MANAGEMENT OF DESIGN INFORMATION IN THE PRODUCTION …488464/FULLTEXT01.pdf · MANAGEMENT OF DESIGN...

Mälardalen University Press DissertationsNo. 119

MANAGEMENT OF DESIGN INFORMATION INTHE PRODUCTION SYSTEM DESIGN PROCESS

Jessica Bruch

2012

School of Innovation, Design and Engineering



Jessica Bruch

2012

School of Innovation, Design and Engineering

Department of Industrial Engineering and Management



Jessica Bruch

Akademisk avhandling

som för avläggande av teknologie doktorsexamen i innovation och designvid Akademin för innovation, design och teknik kommer att offentligen

försvaras fredagen den 30 mars 2012, 13.00 i Filen, Smedjegatan 37, Eskilstuna.

Fakultetsopponent: Professor Bengt- Göran Rosén, Högskolan i Halmstad

Akademin för innovation, design och teknik



Jessica Bruch

Akademisk avhandling

som för avläggande av teknologie doktorsexamen i innovation och designvid Akademin för innovation, design och teknik kommer att offentligen

försvaras fredagen den 30 mars 2012, 13.00 i Filen, Smedjegatan 37, Eskilstuna.

Fakultetsopponent: Professor Bengt- Göran Rosén, Högskolan i Halmstad

Akademin för innovation, design och teknik

AbstractFor manufacturing companies active on the global market, high-performance production systems thatcontribute to the growth and competitiveness of the company are essential. Among a wide rangeof industries it is increasingly acknowledged that superior production system capabilities are crucialfor competitive success. However, the process of designing the production system has received littleattention, ignoring its potential for gaining a competitive edge. Designing production systems in aneffective and efficient manner is advantageous as it supports the possibility to achieve the best possibleproduction system in a shorter time. One way to facilitate the design of the production system is aneffective management of design information. Without managing design information effectively in theproduction system design process the consequences may be devastating including delays, difficulties inproduction ramp-up, costly rework, and productivity losses.

The objective of the research presented in this thesis is to develop knowledge that will contribute to aneffective management of design information when designing production systems. The empirical datacollection rests on a multiple-case study method and a survey in which the primary data derive fromtwo industrialization projects at a supplier in the automotive industry. Each industrialization projectinvolved the design of a new production system.

The findings revealed ten categories of design information to be used throughout the process ofdesigning production systems. The identified design information categories are grouped in thefollowing way: (1) design information that minimizes the risk of sub-optimization; (2) designinformation that ensures an alignment with the requirements placed by the external context; (3)design information that ensures an alignment with the requirements placed by the internal context,and (4) design information that facilitates advancements in the design work. In order to improvethe management of the broad variety of design information required, a framework is developed.The framework confirms the necessity to consider the management of design information as amultidimensional construct consisting of the acquiring, sharing, and using of information. Further, theframework is based on six characteristics that influence the management of design information. Thesecharacteristics are information type, source of information, communication medium, formalization,information quality, and pragmatic information. Supported by the findings, guidelines for themanagement of design information are outlined to facilitate an effective and efficient design of theproduction system and thus contribute to better production systems. The guidelines are of value to thoseresponsible for or involved in the design of production systems.

ISBN 978-91-7485-059-8ISSN 1651-4238

MANAGEMENTOFDESIGNINFORMATIONINTHEPRODUCTIONSYSTEMDESIGNPROCESS

i

ABSTRACT

For manufacturing companies active on the global market, high‐performanceproduction systems that contribute to the growth and competitiveness of thecompany are essential. Among a wide range of industries it is increasinglyacknowledged that superior production system capabilities are crucial forcompetitivesuccess.However,theprocessofdesigningtheproductionsystemhasreceived little attention, ignoring its potential for gaining a competitive edge.Designingproductionsystemsinaneffectiveandefficientmannerisadvantageousas it supports the possibility to achieve the best possible production system in ashorter time. One way to facilitate the design of the production system is aneffectivemanagementofdesigninformation.Withoutmanagingdesigninformationeffectively in the production system design process the consequences may bedevastatingincludingdelays,difficultiesinproductionramp‐up,costlyrework,andproductivitylosses.Theobjectiveoftheresearchpresentedinthisthesisistodevelopknowledgethatwill contribute to aneffectivemanagementofdesign informationwhendesigningproduction systems.The empirical data collection rests on a multiple‐case studymethodandasurvey inwhich theprimarydataderive fromtwo industrializationprojects at a supplier in the automotive industry. Each industrialization projectinvolvedthedesignofanewproductionsystem.The findings revealed ten categories of design information tobeused throughoutthe process of designing production systems. The identified design informationcategoriesaregroupedinthefollowingway:(1)designinformationthatminimizestheriskofsub‐optimization;(2)designinformationthatensuresanalignmentwiththe requirements placed by the external context; (3) design information thatensuresanalignmentwiththerequirementsplacedbytheinternalcontext,and(4)design information that facilitates advancements in the designwork. In order toimprove the management of the broad variety of design information required, aframework is developed. The framework confirms the necessity to consider themanagement of design information as amultidimensional construct consisting oftheacquiring,sharing,andusingofinformation.Further,theframeworkisbasedonsix characteristics that influence the management of design information. Thesecharacteristics are information type, source of information, communicationmedium,formalization, informationquality,andpragmatic information.Supportedbythefindings,guidelinesforthemanagementofdesigninformationareoutlinedtofacilitate an effective and efficient design of the production system and thuscontribute to better production systems. The guidelines are of value to thoseresponsiblefororinvolvedinthedesignofproductionsystems.


iii

SAMMANFATTNING

För tillverkandeglobala företagärhögpresterandeproduktionssystemsombidrartill tillväxt och konkurrenskraft för företaget oumbärliga. Inom en rad olikabranscherärdetalltmererkäntattenöverlägsenproduktionssystemsprestandaäravgörande för konkurrenskraft och framgång. Arbetet med att utformaproduktionssystemhardock fått lite uppmärksamhet ochdesspotential att bidratill konkurrensfördelar försummats. Att utforma produktionssystemet på etteffektivt och ändamålsenligt sätt kan bidra till konkurrensfördelar eftersom detstödermöjlighetenattuppnådetbästamöjligaproduktionssystemetpåenkortaretid. Ett sätt att underlätta utformningen av produktionssystemet är en effektivhantering av designinformation. Utan att hantera designinformation effektivt iutformningsprocessenavproduktionssystemetkankonsekvensernabli förödande,till exempel genom förseningar, svårigheter i upprampning av produktionen,kostsammaomarbetningar,ochförlusterproduktivitet.Måletmedforskningensompresenterasidennaavhandlingärattutvecklakunskapsom bidra till en effektiv hantering av designinformation när produktionssystemutformas. Den empiriska datainsamling vilar på en flerfallstudie och enundersökning där de primära uppgifterna kommer från tvåindustrialiseringsprojekt hos en leverantör inom fordonsindustrin. Varjeindustrialiseringsprojektomfattarutformningenavettnyttproduktionssystem.Resultatenpekadepå tio designinformationskategorier som skall användas underproduktionssystemets utformningsprocess. De identifierade kategorierna kangrupperas på följande sätt: (1) designinformation som minimerar risken försuboptimering, (2) designinformation som säkerställer en anpassning till de kravsom ställs från externa omgivningen, (3) designinformation som säkerställer enanpassning till de krav som ställs internt, och (4) designinformation somunderlättarframstegiutformningsprocessen.Förattförbättraförvaltningenavdenstora variationen av designinformation som krävs, har ett ramverk utvecklats.Ramverket bekräftar nödvändigheten av att beakta hanteringen av designinformation som en flerdimensionell konstruktion bestående av förvärvandet,delandet och användandet av informationen. Vidare är ramverket baserat på sexegenskaper sompåverkar hanteringen av designinformationen.Dessa egenskaperär typav information, informationskällor, kommunikationsmedium, formalisering,informationskvalitet och pragmatisk information. Med stöd av resultaten ärriktlinjeruppdragnaförhanteringavdesignuppgifterförattunderlättaeneffektivochändamålsenligutformningavproduktionssystemetochdärmedbidratillbättreproduktionssystem.Riktlinjernaär avvärde förde somansvarar för ellerdeltar iutformningenavproduktionssystem.


v

ACKNOWLEDGEMENTS

It is with a bittersweet joy that I write the final lines of my thesis. Being a PhDstudenthasbeenaninterestingprocess,atwistedpathwithsurprises.Sometimesithasbeendifficulttoseethepathandothertimesithasbeenveryrewarding.Allinallithasbeenagreatexperience!Thepathhasbeensharedwithmanyfriendsandcolleagueswhorefinedandcontributedtotheresultsofthisthesis.Iwouldliketoexpressmygratitudetoeveryonewhohelpedmemakethispossible.IamdeeplygratefultomysupervisorsMonicaBellgranandChristerJohanssonwhohadthemostimportantinfluenceonthiswork.Monicahasbeenaconstantsourceofnew ideas and inspiration,pushingme togo further.At the same time shehasbeenanexcellentmentorand Ihave truly learnedagreatdeal fromher.Christerdrew my attention to weaknesses in my work through his stimulating andchallengingquestions.I owe a special thanks to Kristina Säfsten. Kristina has always been incrediblysupportivethroughouttheresearchprocessandmadeasignificant impactonthisthesis by her well conducted “final review”. Her feedback and suggestionssharpenedmythinkingandasaresultshehashelpedmetoimprovemywork.Iwouldalsoliketothanktheemployeesatthecasestudycompanies.Withouttheirengagementandopennessthisthesiswouldnothavebeenwhatitistoday.Thecasestudiesareanonymousinthisthesisbutnotforgottenforthatsake.Aspecialthanksto Calle, Göran, John, Mark and Per‐Olof for their valuable assistance. FinancialsupporthasbeengivenbyVinnova ‐ theSwedishAgency for InnovationSystems,for which I am very grateful. I would also like to thank the ProViking ResearchSchoolforfinancingmyresearchvisitinBirmingham,UK.Thisthesishasfromtimetotimebeenamajorundertaking.MypresentandformercolleaguesattheDepartmentofindustrialEngineeringandManagementneedtobeacknowledgedformakingmyyearsherepositiveanddeveloping.SpecialthankstoJohanKarltunwhohasprovidedmuchneededadviceandsupporttomylicentiatethesis. Furthermore, many thanks to my colleagues at the School of Innovation,Design and Engineering for havingmademe feel sowelcome, in particular AnnaGranlundwhohadalwaysherdooropenformeandtookcareofallkindofmatters.CarinRösiöandMalinLöfvingdeserveaparagraphontheirown.Withoutyou,theprocesswouldonlyhavebeenhalfasfunasithasbeen.Theyhavehelpedmewithlargeandsmallproblemsandbuiltmeupwhenthefrustrationbecameunbearable.Whenever needed they have taken their time to discuss, well, everything ... Mywarmestthankstoyouboth!

vi

Thisprocesswouldnothavebeenpossiblewithoutthesupportandencouragementofmyfriendsandfamily.Myparents,WalterandJutta,foralwaysbeingthereandmysiblings,Nicole,NanneandChristinfortheirgoodcompany.Mydeepestgratitude is toLars,whofollowedtheprogressof thisthesiseachdaysharingjoyanddisappointments.MostofallIamthankfulforhisbeingthereformealltheway.JönköpinginFebruary,2012JessicaBruch


vii

APPENDEDPAPERS

Thethesisconsistsoftwomainparts:thesummarizingchapterofthiscompilationthesisandthefollowingsixpapersappendedinfull:

PaperI Bruch, J. and Karltun, J. (2009), Information Requirements in aProactiveAssemblyWorkSetting,Presentedat the3rdInternationalConference on Changeable, Agile, Reconfigurable and VirtualProduction,5‐7October2009,Munich,Germany.

PaperII Bruch, J. and Johansson, G. (2011), Dual Perspective on InformationExchangesbetweenDesignandManufacturing,Presentedatthe18thInternationalConferenceonEngineeringDesign,15‐18August2011,Copenhagen,Denmark.

PaperIII Bruch, J. andBellgran,M. (Inpress),Design information forefficientequipment supplier ‐ buyer integration, Journal of ManufacturingTechnologyManagement,,Vol.23,No.4.

