and a system malfunction. Hit Njord B at 1.2 m/s. Collision energy > 60 MJ.

Oneunresolvedissuerelatedtodigitalizationandautomation

OnoneOSV:ReleasingDPsystemfromautomaticmodetomanualmoderequiresaspecificresponseoftime.Withoutfullcontrol,thisdelaycanbecritical,especiallywhenthevesselislocatedclosetoaninstallation.Normallytheoperatorwillusethejoysticktobringthevesselawayfromtheinstallationbeforeswitchingtomanualmode.

OntheotherOSVThisdoesnotoccurbecausethetransitionoccursinstantaneously.

Expertfromaclassificationsocietystatedthatthetransferofcontrolbetweenmodesofoperationshouldbeimmediate.Adelayof10secondsisnotacceptableforthisvessel.

Thefactthatthevesselisapprovedtooperateandcarrytheassociatednotationcausesconfusion.

Study 2.Exploratory field study

Lessons learned

ModificationoftheautopilotsystemCurrently,theautopilotwillautomaticallydeactivatedwhenthecrewoperatesthejoystick(DNVNAUTOSV2012).

500msafetyzoneisimplemented.Aninstallationcannolongerbeidentifiedasatargetforautopilot(NWEA2006,updated2009)

PotentialcontributionofautomationonOSVoperation:

Smartlookout,enhancedautopilot ‘Smartrouting’~f(weather,fuelcons,t,etc.) Automaticpositioning/‘parking’ Maintenance:‘longdistancesetup,updates,ormaintenance’bythemanufacture Maintenanceschedule,‐ canbeperformedwhenthevesselatport Automaticemergencystop e‐(smart)checklist

• Increase bulwark height to avoid green water

• To secure tubular cargoes on deck, developed & installed:portable, automatic, movable stanchions

25

Lessonslearnedondeck

Lessons(triedtobe)learned,andfail

26

Tohelpdeckhandsconnectbulkhoseatsea,anautomaticbulkcargosecuring&transfer

systemwasdeveloped&installed,butfailtowork

27

Study 2.Qualitative study

Conclusions

HFhasbeenconsideredinOSVsdesign; Thecrewsingeneralaresatisfiedwiththeirvessel Thereisalwaysroomforimprovement HFconsiderationsonOSVswerehappeningduetogoodcommunicationsbetween

thecrew,shipowners,cargoownersanddifferentmanufacturers,includingtheshipyards.

Financialincentivesforresearchanddevelopmentalsoplaysanimportantrole.

Differentproblemsareexperiencedbydifferentvessels HFproblemsareunique(noise,motion,controllability,etc.) Someissuesremain(illumination,layout,space,stairs,access)

Bestsatisfying:Habitability&Workability Lowestsatisfying:Maintainability Fastestgrowingdimension:Controllability There isapotential of unknown riskinthe development of new

equipment/system.

RQ1.Howare humanfactors taken into account inship design?

28

Study 3.Quantitative study

Tocheckifthequalitativefindingsarevalid

Anexplanatory studyusing quantitativeapproachwas conducted

Results:HumanFactorsonOffshoreSupplyVesselsintheNorwegianSea– AnExplanatorySurvey

TransRINA,Vol158,PartA1,InternationalJournalofMaritimeEngineering,Jan‐Mar2016

RQ1.Howare HFtaken into account inship design?RQ2.Isthere any effect of HFconsideration toincidences on board?

Study 3.…Quantitative

RQ1.Howare HFtaken into account inship design?

• Humanfactors aresignificantly addressed.• Dimensionsof humanfactors are notrated

equally: OHSisrated morehighly Maintainability israted lower

There isanindication that humanfactorsrating varies asaresult of OSVdesign,but the finding isinconclusive(duetonumericalcorrection)

Study 3.…QuantitativeHow often do the people on board experience 

the following:How often does the vessel experience the 

following:

RQ2a.Isthere any significant effect of ship designtoincidences on board?

Study 3.Quantitative

• Habitability hasapositiveeffect on the frequency ofpersonnel becoming seasick,fatigue andexperiencingsleep disturbance.

• Maintainability hasanegativeeffect on thefrequency of fireorexplosion on board.

RQ2b.Isthere any significant effect of HFconsideration toincidences on board?

32

Study 4.Evaluationstudy

Tocheckiftheexistingknowledgeofhumanfactorsinshipdesignandoperationareeffective

Anevaluation studywas conducted,byperforming physical measurementson board,combined with some observations anddaily diaries filled inbytheseafarers after every watch

RQ3.Arethe existing knowledge of HFinship designeffective?

33

Study 4.Evaluationstudy

Conclusions:

Theexisting noise criteria donotreflect comfort Disturbing noises [impulsivenoise,high pitch noise,squeaking noise andhammering

noise]are notcovered norcaptured

Motioncriteria need toberevised forOSVoperations They are notrealistic;the criteria are too high (too lenient),

especially MIIandrollmotion.

TheMSI(McCauley etal,1976)isextremely conservative forseafarers population,needs tobeadjusted

Recommendations:

Standardsandcriteria bere‐examined within areasonable periodof time,especially incaseanincident occurs

RQ3.Arethe existing standardsof HFinship designeffective?

34

Study 5.Multivariate …

RQ4.What factors considerably influence crews’performance atsea?

35

Study 6.Theoretical evaluation

Some issues were found when developing HFcheck listsandquestionnaires,indicating that the concept of HFisstilldeveloping

Atheoretical evaluationwas performed toevaluate ortoconfirm the concept of HFinship design&operation,using factor analysis

Report:HumanFactorsinShipDesignandOperations:APreliminarySurveyoftheTheoreticalConstructTransRINA,Vol158,PartA2,InternationalJournalofMaritimeEngineering,Apr‐Jun2016

RQ5.What are HFinship design(andoperation)?

36

Study 6.Theoretical evaluation

Controllability

Workability

Habitability

Cargo facilities

Reliability, Automation and Maintainability

RQ5.What are HFinship design(andoperation)?

37

Study 6.Theoretical evaluation

Reliability, operabilityand maintainability

Interfacing complexity

Ship handling and manoeuvrability

System & procedure

Deck working condition

ER & ECR

Habitability

RQ5.What are HFinship design(andoperation)?

38

Study 6.Theoretical evaluation

A model showinghuman factorsconsiderations in ship design and operation is presented as a result of the theoretical study

RQ5.What are HFinship design(andoperation)?

Study 7.Riskassessment model

RQ6.HowtoaccountHFinmajorriskassessment?

Markov model, adopted from Dhillon (2003)

Safety instrumented system (Rausand & Høyland, 2004

This research adopted the existing hardware reliability framework to develop a method for predicting the probability of accident by focusing on human factors

Human reliability analysis Control theory, Cognitive approach Implemented to several accident cases

Thank you for your kind attention

Theendof the presentation