PaperIV Bruch, J. andBellgran,M. (2012), Creating a competitive edgewhendesigning production systems: facilitating the sharing of designinformation, Status:Re‐submitted to International Journal of ServiceSciencesfollowingonerevision.

PaperV Bruch,J.andBellgran,M.(2011),Managingdesigninformationintheproduction system design process, Status: First round of review inInternationalJournalofProductionResearch.

PaperVI Bruch,J.,Bellgran,M.andAngelis,J.(2011),InformationManagementfor Production System Design with a New Portfolio Approach.Presented at the 21st International Conference on ProductionResearch,31July–4August,2011,Stuttgart,Germany.

viii

Additionalpublicationsbytheauthor,butnotincludedinthethesis

Bruch, J., Johansson, C., Karltun, J. and Winroth, M. (2007), Considering designdemands of a proactive assembly system: a position paper, Presented at the 1stSwedishProductionSymposium,28‐30August2007,Gothenburg,Sweden.Granell,V., Frohm, J.,Bruch, J. andDencker,K. (2007),Validationof theDYNAMOMethodology forMeasuringandAssessingLevelsofAutomation,Presentedat the1stSwedishProductionSymposium,28‐30August2007,Gothenburg,Sweden.Dencker, K., Stahre, J., Bruch, J., Gröndahl, P., Johansson, C., Lundholm, T. andMårtensson, L. (2007), Proactive Assembly Systems ‐ Realizing the Potential ofHumanCollaborationwithAutomation,Presentedat the IFAC‐CEA:ConferenceonCostEffectiveAutomationinNetworkedProductDevelopmentandManufacturing,2‐5October2007,Monterrey,Mexico.Bruch, J., Karltun, J. and Dencker, K. (2008), Assembly Work Settings EnablingProactivity – Information Requirements, Presented at the 41st Conference onManufacturingSystems(CIRP),26‐28May2008,Tokyo,Japan.Bruch,J.,Karltun,J.,Johansson,C.andStahre,J.(2008),TowardsaMethodologyfortheAssessmentofInformationRequirementsinaProactiveAssemblyWorkSetting,Presented at the 2nd Swedish Production Symposium, 18‐20 November 2008,Stockholm,Sweden.Bruch, J. (2009), InformationRequirements inaProactiveAssemblyWorkSetting,Licentiatethesis,ChalmersUniversityofTechnology,Gothenburg.Bruch,J.,Bellgran,M.andJohansson,C.(2009),Exploringrequirementspecificationof the production system – a position paper, Presented at the 3rd SwedishProductionSymposium,2‐3December2009,Gothenburg,Sweden.Fasth, Å., Bruch, J., Stahre, J., Dencker, K., Lundholm, T., Mårtensson, L. (2009),Designing proactive assembly systems – Criteria and interaction betweenautomation, information, and competence, Asian International Journal of ScienceandTechnology inProductionandManufacturing(AIJSTPME),Vol.2,No.2,pp.1‐13.Waefler, T., von der Weth, R., Karltun, J., Starker, J., Gaertner, K., Gasser, R. andBruch, J. (2011),Control capabilityofWorkSystems, InWaefler,T., Fransoo, J. C.,Wilson, J.R. (Eds.) Behavioral Operations in Planning and Scheduling, Berlin,Germany:Springer‐Verlag.Bruch, J., Wiktorsson, M., Bellgran, M. and Salloum, M. (2011), In search forimproved decision making in manufacturing footprint: A conceptual model forinformation handling, Presented at the 4th Swedish Production Symposium, 4‐5May2011,Lund,Sweden.Bruch, J. and Bellgran, M. (2011), The critical role of design information forimproved equipment supplier integration during production system design,Presented at the 44th Conference onManufacturing Systems (CIRP), 31May – 3June,2011,Madison,WI,USA.


ix

TABLEOFCONTENTS

PART1:SUMMARIZINGCHAPTER

1. INTRODUCTION................................................................................................................................1 1.1 GAININGANEDGETHROUGHPRODUCTIONSYSTEMDESIGN.............................................................................1 1.2 THECRITICALROLEOFDESIGNINFORMATION.........................................................................................................3 1.3 RESEARCHOBJECTIVEANDRESEARCHQUESTIONS................................................................................................5 1.4 DEFININGANDDELIMITINGTHERESEARCHAREA..................................................................................................6 1.5 OUTLINEOFTHETHESIS.......................................................................................................................................................7

2. FRAMEOFREFERENCE..................................................................................................................9 2.1 THEPRODUCTIONSYSTEMDESIGNPROCESS.............................................................................................................9

2.1.1 Productionsystems...............................................................................................................................9 2.1.2 Asystematicdesignprocess...........................................................................................................11 2.1.3 Differentapproachestotheproductionsystemdesignprocess....................................13 2.1.4 Difficultiesintheproductionsystemdesignprocess..........................................................15

2.2 THEMANAGEMENTOFDESIGNINFORMATION......................................................................................................17 2.2.1 Definingdesigninformation...........................................................................................................17 2.2.2 Acquiringdesigninformation........................................................................................................18 2.2.3 Sharingdesigninformation.............................................................................................................20 2.2.4 Usingdesigninformation.................................................................................................................23

2.3 SUMMARYOFFRAMEOFREFERENCE..........................................................................................................................25 3. RESEARCHMETHODOLOGY......................................................................................................27

3.1 RESEARCHAPPROACH.........................................................................................................................................................28 3.2 RESEARCHMETHOD..............................................................................................................................................................29 3.3 RESEARCHDESIGN.................................................................................................................................................................29

3.3.1 Numberofcases...................................................................................................................................30 3.3.2 Real‐timeorretrospectivecases...................................................................................................30 3.3.3 Caseselection........................................................................................................................................31 3.3.4 Unitofanalysis......................................................................................................................................32

3.4 DATACOLLECTION................................................................................................................................................................34 3.4.1 CaseBandCaseP:Real‐timecasestudies................................................................................34 3.4.2 CaseEandCaseT:Retrospectivecasestudies.......................................................................39 3.4.3 SurveyS....................................................................................................................................................40

3.5 ANALYSISOFTHEEMPIRICALDATA.............................................................................................................................40 3.6 JUDGINGTHECREDIBILITYOFTHERESEARCH.......................................................................................................42

3.6.1 Constructvalidity.................................................................................................................................42 3.6.2 Internalvalidity....................................................................................................................................43 3.6.3 Externalvalidity...................................................................................................................................44 3.6.4 Reliability................................................................................................................................................44

3.7 CONTRIBUTIONTOTHEPAPERS....................................................................................................................................46 4. EMPIRICALFINDINGS..................................................................................................................47

4.1 CASEB..........................................................................................................................................................................................47 4.2 CASEP..........................................................................................................................................................................................51 4.3 CASEE..........................................................................................................................................................................................55 4.4 CASET..........................................................................................................................................................................................56

x

4.5 SURVEYS.....................................................................................................................................................................................57 5. CONTENTANDSTRUCTUREOFDESIGNINFORMATION................................................61

5.1 CATEGORIZINGDESIGNINFORMATION.......................................................................................................................61 5.2 STRUCTUREFORHANDLINGDESIGNINFORMATION...........................................................................................64

5.2.1 Acquiringofdesigninformation...................................................................................................65 5.2.2 Sharingofdesigninformation.......................................................................................................68 5.2.3 Usingdesigninformation.................................................................................................................72

6. TOWARDSEFFECTIVEMANAGEMENTOFDESIGNINFORMATION.............................77 6.1 REQUIREDDESIGNINFORMATIONWHENDESIGNINGPRODUCTIONSYSTEMS......................................77 6.2 THE CHARACTERISTICS OF DESIGN INFORMATION MANAGEMENT WHEN DESIGNING

PRODUCTIONSYSTEMS........................................................................................................................................................79 6.3 DESIGNINFORMATIONMANAGEMENTFRAMEWORK..........................................................................................82

7. DISCUSSIONANDCONCLUSIONS.............................................................................................87 7.1 GENERALDISCUSSION..........................................................................................................................................................87 7.2 CONCLUSIONS...........................................................................................................................................................................89 7.3 METHODOLOGICALDISCUSSION.....................................................................................................................................90 7.4 DISCUSSIONOFTHECONTRIBUTIONSOFTHERESEARCH.................................................................................92

7.4.1 Scientificcontributions.....................................................................................................................92 7.4.2 Industrialcontributions...................................................................................................................92

7.5 SUGGESTIONSFORFURTHERRESEARCH....................................................................................................................93 REFERENCES……………………………………………………………………….…………..….…………………......95

PART2:APPENDEDPAPERSPAPERI INFORMATIONREQUIREMENTSINAPROACTIVEASSEMBLYWORKSETTING.................111PAPERII DUAL PERSPECIVES ON INFORMATION EXCHANGES BETWEEN DESIGN AND

MANUFACTURING..............................................................................................................................................123PAPERIII DESIGN INFORMATION FOR EFFICIENT EQUIPMENT SUPPLIER – BUYER

INTEGRATION......................................................................................................................................................135PAPERIV CREATING A COMPETITIVE EDGE WHEN DESIGNING PRODUCTION SYSTEMS:

FACILITATINGTHESHARINGOFDESIGNINFORMATION.............................................................155PAPERV MANAGING DESIGN INFORMATION IN THE PRODUCTION SYSTEM DESIGN

PROCESS.................................................................................................................................................................175PAPERVI INFORMATION MANAGEMENT FOR PRODUCTION SYSTEM DESIGN WITH A NEW

PORTFOLIOAPPROACH..................................................................................................................................193


xi

LISTOFFIGURES

Figure1. Outlineofthesummarizingchapterofthiscompilationthesis.......................................................8 Figure2 Typicalworkactivitiescarriedoutwhendesigningtheproductionsystem..........................12 Figure3 Three different approaches to the design process from the manufacturing company’s

perspective(adaptedfromSäfsten,2002).............................................................................................14 Figure4. Modelofthethreedimensionsofmanagingdesigninformationintheproductionsystem

designprocess(basedonFrishammarandYlinenpää,2007).......................................................18Figure5. Taskcomplexityandearlierexperience(R=routineproblem,T‐f=trained‐forproblem,

N=novelproblem)(basedonFjällström,2007;Säfstenetal.,2008).......................................19Figure6. The different needs of information sharing to cope with uncertainty and equivocality

reduction(adaptedfromDaftandLengel,1986)................................................................................21Figure7. Informationqualityframework(adaptedfromEppler,2006)......................................................24Figure8. TheindustrializationprojectastheunitofanalysisofCaseBandCaseP...............................33Figure9. The production equipment acquisition project as the unit of analysis in Case E and

CaseT.......................................................................................................................................................................33Figure10. Datacollectionprocess....................................................................................................................................34Figure11. The data reduction and analysis process based on a continuous interaction between

theoryanddata(basedonRichtnér,2004)............................................................................................41Figure12. IllustrationoftheverificationplaninCaseB........................................................................................50Figure13. Detailed list of work activities that should be accomplished in the production system

designprocess.....................................................................................................................................................53Figure14. Theproportionof harddesign informationusedby equipment suppliers in relation to

totalinformation................................................................................................................................................58Figure15 Tenidentifiedcategoriesofdesigninformationrequiredforthedesignoftheproduction

system.....................................................................................................................................................................64Figure16 Categories of design information acquired in the different phases of the production

systemdesignprocess.Thedifferentgreyscalesindicatetheoriginoftheinformation..68Figure17.Design informationreceivedfromandsenttothe industrializationprojectmanagerand

productionengineeringmanagerwhendesigningthepreproductionsysteminCaseB..70Figure18. Processforselectingaproductionequipmentsupplierbasedonthefourcasestudies...71Figure19. The management of design information should be viewed as a looping of acquiring,

sharing, and using of design information with a high dependency between the threedimensions............................................................................................................................................................79

Figure20. Themanagementofdesigninformationisacontinuousloopingofacquiring,sharing,andusing of design information through all phases of the production system designprocess....................................................................................................................................................................80

Figure21. Overview of the identified characteristics affecting the management of designinformation...........................................................................................................................................................80

Figure22. Overview of the key factors influencing the management of design information whendesigningproductionsystems......................................................................................................................83

Figure23 Thedesigninformationmanagementframeworktobeusedwhendesigningproductionsystems...................................................................................................................................................................85


xiii

LISTOFTABLES

Table1. Informationqualitydimensionsandtheirdefinitions(Eppler,2006)......................................23 Table2. Overviewofthecompaniesselectedandthecasesstudied...........................................................31 Table3. OverviewofthedatacollectedforCaseBduring37daysbetweenNovember2009and

August2011..........................................................................................................................................................37 Table4. Overviewof thedata collected for CasePduring34daysbetweenFebruary2011 and

April2011..............................................................................................................................................................38 Table5. Dataforthesemi‐structuredinterviewsduringCaseE....................................................................39 Table6. Dataforthesemi‐structuredinterviewsduringCaseT...................................................................40 Table7. The applied methods for strengthening validity and reliability (based on Olausson,

2009)*.....................................................................................................................................................................45 Table8. Case study denotations in the thesis and papers and the authors’ contributions to the

papers......................................................................................................................................................................46 Table9. Categoriesof informationusedbytheequipmentsuppliersandrankedbyrelevanceto

thesuppliers.........................................................................................................................................................58 Table10. The four most important factors contributing to an effective production equipment

acquisitionprocessasidentifiedby25equipmentsuppliersinthesurvey............................59 Table11. Thetendesigninformationcategoriesclassifiedaccordingtotheirorigin............................66

PART1

SUMMARIZINGCHAPTER


1

CHAPTER1

:INTRODUCTION

CHAPTERINTRODUCTIONThe first chapter of this thesis establishes the importance of the research area –managing design informationwhen designing production systems and framing itinto a context. Based on a need for a more effective management of designinformationwhendesigningproductionsystems, theresearchobjective isdefinedandtheresearchquestionsareformulated.Further,thescopeandstructureofthethesisarepresented.

1.1 GAININGANEDGETHROUGHPRODUCTIONSYSTEMDESIGNArguablytheprerequisitesforeconomicsuccessofmanufacturingcompanieshavechanged tremendously during the last two decades. Several uncontrollable forceshaveemergedincludingagrowinginternationalenvironment,fragmentedmarketswith sophisticated customers, fast‐evolving technology, and shrinking productlifetimes(Chryssolouris,2006;ClarkandFujimoto,1991;ElMaraghyandWiendahl,2009).Asthecompetitioninwhichthecompaniesoperateisincreasing,frequentlyintroducingnewproducts to themarket in time is crucial forbusinessprosperity(Girotraetal.,2007;StalkJr.andHout,2003).Forexample,HendricksandSinghal(2008)pointoutthatdelaysinproductintroductionhaveasubstantialandnegativeimpactonprofitability.Thesenegativeconsequencescanbeexplainedbycustomersthat cancel orders, a reduced window for generating revenues, products fasterbecomingobsolete,or lowerproductprices(HendricksandSinghal,2008).Delaysminimize a company’s possibility to benefit from first‐mover advantages and canlead to decreases in market share and sales growth as well as potentiallystrengthening the market position of competitors. Thus, being late with theintroduction of a product can be devastating if competitors succeed in gaining asuperiormarketposition.Hence the ability to identify the needs of the customer and to quickly createproductsthatwillmeettheseneedsandthatcanbemanufacturedatlowcostwillhave major implications on the survival, growth, and profitability of companies(Trott, 2008;Ulrich andEppinger, 2007).This being said, extensive scientific andindustrial attention has been devoted to finding the most efficient methods andtoolsinordertoimprovetheproductdevelopmentperformanceofcompanies.Theresult is a substantial body of knowledge with a great deal of homogeneityconcerning tools and methods that support the performance of structured andefficientproductdevelopment.Cochranetal. (2001/2002)concludethatalthough

INTRODUCTION

2

thefieldofproductdevelopmentisstillgrowinganddynamic,thereisanagreementonwhatitmeanstodevelopaproduct.However,thesuccessofmanynewproductsishighlyrelatedtotheabilityofintegratingthedevelopmentofproductionsystems(BellgranandSäfsten,2010).The area of production system development has generated far less excitementamongacademicsandpractitioners.This isperhapsnotsurprisingas thewesternworldhaslongbeenemphasizingthedevelopingofproducts,whileitwasassumedthatthemanufacturingofproductscouldbecarriedoutinlow‐wagecountriesorbycompetitors with stronger operating competencies (Karlsson, 2009). Theperceptionthat industrialproductionisnotacorecompetenceandofnostrategicuse ignores the integrating role of production systemdevelopment capabilities innew product development (NPD) performance. The real power of superiorproductionsystemdevelopmentisnot itscontributiontoreducedoperatingcosts,but how it supportsmanufacturing companies in their attempts to achieve fastertime tomarket, smoother production ramp‐up, enhanced customer acceptance ofnewproducts, and/or a stronger proprietary position (Hayesetal., 2005; Pisano,1997). Further, the real valueof theproduction system is not its often extremelycostlyproductionequipmentbuttheintellectualcapitalembeddedwithinitsdetailssuch as assembly sequences or quality assurance (Hayes et al., 2005). Theknowledge relevant for creating these details resides in the heads of the peopleinvolved in the creation process and is thus difficult to observe and imitate bycompetitors.Consequently, research into production system development is of high relevancewhenspeedinNPDisacritical issueformanufacturingcompanies.Ingeneral,theprocessofdevelopingaproductionsystemincludesboththedesignofaproductionsystem solution and the implementation of the solution (Bellgran and Säfsten,2010). Previous research reveals that the design process is the foundation of acompetitiveandprofitablemanufacturingbusiness.Forexample,Bennett(1986,p.2)pointsoutthat“thewayinwhichaproductionsystemisdesignedwillenableorprecludethepossibilityofachievingbestresults”.Thedecisionsmadeinthedesignphase have major implications on factors such quality, speed, dependability,flexibility, and cost of the production system (Slack et al., 1998). Early designdecisions aremuchmore significant than later production decisions due to theirimpact on the downstream business activities by being technically feasible orpracticallyviable(Bartonetal.,2001).Further,manufacturingcompaniesthatshiftthe identification and solving of design problems to earlier phases of thedevelopmentprocesscanenhance theoveralldevelopmentperformance(Thomkeand Fujimoto, 2000). The right design before implementation facilitates thatsystemscanberapidlycommissionedtoallowforrapidrepaymentoftheinvestedcapitalaswellasbringingnewproductstothemarketpromptly,thusreducingthecostforthemanufacturingcompany(Wu,1994).Thebenefitsassociatedwithfront‐loading problem solvingmake the production system design process particularlyinteresting.Althoughithasbeenarguedthatthedesignofproductionsystemsiscrucial,thereisa lack of theory supporting practitioners in their critical task of designing aproductionsystem.Concernshavebeenraisedthatresearchinproductionsystemdesign as one of the determinants of the effectiveness of operations and NPD is


3

seriously underexposed (Ruffini etal., 2000). As a result, production systems aregenerallydesignedrelativelyshortlybeforetheirinstallation(Chryssolouris,2006;Duda,2000),whichlimitstheabilitytoevaluatetheconceptualsolutioninatimelyand financially soundmanner.However, a poor conceptual solution can never becompensated for by the later phases in development projects (Cross, 2000).Therefore, this thesis considers the production system design process as a non‐trivialproblemthatwillincuraconsiderableamountofriskunlessdesignactivitiesarehandledinanefficientandeffectivemanner.Many approaches to an effective and efficient production system design processoriginated before production systems needed to have the ability to continuouslyadapt and evolve. In the past,manufacturing companies could plan for relativelylong product life cycles that more or less followed an s‐shaped diffusion curveaccordingtotheintuitivelogicofintroduction,growth,maturity,anddecline(Mataetal.,1995;PowellandDent‐Micallef,1997).Oneof theresultswas thatboth thetime and the costs required for designing a new production system onlyrepresentedasmallproportionofitstotallifetimeandtheunit’sfullcost(Hayesetal., 2005). Today’s dynamic markets, on the other hand, imply new types of lifecycles and an enormous number of productmodels and variants (ElMaraghy andWiendahl, 2009; Wiendahl et al., 2007), causing an increasing divergence of theproductandtheproductionsystemlifecycle(KellerandStaelin,1987).Realityshowsthatdespitetheintroductionoftoolsthatfacilitatetheintegrationofproduction issues inNPD, such as integrated product development or concurrentengineering (see, among others, Andreasen and Hein, 1987; Gerwin andBarrowman, 2002; Magrab et al., 2010; Terwiesch et al., 2002), the productionsystemisoftenanobstacletofutureproductintroductions.Theintroductionofnewproductsandtheincreasingnumberofproductvariantstriggerfrequentchangesintheproductionsystem,whichoftendictatescostlyandtime‐consumingchangesto,for instance, jigs, fixtures, and machinery (ElMaraghy, 2009). Due to the highinvestment costs of new production equipment, many manufacturing companiesevenhesitatetointroducenewproductsthatwouldmaketheirexistingproductionequipment outdated (Hayes etal., 2005). Thus there is clearly a need for amoreeffective and efficient production system design process resulting in betterproductionsystems.1.2 THECRITICALROLEOFDESIGNINFORMATIONTheproductionsystemdesignprocessgreatlyaffectsNPDperformance,whichhascontributed to improved knowledge about designing production systems in asystematicwaybasedonapredefined structure (see, amongothers,BellgranandSäfsten, 2010; Bennett and Forrester, 1993; Schuh et al., 2009). Yet, there is agenerallackofempiricalstudiesanalysingandidentifyingresourcesrequiredwhendesigning production systems and capabilities needed to deploy, integrate, andprotectthoseresources.Informationisoneimportantresourcetobeabletocarryoutthedesignprocessinan effective and efficient manner. Kehoe et al. (1992) regard information as themostvaluable resource that amanufacturing companyowns,beyondanydoubtapowerful weapon. Drawing on the resource‐based view, information can beconsideredasastrategicresource,i.e.aresourcethatincludesallassetscontrolled

INTRODUCTION

4

byafirmthatenablethefirmtoimplementstrategiestoimproveitsefficiencyandeffectiveness (Barney, 1991; Barney et al., 2001). A strategic resource leads toperformance differences across organizations, and competitors find it difficult tosubstituteorimitatetheresourcewithoutgreateffort(Hoopesetal.,2003;Peteraf,1993). In other words, a manufacturing company highly proficient in providingrelevantandnecessarydesigninformationthatfitstheneedsofparticularusersonspecific occasions might develop a competitive advantage over less skilledcompetitors.Prior research has shown that approaching information from the resource‐basedtheory can contribute to improved understanding of information in thedevelopmentprocess(e.g.Frishammar,2005;Mataetal.,1995;Zahayetal.,2004).Tomeet the general requirements of the resource‐based theory and to allow forcompetitive advantages, information has to be heterogeneous across firms andimperfectly mobile (Barney, 1991). The design information in the productionsystem design process has no intrinsic value; instead its value depends on theparticularcontextandwhetheritenablesnecessaryactivitiesanddecisionstotakeplace (Galliers, 1987). Organizations also have a natural tendency to create aterminologyandsystemofmeaningoftheirown(Weick,1969).Finally,mostoftheinformationrequiredintheproductionsystemdesignprocesscanonlybefoundinthemindsof experienced systemdesigners (Bellgran, 1998). Thus, information isoften unique and deeply embedded in the organization, by which it satisfies thedemandtobescarceanddifficulttoimitateandsubstitute(Lewisetal.,2010).Tofullyusethepotentialofdesigninformation,manufacturingcompaniesalsoneedtohave thecapability tomanagedesign information inaneffectiveway. In fact, ithas been argued that the managing of information “plays a pivotal role indetermining the success or failure” of new products (Ottum andMoore, 1997, p.258). The effectiveness of information management needs to be based on thecapability to avoid situations in which the production system design process iseitherbeingsubjectedtoinformationoverloadorgettinginformationtoolateornotat all. Hence, capabilities are needed to deploy, integrate, and protect the designinformation resource. The term “capability” refers to tangible or intangibleprocessesthatarefirm‐specificandaredevelopedovertime(Makadok,2001).Thuscapabilities cannot easily be bought; instead they must be built within theorganization(Teeceetal.,1997).Frishammar(2005)arguesconvincinglythat themanagementofinformationisacapabilitythatmayallowforeffectiveandefficientNPDandsubsequentlycontributetocompetitiveadvantages.However, prior research offers insights and evidence that the capability ofmanagingdesign information is challengingandall but trouble‐free. For example,literature frequently stresses the paradoxical situation that, although there is anabundanceofinformationavailable,itisextremelydifficultforthepeopleinvolvedtoobtainnecessaryandrelevant informationwhensuch isneeded(EdmundsandMorris, 2000). If too much information is provided, the person receivinginformationcannotuseiteffectivelysincehe/sheisburdenedwithalargesupplyofunsolicitedinformation,someofwhichmayberelevant(Butcher,1998).Searchingfor and accessing design information can take up to 34 per cent of engineers’workingtime(MacGregoretal.,2001).Furthermore,theinabilitytohandledesigninformation may have major severe consequences including delayed launch to


5

market, exceeding the budget, corrections in operation, customer dissatisfaction,reducedmarketshare,andtheimpossibilityofaccomplishingdevelopmentprojects(Baxter,1995;Cooper,1999).Ithasalsobeenarguedthataneffectivemanagementofdesigninformationcontributestotheinnovationcapabilityofthemanufacturingcompanies(FrishammarandHörte,2005;MillerandFriesen,1982).Thereasoningabovehighlightsthecriticalroleofdesigninformationanditsmanagementforthesuccess of the designprocess. Consequently, there is a need for empirical studiesfocusing on the management of design information when designing productionsystems.1.3 RESEARCHOBJECTIVEANDRESEARCHQUESTIONSThe discussion so far can be summarized by stating two conclusions. First, theintroductionshowsthat thedesignofproductionsystemscancreateasubstantialedge over less skilled competitors and contribute to a company’s prosperity.Therefore,theeffectivenessandefficiencyindesigningproductionsystemscanbeastrategic weapon for competition in a global environment with sophisticatedcustomers, fast‐evolving technology, and shrinking product lifetimes. Second, theeffectiveness and efficiency of the production system design process is largelydependent on the capability of managing relevant and necessary designinformation,thusbringingtheresearchareainthisthesisintofocus.However,althoughpreviousstudiesclearlycontributetotheliteratureofimprovedmanagementofdesigninformation,theydonotfocusexplicitlyontheimplicationsfor themanaging of design information in the production systemdesign process.The majority of theories on managing information originate from the field ofproduct design and development,while theories in the production systemdesignprocessrarelyfocusonthemanagingofdesigninformation.Addressingthegapsintheliterature,theresearchobjectivewasformulatedasfollows:The objective is to develop knowledge to contribute to an effectivemanagement ofdesigninformationwhendesigningproductionsystems.

Thistypeofresearchisimportanttosupportthemanagementofdesigninformationwhen designing the production system. An effective management of designinformation is seen as a means of contributing to and effective and efficientproductionsystemdesignprocess,thussupportingthecreationofthebestpossibleproduction system in a shorter time. To address the objective of the thesis, thethesis expands the analysis of the role of design information in the productionsystemdesignprocessintotwosub‐areas,thetypeofdesigninformationrequiredandthemanagingofdesigninformation.Theformerreferstotherequiredresourcetocarryoutdesignactivitiesinaneffectiveandefficientway,whilethelatterrefersto the capability required to deploy design information resources. Therefore, thethesisfocusesontworesearchquestions:

RQ1: Whatdesigninformationisrequiredwhendesigningproductionsystems?RQ2: Whatcharacterizesthemanagementofdesigninformationwhen

designingproductionsystems?

INTRODUCTION

6

1.4 DEFININGANDDELIMITINGTHERESEARCHAREAIn this thesis, the objective is to develop knowledge to contribute to an effectivemanagementofdesigninformationwhendesigningproductionsystems.Toachievethis, calls for a structure that enables the assimilation and utilization of thedevelopedknowledge.Therefore,adesigninformationmanagementframeworkwillbecreatedinordertovisualizethefindingsoftheresearchpresentedinthisthesisand to support the management of design information when designing theproductionsystem.NPDcanbedefinedas“thetransformationofamarketopportunityintoaproductavailableforsale”(KrishnanandUlrich,2001,p.1).ThismeansthatthetermNPDasused in this thesis concernsbothdevelopment andmanufacturingof products.Consequently, NPD is employed to refer to a broader concept than only productdevelopment; NPD rather considers both product and production systemdevelopment as integrated processes that are dependent on each other forsuccessfulNPDprojects.Production system design can be defined as the conception and planning of theoverallsetofelementsandeventsconstitutingtheproductionsystem,togetherwiththerulesfortheirrelationshipsintimeandspace(Chisholm,1990).Theresultofaproduction system design process is a detailed description of the proposedproductionsystemsolution,whileproductionsystemdevelopmentasused in thisthesisalsoincludestherealizationoftheproductionsystem(BellgranandSäfsten,2010). Consequently, the focus of the empirical studies has been on the actualdesign task, which corresponds to the early phases of the development process,while the implementationandproduction start‐upof theproduction systemwereexcludedfromtheempiricaldatacollection.Further,inordertoachievetheobjectiveofthethesis,studyingthedesignprocessof the production system is crucial. Despite the fact that a significant amount ofdesign research has been carried out over the last decades, it is still questionedwhetherdesigncanbeatopicsuitableforscientificinvestigations(Berglundetal.,2001;DavenportandPrusak,1998).Therefore, thewaydesign isconsidered isofrelevance for assessing the research. In the terms used by Cross (1999), ”designscience” and “science of design” refer to research that aims at improving ourunderstanding of design and the development of support through the use ofscientific methods during the investigation. The current thesis studies principlesand practices of designing production systems as well as the relevance ofdevelopmentofsupport.The empirical data were collected in the automotive industry. The automotiveindustry has long experience of combining novelty with complexity and thusfrequently applies a concurrent development process (Terwiesch et al., 2002) inorder to develop a product in the best way regarding value‐added time andresources. Therefore, the automotive industry was a natural candidate for theresearchpresentedinthethesis.However,theresultsofthethesisaretransferabletoother industries,where it is imperative to findmore resource‐efficientwaysofdesigningproductionsystems.Whendesigning anewproduction system, thedegreeof change, i.e. the extentofchange required in relation to the properties of an existing production system,


7

varies (Bellgran and Säfsten, 2010). Thus the extent of the changes in theproduction system can be seen as a continuum ranging from minor to majorchanges.Attheoneendofthecontinuum,itispossibletofindproductionsystemsthatlargelypossessthecapabilitiesrequiredfortheintroductionofnewproducts,whileattheotherendofthecontinuumnoactualproductionsystemcanfulfiltherequiredcapabilities.The two industrializationprojects1 followed in this researchrequiredthedesignofanewproductionsystem,i.e.thenewproductscouldnotbemanufactured in the existing production systems and a new physical productionsystemneededtobecreated.Theseprojectswereselectedbecausethecreationofanewproductionsystemdemandedmoreworkactivitiesthanameremodificationofthe production system. This gave valuable insights regarding themanagement ofdesign information. If there had not been so large changes, many of the insightsgainedwouldhavebeen lost.However, it is important tonote that anentirenewproductionsystemdesigndoesnotoccurasfrequentlyaschanges.Evenwhennewproducts cannot be manufactured in actual production systems, conceptual ortechnicalsolutionsareoftenreusedforthedesignofanewproductionsystem.Finally, the managing of design information can be studied from differentperspectives. The perspective taken in this research is what Frishammar andYlinenpää (2007) call the “people‐side” ofmanaging information. As a result, thethesis ignores how the managing of design information could be organized bymeansof information technology.That is, the thesisdoesnotaddress informationtechnology aspects such as the retrieving, processing, and storing of designinformationortheuseofmanagementinformationsystems.However,althoughthethesis emphasizes the “people‐side” of design information management,informationissuesassociatedwithcognitionofinformationsuchastheperceptionand processing of information by individuals are not considered in the research.Thisdoesnotmeanthattheseissuesareirrelevant,buttheyfalloutsidethescopeofthe thesis. Further, although the research presented in this thesis studiedindustrialization projects, the research does not focus on project managementissues. The scope of the is research is on themanagement of design informationwhen designing the production system and thus project management issues areonly discussed in relationwith the aspectswhere they affect the outcome of thedesignprocess.1.5 OUTLINEOFTHETHESISThe thesis comprises two parts: (1) the summarizing chapter of this compilationthesisand(2)theappendedpapersPart 1 consists of seven chapters. In the introductory chapter the overallimportanceofthisresearchismotivatedandtheresearchobjectiveandquestionsaredefined.Chapter2outlinestheframeofreferencecoveringtwomainsections:productionsystemdesignanddesigninformationmanagement.Chapter3presentstheresearchmethodology,whichstartswithadescriptionoftheresearchapproach.This is followed by a discussion of the research method, design, the process of1ToseparatebetweentheactivitiescarriedoutduringNPD,manufacturingcompaniesusethetermindustrializationtorefertotheprocessrequiredtotransferaproductdesignintoproduction,thusincludingthedesignofproductionsystems.

INTRODUCTION

8

collectingandanalysingdata,concludingwithanassessmentofthecredibilityoftheresearch. In Chapter 4 the results of the four case studies and the survey arepresented.Theanalysisof the results isdone inChapter5.The following chapter(Chapter6)synthesizesthefindingsaroundthetworesearchquestionsandbringsthemtothedevelopmentofadesigninformationmanagementframework.Chapter7concludesthethesisandtheresearchresultsaresummarisedanddiscussed.Thechapterendswithrecommendationsforfurtherresearch.Figure1outlinesPart1–thesummarizingchapterofthiscompilationthesis.

Figure1. Outlineofthesummarizingchapterofthiscompilationthesis.Part2consistsofsixpapersproducedduringthePhDstudies.PaperIsummarizesthe conclusionsdrawn in the licentiate thesis. Paper II examines themanagingofdesign information between design engineers and production system designers,whilePaperIIIfocusesonthesharingofdesigninformationbetweenmanufacturingcompaniesandexternalequipmentsuppliers.Paper IV investigatescritical factorsfacilitating effectivemanagement of design information in the production systemdesign process. PaperV focuses on themanagement of design information in theproductionsystemdesignprocess.PaperVI investigates theconsequences for themanagingofdesigninformationwhenthereisaneedtodesignproductionsystemswithalonger‐termperspective.


9

CHAPTER2

:FRAMEOFREFERENCE

CHAPTERINTRODUCTIONIn this chapter, the frame of reference is summarized. The theoreticalconsiderationsaredividedintotwopartscentraltotheresearcharea:thedesignofproductionsystemsandthemanagementofdesigninformation.

Arecenttrendinmanufacturingindustryincludesanefforttowardsapplyingleanproductiontoachievegreaterefficiencyandproductivity(Liker,2004).Atthesametime, companies have to cope with the demands to frequently introduce newproducts to themarket (Girotra etal., 2007; Stalk Jr. andHout, 2003).While theconceptofleanproductionhasdoubtlesslycontributedtowardstheunderstandingthatproductioncanbeadecisivetoolforcompetiveness,itfocusespredominantlyon improvements in operational performance. The emphasis on improvingproductionperformancemayappear efficient tomanufacturing companies, but inthelongerterm,theprocessbywhichaproductionsystemisdesignedprovidesthelargest potential of themost cost‐effective solution (Bennett, 1986).The focusonlean production in manufacturing companies may thus not be enough to beresponsive to changes in product design and demand patterns. The propositionunderlying this thesis is that manufacturing companies need to pay increasingattention to the production systemdesign process in order to obtain competitiveadvantages. Thus, this chapter starts by reviewing previous literature about theproduction system design process before addressing critical issues regarding themanagingofdesigninformationintheproductionsystemdesignprocess.2.1 THEPRODUCTIONSYSTEMDESIGNPROCESS

2.1.1 ProductionsystemsBeforedealingwithresearchconcerningthedesignof theproductionsystem, it isessentialtounderstandtheunderlyingtermsoftheresearch.Sincethemeaningoftermsvariesamongdifferentauthors, thoseadopted in this researchareoutlinedbelow.Production is regarded as “the act or process (or the connected series of acts or processes) of physically making a product from its material constituents, as distinct from designing the product, planning and controlling its production, assuring its quality” (Chisholm, 1990, p. 736). This definition implies that production refers only to the process of convertinginput into desired products and services and thus production is seen as one of

FRAMEOFREFERENCE

10

several activities required to put a product on the market, an activity which isregarded as manufacturing. Thus, manufacturing is transforming something ofmuchgreaterscopethanproduction.Thetermsystemreferstoafinitesetofelementsthathavearelationshiptoeachother and to the environment and that under well‐defined rules should form awhole (Hubka and Eder, 1988). Hubka and Eder (1988) point out that systemsconstituteahierarchymeaningthatasystemisalwaysaconstituentpartofasupersystem,whileatthesametimeitcanitselfbedividedintosubsystems.Whenaddingthewordsystemtoproduction,itreferstotheactualphysicalsysteminwhichthetransformationfrominputintodesiredoutputstakesplace.Asystemcantherebybedefinedas“a collection of different components, such as for example people and machines, which are interrelated in an organised way and work together towards a purposeful goal” (Bellgran and Säfsten, 2010, p. 38). Asystemisaseparateunitwithsystemboundariesthatcanbedrawnatdifferentlevels, and everythingoutside of the systemboundaries canbe considered as theexternalenvironment(Wu,1994).Inaddition,asystemcanbeeitheropenorclose.Theformerreferstoasystemthatdependsonandinteractswithitssurroundings,which is not the case in a closed system. The production systems studied in thisresearchwereopensystemsthatdependedonandwereaffectedbythecontext,i.e.the production systems had to be adaptable to the changing context such ascustomerdemandsorvolumefluctuation.Itis importanttonotethatalthoughtheenvironment influences the production system, the production system cannotinfluence the environment; for a more detailed discussion, see e.g. Churchman(1978).Basedonthediscussionaboveitcanbeconcludedthattheproductionsystemcanbeseenasasubsystemofthemanufacturingsystem,wheretheproductionsystemincludesallactivitiesandelementsneededtotransferasetofinputsintoproductsand services. The production system comprises a number of elements withreciprocalrelations.Tostudythe transformationprocessrequiresconsidering thetotalityofallsubsystemsandelementsincludingtherelationshipbetweenthemandto their environment. Consequently, in this thesis the production system isconsideredas“an interacting combination at any level of complexity, of people, material, tools,machines, software facilities, and procedures designed to work together for somecommonpurpose”(Chapanis,1996,p.22).The definition emphasizes the need for taking a comprehensive view of allsubsystems,theirelementsandtheirrelationswhendesigningaproductionsysteminordertominimizetheriskofsuboptimizationbyhavingaone‐sidedfocusononeoftheproductionsystemsubsystems.Groover(2008)identifiesfoursubsystemsoftheproductionsystem: Technical system – represents the hardware that is directly linked to the

productionprocessincludingmachines,tools,fixtures,etc. Materialhandlingsystem– represents thehardware that is related to loading,

positioning, and unloading as well as transportation and storage betweenstations.


11

Humansystem–representsdirectandindirect labourrequiredtooperateandmanagetheproductionsystem.

Control system – represents the planning and control capabilities required tocoordinatetheothersystemelements.

2.1.2 AsystematicdesignprocessThe introduction stated that the designing of production systems is important inorder to achieve fast and effective product introductions to themarket. Once theproductionsystemisinoperation,theabilitytomakemajorchangesislimiteddueto cost and time restrictions. The work of designing the production system isregardedasaprocess.Aprocess is “anetworkof interrelatedactivities thatare repeated in time,whoseobjectiveistocreatevaluetoexternalandinternalcustomers”(BergmanandKlefsjö,2010,p.42).

In contrast to the system perspective, which facilitates the understanding of thecomplex production system (Checkland, 1999), the process perspective supportsthecoordinationofthework(KeenandKnapp,1996).Thus,theproductionsystemdesignprocess is understood as a tool needed tomanage and support thedesignactivities. It is important tonote that the taskofdesigningaproductionsystemisoftencarriedoutinaproject.Aprojectreferstoatemporaryendeavourundertakento solveaunique task suchas thedesignof theproduction systemwithinawell‐definedtimeframe,whichshouldbeguidedbythecompany’sdesignprocess.Ingeneral,theprocessofdesigningaproductionsystemcanbedividedintoseveraldistinct phases comprising all necessary activities from an analysis to a detaileddesignoftheselectedsystemsolution.Onewaytostructurethedesignactivitiesisto separatebetweenapreparatorydesignphase and adesign specificationphase(Bellgran,1998;BellgranandSäfsten,2010).The formerphase isvery crucial forthepossibilityofdesigningproductionsystemsthatsuit thepreconditionsofeachcompanyandsituationandmainlyinvolvesanalysis,whilethelatterphaseinvolvestheutilizationofbothcreativityandanalysis.Each phase can be further divided into subsequentwork activities. The first andsecondstepsarepreparatorydesignactivitiesand include lookingbackwardsandinwardsinordertobringobtainedexperienceintoforthcomingproductionsystemsbutalsolookingaheadandoutwardsaimingatcapturingthecompany’sgoalsandstrategies into the production system design. The latter three steps (steps 3‐5)concern thedesignspecification,whichdealswithactivities important to createacomplete and appropriate system solution. Thus, each step in the design processincludesdifferentactivitiesthatneedtobecarriedout.Figure2belowreviewstheactivitiesthatshouldbecarriedoutinthedifferentdesignphasesassuggestedbyseveral scholars (Bellgran and Säfsten, 2010; Cross, 2000; Pahl and Beitz, 1996;RoozenburgandEekels,1995;UlrichandEppinger,2007;Wu,1994).TheactivitiesshowninFigure2areillustratedinasequentialflow.However,inordertoachievean effective and efficient production system design process, it is essential toemphasize the necessity for an iterative process with many cycles and partlyoverlappingactivities.

FRAMEOFREFERENCE

12

Figure2 Typical work activities carried out when designing the production

system.AccomplishingtheworkactivitiesspecifiedinFigure2requirestheinvolvementofdifferent functions2 that contributewith inputanddecisionmaking.However, themanufacturing company does not need to be responsible for all work activitiescarried out in the production system design process; rather the manufacturingcompany has the choice among internal designers and external designers or acombinationofboth(Bellgran,1998).Byutilizingexternalexpertise, thecompanycan benefit in the design process from new and innovative ideas and detailedknowledge.Oneactivitythatisfrequentlycarriedoutincollaborationwithseveralindustrial actors is the design and subsequent building of the productionequipment,i.e.thereisageneraltrendtowardsacquiringtheproductionequipmentfromexternalequipmentsuppliers.As early as the 1970s, Abernathy and Wayne (1974) pointed out that verticalintegrationexpandsandspecialization inproductionequipment increasescausingan increase in capital investment. Equipment suppliers are sources of major2Sincedepartmentscanincludemorethanonefunctionandalsoevolveovertime,inthisresearchthe term function is used to denote responsibilities and work areas required to design theproductionsystem.


13

innovations inmanufacturing technology forwhich the incentivesaregreaterandadopted by the larger user firms (Hutcheson et al., 1996; Reichstein and Salter,2006). Further, it has been argued that equipment suppliers need to take moreresponsibility forrefiningexistingtechnologyandimprovingequipmentreliabilityand capabilities (Hutchesonetal., 1995).However, those companies that procureproduction equipment become dependent on the equipment suppliers’ efforts toprovidetheequipmentandtosecureorimprovetheoperatingperformanceoftheequipment (Lager and Frishammar, 2010). Thus, the design and building ofproduction equipment is of significance for themanufacturing industry and oftenaccountsforafairlylargeshareofcostsinproductiondevelopmentprojects.Long‐term collaboration between the manufacturing company and the equipmentsupplier is therefore of great value for successful acquisition of productionequipment(Bellgran,1998).Inordertoensurelong‐termcollaboration,effortshavebeen made to define a structured and systematic acquisition process for theproductionequipment(JohanssonandNord,1999;RönnbergSjödinandEriksson,2010).2.1.3 DifferentapproachestotheproductionsystemdesignprocessAdistinctioncanbemadeconcerningtheapproachtakentotheproductionsystemdesignprocess,whichaffectstheactivitiesthatarecarriedoutbythemanufacturingcompany. Expanding on Wu (1994) and Engström et al. (1998), Säfsten (2002)identifiesthreemainapproachestothedesignofproductionsystems: The concept‐generating approach – The design process is driven by different

constraintssuchastypeofproduct,volume,andnumberofvariants. The concept‐driven approach – The design process is driven by something

externalsuchasapre‐existingdesignortheinterestofanactor. The supplier‐driven approach – The design process is driven by an external

supplier suggesting possible alternatives based on more or less detailedrequirementspecifications.

Thethreeapproaches,seeFigure3, implydifferentdegreesofinvolvementbythemanufacturing company in the production system design process. Figure 3illustrates that in a concept‐generating approach the manufacturing company isresponsible for all activities from the analysis of the situation to a completeproductionsysteminoperation.Ontheotherhand, inasupplier‐drivenapproach,thesuppliertakescareofpartsoftheactivities.Inthemostextremecaseallworkactivitiesareoutsourcedtoasupplier.However,eveninsituationswherethedesignoftheproductionsystemiscompletelyoutsourcedtoasupplier,itseemsnecessaryto maintain certain competencies also within the manufacturing company. Forexample, Hobday etal. (2005) or VonHaartman andBengtsson (2009) point outthattobeabletobenefitfromsupplierintegration,manufacturingcompanieshavetopossesscorrespondingin‐housecompetences.

FRAMEOFREFERENCE

14

Figure3 Three different approaches to the design process from themanufacturingcompany’sperspective(adaptedfromSäfsten,2002).

Further, it is widely acknowledged that product and production system designactivitiesshouldbeintegratedinordertoreducethetimerequiredforintroducingnew products on the market (Andreasen and Hein, 1987; Magrab et al., 2010).GerwinandBarrowman(2002,p.939)defineintegratedproductdevelopmentas“amanagerial approach for improving new product development performance (e.g.,development time), which occurs in part through the overlap (partially orcompletely parallel execution) and the interaction (exchange of information) ofcertain activities in the NPD process”. As a result, several issues related to thedevelopmentoftheproductareconsideredsimultaneouslyratherthansequentially.However, in contrast to a non‐overlapping and non‐interacting development ofproducts, an integrated approach also increases the need for coordination. Onepossibilitytoensureahighdegreeofcoordinationofthedifferentworkactivitiesistheabilitytoapplyaproductdevelopmentprocess.Aproductdevelopmentprocessdescribes the sequence of steps and activities the company has to deploy (UlrichandEppinger,2007).Thus,theproductionsystemdesignprocesscanbeconsideredas a part of the new product development process. Prior research on productdevelopmentbestpracticeshighlights that successfulprojects followa formalizedand structured cross‐functional stage‐gate model for the product developmentprocess (Griffin, 1997). Cooper (2008) describes that a stage‐gate process in itssimplest form consists of a series of stages which are followed by gates. In thestagestheprojectteamundertakestheprescribedworkactivities,whileinthegatesdecisionsaremadeonwhethertheprojectshouldcontinueornot. Includingboththeproductandtheproductionsystemdesigninthesameprocessrequirescreatinga balance between the two design processes by not solely focusing on either theproductortheproductionsystemdesign.Further, as the production system is hierarchical and consists of a number ofinterrelatedelements, therearemanysimilaritiesbetween theproductionsystem


15

and a complex product. Hobday (1998, 2000) defines complex products as high‐technology,business‐to‐business capital goodswhereeachproduct isofhigh costand made up of many interconnected, often customized parts, designed in anhierarchical manner and tailor‐made for specific customers. Due to high costs,physicalscale,andcomposition,complexproductsaretypicallyproducedwithinarecognizable, singleprojector small batches (Hobday,1998,2000). In linewith acomplexproduct,theproductionsystemisoftenaone‐offthatishighlyadjustedtothe requirements placed upon it and whose creation requires a high degree ofknowledge. However, one should be aware that there is also a fundamentaldifferencebetweenacomplexproductandtheproductionsystem.Theproductionsystemisnotaproduct that isusedbyhumans;ratherthehumansubsystem isapart of theproduction system (Bennett andForrester, 1993;Groover, 2008). Theimplication of the difference is that although a great deal of relatedwork can befound in the newproduct development theory, not all theorymay be validwhendesigningaproductionsystem.2.1.4 DifficultiesintheproductionsystemdesignprocessAlthough manufacturing companies have started to focus on production systemdesign,many find it difficult to coordinate the production system design processand work in a structured and systematic way; see discussions by Bellgran andSäfsten (2010) or Cochran et al. (2001/2002). Multiple explanations for thedifficulties in production system design are possible. First, the nature of theproductionsystemdesignprocessisnotwelldefined,i.e. therearemanydifferentdefinitionsandinterpretationsoftheprocessandworkactivitiesinvolved(Cochranetal.,2001/2002).Partofthereasonisthatcompanieshavefocusedontheproductdesignbecause theysaw itasaway toachievecompetitiveadvantages,while theproduction system design process is seldom seen as ameans to achieve the bestpossibleproductionsystem(BellgranandSäfsten,2010).Thus, although the termdesignprocess iswellknown inmanufacturingcompanies, it isusuallyapplied inthe product design andnot in the system required toproduce theproducts. It is,however,importanttonotethat“processdevelopment3isatechnicallydifficultandorganisationally complex activity on its own right, and operates in amuch richercontext than is generally portrayed in the simultaneous‐engineering literature”(Pisano,1997,p.31).Asaresult,thedesignofproductionsystemsneedstotriggerseparatecontrolandcoordinationofthespecificsetofactivitiesrequiredtomovetheprojectalongfrominceptionanalysistodetaileddesign.Second,ageneralimplicationofthedefinitionoftheproductionsystemistheneedto extend the activities of the production system design process beyond those ofproductionequipment,plantlayout,andjobdesign(Love,1996).Applyingaholisticperspective increases the complexity of the production system design processsignificantly, implies that the design itself is organizationally complex, and spansovermultiplefunctions.Consequently,productionsystemdesignprocessesrequire

3The termprocessdevelopment isdefinedascreatingandrefininganorganization’scapability tomanufacture a product or set of products commercially, which is similar to the term productionsystemdesignappliedinthisthesis.

FRAMEOFREFERENCE

16

setting up a project team with members from different specialized functions tohandleboththerequiredskillsandthevolumeoftheworkinvolved(Love,1996).Third,Chryssolouris(2006)pointsoutthatatthetimeaproductionsystemshouldbe designed, the objectives of a production system are neither well‐defined noraligned to changes. Further, information about the elements of the productionsystemsuchastheproductionequipmentormaterialhandlingsystemisimpreciseatthebeginningoftheproductionsystemdesignprocess,particularly if there isaneedtohandleahighdegreeofnovelty(Chryssolouris,2006).Inordertominimizethedegreeofnovelty,itissuggestedthatproductionsystemdesignersshouldreuseandretainasmany thoughtsandconceptsaspossibleofpreviouslydesignedandimplementedsystems(HubkaandEder,1988).Finally,theproductionsystemdesignprocessisinterdependentwiththecompany’smanufacturingandbusinessstrategy(BennettandForrester,1993;Pisano,1997).Thus,thedecisionsmadeintheproductionsystemdesignprocessshouldstrivefortheachievementofbothaninternalandexternal(environmental)fit(e.g.Choudhariet al., 2010; Miller, 1992; Ruffini et al., 2000). An internal fit ensures that theelementsof theproductionsystemaremutuallysupportiveandthat thedecisionsabouttheproductionsystemdesignalignproperlywiththecompany’soperationalgoals and objectives, while an external fit provides consistency between thecapacities and capabilities of the production system and the context. However,accomplishing an internal fit among the different subsystems of the productionsystem poses challenges due to the large number of design variables and theirinterdependence. As has been argued by Bozarth and McDermott (1998), toaccomplish an internal fit one must consider simultaneous, complex interactionsamong a wide range of interdependent variables when designing the productionsystem.Ontheotherhand,inordertoobtainanexternalfitrequiresthatthecapabilityofthe production system corresponds to the competitive priorities of the company(Miltenburg, 2005). One possibility is to consider the market and context,manufacturingstrategy,productconcept,etc. (BennettandForrester,1993;Duda,2000).As theconditionschangeover theproduct lifecycle, it isalsonecessary toidentifyorderwinnersandorderqualifiersforeachphaseoftheproductlifecycle.For example, Aitken et al. (2003) and Slack and Lewis (2008) illustrate theimportanceofconsideringdifferentcompetitivestrategiesduringdifferentphasesof the product life cycle. Cost, for example, may in the introduction and growthphasebeanorderqualifier,whileitcanbeanorderwinnerinmaturityandmarketsaturation phases. However, any change in the external fit may also demandsuitablemodificationstotheproductionsystemdesigntoprovideaninternalfit.Although one key requirement in the production system design process is toachieveinternalandexternalfit,thereisthepossibilitythatthedemandcancauseconflictsaswheneffortstomaintainenvironmentalfitpreventordestroyinternalcomplementarities, or when the emphasis on internal consistency detractsmanagers from changes outside the organization (Miller, 1992). Thus, achievinginternal and external fit is not an easy task. Further, concerns have been raisedwhether there ismore thanoneproduction systemdesign that can represent thebestfit,asdifferentproductionsystemsmayresultinsimilarperformance(Ruffini,


17

1999).SimilarresultshavebeenfoundbyDraaijerandBoer(1995),whofoundthaton the one hand similarmarket demands were satisfied althoughmanufacturingcompanieshadorganizedtheproductionsystemsverydifferently,andontheotherhand very different sets of market demands were met by similar productionsystems.Thus, it isnotpossibletopointoutanexactwayofhowamanufacturingcompanyshoulddesignitsproductionsystems; insteadtherearemultiple,equallyeffectivewaysinwhichamanufacturingcompanycanachieveinternalandexternalfit (Van de Ven and Drazin, 1985). As has been argued by Bellgran (1998), theresulting production system is the product of how people, influenced by theorganizational and historical context, apply the available options to design theproductionsystem.Overall, it can be concluded that the literature reviewed so far provides manyimportant insights into theproduction systemdesignprocess and challenges thatneed to be met. However, the above‐discussed research provides only limitedinsights into how those challenges can be handled by critical resources such asdesign information. This thesis contributes to the important work conducted byBennett and Forrester (1993), Ruffini (1999), and others by investigating thecritical role attributed to design information in the production system designprocess.2.2 THEMANAGEMENTOFDESIGNINFORMATION

2.2.1 DefiningdesigninformationThe term information is used in a variety of ways and is difficult to define. Forinstance, Rauterbeg and Ulich (1996) present six different interpretations of theterminformation.Itisseenthatinformationisoftendefinedinrelationtothetermsdata and knowledge. Data, information, and knowledge can be arranged in acontinuum(Davenport,1997;Kahnetal.,2002),wheredifferencesarebasedontheextenttowhichtheyreflecthumaninvolvement.Information isdefinedas “collection of data, which, when presented in a particular manner and at an appropriate time, improves the knowledge of the person receiving it in such a way that he/she is better able to undertake a particular activity or make a particular decision”(Galliers,1987,p.4).Galliers refers to thedifferencebetweendata, information, andknowledge,whichhas two important implications. First, information is enlightening and has realmeaninginagivencontextorsituation, i.e. informationiscontextualandenabling(Galliers, 1987). Second, because knowledge is valuable information from thehuman mind (Davenport and Prusak, 1998), it consists of truths and beliefs,perspectivesandconcepts, judgmentsandexpectations,methodologiesandknow‐how.Toeffectivelyusesuchassetsrequirestheuserofinformationtoacknowledgeandapplyinformationduringtheproductionsystemdesignprocess.Inthisthesis,thetermdesigninformationisusedtodenotetheinformationneededtocarryoutthenecessarydesignactivities.In general, manufacturing companies need to have the capability to deploy,integrate,andprotectthedesigninformationresourceinaneffectiveway,i.e.they

FRAMEOFREFERENCE

18

need to manage design information. Three reasons have been found why designinformationisnotincorporatedinthedesignprocess:1. Informationisnotacquired(Cooper,1975;Omaretal.,1999).2. Information is not shared among different specialized functions (Sivadas and

Dwyer,2000;Souder,1988).3. Information isnotused inspiteofbeingacquiredandshared(Deshpandeand

Zaltman,1982;Zahayetal.,2011).Basedontheseconclusions,priorresearchinnewproductdevelopmentpointsoutthat management of information should not be considered as a single one‐dimensional construct; rather the managing of design information is amultidimensional construct consisting of the three dimensions acquiring, sharing,andusing(Frishammar,2005;FrishammarandYlinenpää,2007;OttumandMoore,1997).Forthedesigningoftheproductionsystem,itmeansthatthemanagementofdesigninformationshouldbeconsideredfromthreedimensions,seeFigure4.

Figure4. Modelof the threedimensionsofmanagingdesign information in the

production system design process (based on Frishammar andYlinenpää,2007).

2.2.2 AcquiringdesigninformationAcquiringof design information refers to the gatheringof relevant andnecessaryinformation required to make the design process more effective and efficient(Frishammar and Ylinenpää, 2007). In general, the definition of the productionsystem and the identified activities carried out in the design process imply thatmultiple types of design informationneed to be considered in the designprocessranging from strategic to operational information. The need to differentiatebetween strategic and operational information is confirmed by Bruzelius andSkärvad(1989),whocategorizeinformationintobroadanddeepinformation.Theformer concerns the strategicmoregeneral information required,while the latterrefers to the operative information and is often deep in details. Informationmayalso be classified as operational or background information (Petersson andPetersson,1992).Thepurposeofoperational informationisto facilitatefulfilmentofthetask.Backgroundinformation,ontheotherhand,aimsatprovidingabettercontextforthedecisionmakingandstrengtheningthemotivationoftheemployees.


19

Inaddition,Frishammar(2003)andHäckner(1988)classifyinformationintohard(quantified)andsoft(qualitative)informationanddiscoverthatcompaniesusuallyrelyonacombinationofbothhardandsoftinformation.Hardinformationregardsnumericalinformationthatcaneasilybequantifiedandprocessedwiththehelpofanalytical models, whereas soft information refers to images, visions, ideas, andcognitive structures (Häckner, 1988). In addition, soft information may becharacterized as broad, general, and subjective, as it concerns holistic images ofrealityandislinkedtoindividuals(Shrivastava,1985).It has been argued that the type of information required depends on taskcomplexity and previous experience (Byström, 1999; Fjällström, 2007). Taskcomplexity refers to whether the event is planned or unplanned, and previousexperiencerefers towhether thehandlingof theevent isknownorunknown,seeFigure 5. Fjällström (2007) found thatwhen a routineproblemoccurred, domaininformation,i.e.informationaboutfacts,concepts,andtheoriesinthedomainoftheproblem and problem solving information was available, while in cases of novelproblemsthishadtobefiguredoutbythepeopledealingwiththeproblem.

Figure5. Task complexity and earlier experience (R = routine problem, T‐f =

trained‐for problem, N = novel problem) (based on Fjällström, 2007;Säfstenetal.,2008).

Another classification is made by Zahay et al. (2004), who distinguish betweendifferent information types according to their origin, i.e. whether the designinformation was internally developed or obtained from the external context.Consequently,designinformationmaybeobtainedfromdifferentsources.Asourceis expected to contain the relevant design information required for performing aworkactivity.Sourcescannotonlybedividedintointernalandexternalsources,butalso into personal and impersonal sources (Aguilar, 1967; Daft et al., 1988). Theformer include direct human contact, while the latter regard written/non‐verbalsourcesofdesigninformation.Amoredetailedclassificationofimpersonalsourceshas been provided by Johansson (1999) andHarlin (2000)who separate processand documentation as impersonal sources. Documentation refers to any type ofwritten information, while process regards information that can be obtained byanalysing existing production systems. Aguilar (1967) found that although thereceiver of information often depends on a combination of information sources,personalsourcesfarexceedimpersonalsourcesinimportance.Similarresultshave

FRAMEOFREFERENCE

20

beenprovidedbyFjällström(2007)andHertzumandPejtersen(2000),whostatethat personal sources are the preferred choice. Further, the results of previousresearchindicatethatinternalsourcesarefavouredincontrasttoexternalsources.For example, Aguilar’s (1967) study shows that even in cases of externalinformation, i.e. informationabout the company’s outside environment,managerstend to rely almost as much on inside sources than on outside sources. Thisphenomenonmaybepartlyexplainedbythefactthatperceivedsourceaccessibilityispositivelyrelatedtothefrequencyofusage(Sawyerretal.,2000).2.2.3 SharingdesigninformationTo Frishammar and Ylinenpää (2007), information sharing is the transfer ofinformationacrossspecializedfunctions.Thesharingofdesigninformationisakeyfor successful development since it reduces uncertainty and equivocality amongparticipants in development projects (Daft and Lengel, 1986) and allows forintegration between the specialized functions (Moenaert and Souder, 1990;Turkulainen,2008).A key implication fromprior research is that themanagementof uncertainty andequivocality has different information needs (see e.g. Daft and Lengel, 1986;Frishammaretal.,2011).Uncertaintycanbedefinedas“thedifferencebetweentheamount of information required to perform a particular task, and the amount ofinformation already possessed by the individual” (Galbraith, 1973, p. 5).Consequently,uncertaintyrequirestheprovisionofadditional/newinformationtoclose the information gap causing uncertainty. The assessment of uncertainty isbasedonaspectssuchasclarityandtheamountofdetailthatisavailable(Galbraith,1973). Daft and Lengel (1986) propose that in order to reduce uncertainty,companies have to ask a large number of questions and gather additionalinformation to seek answers to the questions. They conclude that to support theamount of information needed to dealwith uncertainty and achieve desired taskperformance, it is possible to implement specific structural mechanisms. Overall,the reduction of uncertainty requires the provision of new but target‐orientedinformation(Moenaertetal.,1995).Equivocality,ontheotherhand,isprimarilyassociatedwithambiguity,i.e.multipleandconflicting interpretationsamongprojectparticipants(Weick,1995). Inorderto reduce equivocality among project members, the exchange of subjectiveinformationisrequired.Insituationsofhighequivocality,thekeyistoconstructorenactareasonableinterpretationthatmakespreviousactionsensibleandsuggestshow to proceed (Daft and Weick, 1984). Thus, additional information does notcontributetotheresolvingofmisunderstandings;ratherequivocalityisresolvedbydefiningorcreatingananswer(Weick,1995).Ithas,therefore,beensuggestedthatthecommunicationmedium,i.e. themediumbywhich information is transferred, has to be carefully chosen since the appliedcommunicationmediumhasconsequencesonthecapacitytoprocess information.Forexample,HertzumandPejtersen(2000)pointoutthatwritteninformationsuchas documents lack information about the surrounding context, i.e. they missexplanations about why specific decisions are made as well as what purpose isserved. Therefore, in situations of high equivocality the processing of richinformation, i.e. informationthatwillchangeunderstandinginatimelymanner, is


21

crucial (Daft and Lengel, 1984, 1986). Consequently, there is a need to use acommunication medium that allows for immediate feedback, debate andclarification such as face‐to‐facemeetings.On the other hand,when the aim is tominimizeuncertainty,a lessrichmediumis recommended, i.e.amediumthatcantransferrather largeamountsof informationsuchasdocuments(DaftandLengel,1984, 1986). This category is effective for the processing of well‐understoodmessages and data since it involves fewer cues and thus restricts feedback. Thediscussion above is summarized in Figure 6, which illustrates the differentstructural characteristics a company can implement to facilitate the sharing ofeitherrichorlessrichinformation.

Figure6. The different needs of information sharing to cope with uncertainty

andequivocalityreduction(adaptedfromDaftandLengel,1986).Furthermore, it has been argued that a company’s ability to effectively integratecross‐functionalactivities isbasedontheprocessofsharingreal‐timeinformationinordertoidentifyandsolveproblems(MoenaertandSouder,1990;Wheelwrightand Clark, 1992). Nevertheless, although the transfer of information is essential,researchbyOttumandMoore(1997)showsthattheinformationcollectedbyonefunctionisnotautomaticallysharedwithotherfunctions.Thediversityoffunctionsimpliesahighdegreeofdifferentiationand interdepartmentaldifferencescausingproblems in the sharing of information (Lawrence and Lorsch, 1986; VandeveldeandVanDierdonck,2003).GriffinandHauser(1996)arguethatsubtleterminologydifference between different functions may imply vastly different solutions. Theterminology and detail used can cause frustration when people with differentexpertise and skills share information (Vandevelde and VanDierdonck, 2003). Ingeneral,itbecomesmoredifficulttoshareinformationamongpeoplewithdifferentfunctionsandbackgroundsthemoreabstracttheinformationis(Jacobs,1996).Forpeoplewith thesamebackgroundandknowledge it ismucheasier tounderstandthesharedinformation.Prior research emphasizes the important role dedicated to external equipmentsuppliers in the production system design process (Reichstein and Salter, 2006;Skinner, 1992). However, the sharing of information should become even morechallengingwhenexternalcompetenciessuchasequipmentsuppliersareincludedin the design process. Several studies illustrate that distance and organizational

FRAMEOFREFERENCE

22

bondshavean impacton informationexchange (e.g.Dankbaar, 2007;VandeveldeandVanDierdonck,2003).Thesestudiesarguethatphysicalproximity,i.e.ashortdistance, is beneficial for the exchange of information. For example, Allen (1977)points out that the frequency of information sharing among people normallydecreases in linewith increased physical separation. It has been highlighted thatpeople working in dispersed teams are more likely to ‘drop’ members who aredistant (Mortensen and Hinds, 2002).With an increase in distance, informal andface‐to‐facecommunicationbecomesinconvenient(VandeveldeandVanDierdonck,2003).Another reasonwhy the sharingof information shouldbecomeevenmorechallenging is organizational bonds. These bonds can hamper the sharing ofinformation because of a lack of clarity, roles, and responsibilities (Gupta et al.,1987).Based on the discussion above, it can be concluded that there is a correlationbetween the sharing of information and the integration of specialized functions.Interaction and collaboration are two important elements of integration betweendifferent functions. Interaction is associated with structural and formallycoordinatedactivitiesincludingroutinemeetings,plannedteleconferencing,andtheflow of standard documentation (Kahn, 1996). Collaboration, on the other hand,refers to the unstructured, affective nature of relationships between differentfunctions and stresses the importance of mutual understanding, common vision,shared resources, and achievement of collective goals in a processwhere severalfunctionsworktogetherinamutual/sharedprocess(Kahn,1996).FrishammarandHörte (2005) argue that although interaction is primarily concerned with thesharingofinformation,alsocollaborationisanappropriatemeasureofinformationsharingsince theunderlying foundation forahigh levelof collaboration isagoodflow of information between the different functions. Consequently, integration isapproachedasaninformation‐sharingphenomenon,whereshortagesoccurduetoa lack and asymmetry of information or an inability to share information(Turkulainen,2008).Further, it can be concluded that integration can be promoted by structural andformal coordination activities among functions and through amore unstructuredinformal coordination process that stresses continuous relationship (Frishammarand Ylinenpää, 2007). Coordination can be defined as “all informal and formalmechanisms that establish and integrate the roles of project participants and itinvolvesthetimingandfrequencyofactivitiesthatarerequiredtomeettheproductgoals”(OlaussonandMagnusson,2011,p.283).Formalizationconcernsthedegreeto which rules or standard operating procedures are used to govern interactionbetween different functional areas (Ruekert and Walker, 1987). Expanding onThompson(1967)andAdler(1995),VandeveldeandVanDierdonck(2003)discussfourcategoriesofformalcoordination: Standardsorrules Plansandschedules Formalmutualadjustments Dedicatedteams


23

In a study of the design‐manufacturing interface in the UK automotive industry,Twigg(2002)foundthatthosecoordinationmechanismscanalsobeappliedwhencompany bounds are crossed. Informal coordination mechanisms, on the otherhand, concern formless relations that cut across formal structures or informalcommunication supplementing formal (Martinez and Jarillo, 1989). Overall, it hasbeenargued that informationobtained through informalpaths is far greater thantheinformationobtainedthroughformalpaths(BruzeliusandSkärvad,1989).2.2.4 UsingdesigninformationUsing design information in this thesis is regarded as responding to or takingappropriate actions on the acquired and shared information (Frishammar andYlinenpää,2007).Ingeneral,itisimportanttoensurethattheinformationisofhighvaluetotheuser, i.e. theuserhastoperceiveinformationasuseful forthetaskathandtoactuallyemploythe informationinhis/herworkactivities(Eppler,2006).Onewayofestablishingahighdegreeofinformationutilizationintheworktasksistofocusonqualityandnotquantity,meaningthatmanufacturingcompaniesshouldstart to focus on information quality instead of gathering enormous amounts ofinformation. The goal has to be to improve the usefulness and validity of theavailable information rather than to overwhelm peoplewith information. Severalstudies,suchasthosebyKahnetal.(2002),LescaandLesca(1995),andStrongetal. (1997), have been undertaken to investigate the origin and consequences ofdeficiencies in information quality. It is clear from these studies that informationquality is amultidimensional construct that considers several qualitative criteriathatinformationshouldfulfiltoeffectivelymeetuserrequirements.Eppler(2006)suggestssixteendimensionsofinformationquality,whicharesummarizedinTable1.Table1. Informationqualitydimensionsandtheirdefinitions(Eppler,2006)Dimensions DefinitionsComprehensiveness Isthescopeofinformationadequate(nottoomuchnortoolittle)?Accuracy Istheinformationpreciseenoughandcloseenoughtoreality?Clarity Istheinformationunderstandableorcomprehensibletothetarget

group?Applicability Cantheinformationbedirectlyapplied?Isituseful?Conciseness Istheinformationtothepoint,voidofunnecessaryelements?Consistency Istheinformationfreeofcontradictionsorconventionbreaks?Currency Istheinformationup‐to‐date andnotobsolete?Correctness Istheinformationfreeofdistortion,bias,orerror?Convenience Doestheinformationprovisioncorrespondtotheuser’sneedsand

habits?Timeliness Istheinformationprocessedanddeliveredrapidlywithoutdelays?Traceability Isthebackgroundoftheinformationvisible(author,date,etc.)?Interactivity Cantheinformationprocessbeadaptedbytheinformationconsumer?Accessibility Isthereacontinuousandunobstructedwaytogettheinformation?Security Istheinformationprotectedagainstlossandunauthorizedaccess?Maintainability Canalltheinformationbeorganizedandupdatedonanon‐goingbasis?Speed Cantheinfrastructurematchtheuser’sworkingpace?

FRAMEOFREFERENCE

24

If problems in information quality are not handled adequately, there is a risk fordifficulties such as the inability to find the right information or to use theinformationinthetaskathand.Therefore,Eppler(2006)proposesanempirically‐basedframeworkfor informationquality,seeFigure7.The frameworkfocusesonimproving information quality in knowledge‐intensive processes, i.e. non‐routineprocesses with demanding requirements in terms of continuous learning andinnovation; processes that rely on the individual’s expertise and personalcontribution in the form of information. The information quality frameworkillustratedinFigure7comprisesfourelements(Eppler,2006): The first element consists of the four information quality levels: relevant

information,soundinformation,optimizedprocess,andreliableinfrastructure. The second element consists of four phases and concerns the life cycle of

information from a user’s perspective: information is searched and found(identification),evaluated,adaptedtoanewcontext(allocation),andapplied.

Thethirdelementconsistsofthesixteenidentifiedinformationqualitycriteria,which are placed along the phases according to their importance for thedifferentphases.

Thefourthelementconsistsofthemanagementprinciplesthatcanbeappliedtoimprove the quality of information in every phase: integration, validation,context,andactivation.

In Figure 7 the four information quality levels are further divided into contentqualityandmediumquality.This isdone inorder toemphasize that the first twolevels, relevance and soundness, relate to the actual information itself, while thelowerlevels,processandinfrastructure,relatetotheadequatenessofthechannelsbywhichinformationistransported(Eppler,2006).

Figure7. Informationqualityframework(adaptedfromEppler,2006).


25

It is important to note that using information can be particularly challenging inprojects where tasks are often processed in parallel, i.e. concurrent engineering(Terwieschetal.,2002). Johansson(2009)discusses indetail theconsequencesofconcurrentengineeringforinformationqualityandconcludesthattheneedtoworkwithoverlappingactivitiesseemstoaffectseveralinformationqualitycriteria.Oneexplanation is the need to use partial/preliminary information in developmentprojectswithoverlappingactivities,i.e.onecannotexpecttoreceivethesametypeof informationwith the same accuracy in development projectswith overlappingactivitiescomparedtoprojectsthatarecarriedoutinasequentialflow(SmithandReinertsen, 1998). Previous research points to a general reluctance amongengineers to release early information on the one hand and to use incompleteinformationontheotherhand(e.g.ClarkandFujimoto,1991;HauptmanandHirji,1996). In the traditionalmoresequentialdevelopmentprocess,where tasksoftenproceededinasequentialorder,informationcouldbefinalizedbeforeitwassharedwith other functions. However, the need to shorten the time to market requiresengineers to release and use preliminary information, which influences theprecisionandstabilityoftheinformation(Terwieschetal.,2002).Anotheraspect that facilitates theuseof information is thepragmatic levelof theinformation(Fjällströmetal.,2009;VonWeizsäcker,1974).Pragmaticinformationadds new knowledge to the receiver by combining pre‐knowledge of the matter(confirmation)withnovelty,i.e.newaspectsthatthereceiverdidnotalreadyknow.Forinformationtobeunderstood,ithastorelatetopre‐understandingandcontainconfirmation.When no confirmation exists, the receiver is not able to relate theinformation to any meaning and thus no understanding is possible (VonWeizsäcker,1974).Ontheotherhand,informationalsohastomakeadifferencetothe previous knowledge of the receiver, i.e. people expect to receive informationthatcontainssomedegreeofnovelty(AlmandFjällström,2003).2.3 SUMMARYOFFRAMEOFREFERENCEThe frameof reference reveals that there is only a small amount of research thatexplicitly focuses on the critical role of the design information resource in theproductionsystemdesignprocess.Themajorpartofresearchconcernseithertheproduction system design process or the management of information in newproductdevelopment.Thus,thereisaneedforfurtherresearchwithrespecttoatleast two areas when designing production systems: the required designinformationandthemanagingofdesigninformation.Thepresentedframeofreferencedoesnotonlydefinethefocusoftheresearchandits limits; it is also a central tool that guides the data collection and analysis. Toavoid ending up in a situation where the evidence does not address the initialresearch questions requires a framework of reference that directs the datacollection (Yin, 2009). In addition, the frame of reference is also used during theanalysis of the data, which is a continuous interplay between theory and data(Eisenhardt, 1989). Going into details regarding the researchmethodology is thethemeofthenextchapter.


27

CHAPTER3

:RESEARCHMETHODOLOGY

CHAPTERINTRODUCTIONThischapterdescribestheresearchmethodologyadoptedstartingwithadiscussionabout the research approach. Thereafter follows a description of the case studymethod including a discussion about research design, data collection, and dataanalysis. The chapter concludes with a discussion about the credibility of theresearch.

Theresearchprocessof thePhDstudiesstartedinJanuary2007andtheresearchcan be divided into two phases. The first phase of the research process aimed atidentifying and analysing information that supports a proactive behaviour ofassemblyoperators.The focusof theresearchwasonoperational information, i.e.the information required by the operator to plan, control, and execute the workactivities that need to be carried out once the production system is in operation.This phasewas finalized by the presentation of the licentiate thesis “InformationRequirements in a Proactive AssemblyWork Setting” (Bruch, 2009); the findingsaresummarizedinPaperI(BruchandKarltun,2009)appendedtothisthesis.Afinalconclusionfromtheresearchpresentedinthelicentiatethesiswasthatitisnecessarytoprovideoperatorswithsocalled‘whyinformation’toenableproactivebehaviour4(Bruch,2009;BruchandKarltun,2009).Thismeansthatoperatorsneedto get access to information that is related to a longer time horizon such asinformation about the introduction of new products leading to changes in theproduction system. However, the study also revealed that the manufacturingcompany had difficulties identifying and providing relevant and necessaryinformationrelatedtothedesignandimplementationofproductionsystems.Basedon the conclusions drawn, a further literature review was conducted, whichconfirmedalackofknowledgeconcerningthedesigninformationrequiredandhowtoactuallymanagethedesign informationwhendesigningtheproductionsystem.Inaddition,theliteraturereviewrevealedthepotentialbenefitsthatmanufacturingcompaniescangainbyfocusingonthedesignprocesscomparedtothemanagementandcontrolofanexistingproductionsystem.Asaresult,theresearchofthesecondphase,whichstartedinmid‐2009afterthepresentationofthelicentiatethesis,has4Foramoredetaileddiscussionaboutthefindingsoftheresearchpresentedinthelicentiatethesis,seeBruch(2009)andPaperI(BruchandKarltun,2009)appendedtothisthesis.

RESEARCHMETHODOLOGY

28

achangedfocusandthustheresearchmethodologyandfindingspresentedinthisthesisrepresent thesecondphaseof theresearchprocess.However, theactivitiescarriedoutduring the firstresearchphaseprovidedarelevantbackgroundto theinformationmanagementarea.3.1 RESEARCHAPPROACHTheobjectiveistodevelopknowledgetocontributetoaneffectivemanagementofdesigninformationwhendesigningproductionsystems.Understandingknowledgeas a contribution also needs admitting that knowledge is cumulative, i.e. theexpandingofknowledgeinanareamaygothroughdifferentphasesdependingonthe existing knowledge (Karlsson, 2009). Both themanagement of information innewproduct development (NPD) and theproduction systemdesignprocesshavebeenstudiedpreviouslybutthereisalackofstudiesfocusingonthemanagementofinformation when designing the production system. Therefore, it was foundnecessary tohaveanexplorative researchapproach.Thismeans that the study ismoredirectedtowardsknow‐whatresearchratherthantounderstandhowthingswork(know‐how)orwhyitisso(know‐why).Eveniftheproductionsystemdesignprocessisusuallycarriedoutinaprojectwithadefinedgoalandtimeframe,itcanbeconsideredasacomplexprocess.Notonlyarethenumerouselementsandsubsystemsoftheproductionsysteminterrelated,but the activities carried out in the design process are also related. Further, theproductionsystemdesignprocessrequiresinputfromvariousspecializedfunctionsincludingthosenotdirectly involvedinthe industrializationprojectandissubjectto influences from theexternal context.Thismakes itdifficult to explicitly isolatethe process of designing theproduction system from its surroundings. Therefore,the research problem has been approached from the system perspective, whichmeansthattheindustrializationproject,wherethedesignoftheproductionsystemisperformed,isthoughtofasasystem.Applyingasystemapproachinthisresearchhas some consequences to the results of the research. Although the systemapproach assumes the existence of an objective reality, the findings of a systemstudydonotresultinabsolutetheory(ArbnorandBjerke,2009).Thusthesystemapproachseesrealityasconsistingofcomponentsthataremutuallydependentoneachotherbutcannotbesummedup.Itisimportanttonotethatthedescriptionofthesystemisalwaysaninterpretationandsimplificationoftherealitystudiedbuthastobebasedonastudyoftherealsystem(ArbnorandBjerke,2009).While the system approach determines the conception of reality, the researchapproach is also influenced by the reality that is studied. This thesis can becategorized as applied research, meaning that the knowledge produced in theresearchshouldnotonlybeofscientificrelevancebutalsoofindustrialusefulness.Oneapproachemphasizingthecontributiontopracticaluseandusefulresearchforpractitioners concurrent with theory development is the interactive researchapproach. Interactive research emphasizes the need for collaboration betweenresearchandpracticeandaimsatjointlearningandcriticalknowledgeproductionbetween the participants and the researcher (Svensson and Nielsen, 2006).Interactiveresearchshouldnotbeconsideredintermsofthemethodsused;ratheritisanapproachtowardsacertainwayofunderstandingandconductingresearch,namely to do research together with, not on, the participants (Larsson, 2006).


29

Therefore, the researchwas influencedby the interactive research approach. Theinteractionwith the employees at themanufacturing companywas based on thecommon interest inunderstanding theproduction systemdesignprocess inmoredetailasafoundationforimprovements.Thecollaborationwiththemanufacturingcompanytookplaceover fourperiods: (1) introduction, (2)preliminary feedback,(3)participationindevelopmentactivities,and(4)evaluationandfeedback,whicharedescribedinmoredetailinSection3.4.1.Ininteractiveresearch,newknowledgeiscreatedduringvarioustypesofmeetingsandcloseinteractionbetweentheresearchersandpractitioners.However,despitetheimportanceofcooperationthatbenefitsbothparties,thenecessityofacademicindependence for interactive research is stressed (Svensson and Nielsen, 2006;WigrenandBrundin,2008).Althoughinteractiveresearchaimsatcomingascloseas possible to the phenomena studied, it does not point out the importance ofdistanceforreflection.Thus,althoughtheresearchprocesshasbeencollaborativetoaveryhighdegree,theprocessofwritingthepresentthesisandsummarizingtheacademicresultshasbeenanindividualone.3.2 RESEARCHMETHODIn order to answer the two research questions presented in Chapter 1, it isimportanttoselectasuitableresearchmethod.Theresearchmethodrepresentsawaytoplan for,collect,andanalyseempiricaldata.Larsson(2006)argues that ininteractive research different research methods can be used depending on thesituation and the research questions. For the present research, the case studymethodwaschosen.Thefirstrationaleforchoosingthecasestudymethodwastheneed to gather a rich set of data from actual practice in order to facilitate theunderstandingofthephenomenonstudied(Vossetal.,2002).Thelackofempiricalstudies exploring the management of design information when designingproduction systems called for a more flexible design strategy. Having a flexibledesign approach implies rigorous data collection procedures but the detailedframework of design emerges during the research progress (Robson, 2002). Bytaking a flexible approach, the research was not restricted by the clear‐cut andlinear researchdesignused in fixed studies; instead, therewas thepossibility forthe researchquestions to evolve, anddata collection options coulddevelopor beabandonedduringtheresearchprocess.Thesecondrationaleforchoosingthecasestudymethodwastheabilitytostudytheentireproductionsystemdesignprocessfrom both the strategic and the operational perspective. An investigation of theprocessonlyforagivenmomentintimewouldnothavereflectedthedynamicsofthe process. As Leonard‐Barton (1990) points out, a case study allows for theinvestigationofpastandcurrentphenomenafrommultiplesourcesofevidence.3.3 RESEARCHDESIGNOncetheresearchmethodhadbeenselected,thenextstepwastoplanandorganizetheresearch,i.e.todesigntheresearch.Theresearchdesignguidestheresearcherin collecting, analysing, and interpreting the observationsmade in the case studyandthusworksasaplanforhowtheresearchershouldgofrominitialquestionstothe conclusions answering the questions (Yin, 2009). Based on the conclusionsdrawninthelicentiatethesisandonanotherreviewoftheacademicliterature,the

RESEARCHMETHODOLOGY

30

initial research questions were formulated. Gaining pre‐understanding aboutgeneral constructs and categories that are of relevance for the research is aprerequisitewhenusingacasestudymethod(Vossetal.,2002).Althoughitisimportanttoformulateresearchquestionsanddevelopatheoreticalconstruct, it is commonthat theresearchquestionsevolveover timeand that thetheoretical constructs become modified. In addition, doing applied research of adynamicprocessmakesitdifficulttoplanfortheresearchprocessindetail.Makingadjustmentstotheobservationsmadewasalsonecessaryinthecurrentresearchtobeabletogiveacomprehensivedescriptionofaspectsthatplayedanimportantroleregarding the management of design information. The inherent flexibility of theresearch design can be a strength but should not be an excuse for inadequatespecifications of research questions or constructs nor should it lessen the rigourwithwhichcasestudyproceduresarefollowed(Vossetal.,2002;Yin,2009).Inthepresentresearchdesign,minorchangesweremadeduetotheneedtopursuenewleads. These changes did, however, not require a change in the overall researchdesignnorintheresearchobjective.Besidesformulatingtheresearchquestionsanddevelopingageneralunderstandingoftheareastudied,theresearchdesigninvolvesmakinganumberofdecisions.Inthe following, five aspects are discussed: number of cases, real‐time versusretrospectivecases,caseselection,unitofanalysis,andmixedmethodresearch.3.3.1 NumberofcasesTheresearchisperformedinthebeliefthatthemanagementofdesigninformationwhendesigningproduction systems is influencedby the specific situation.Ratherthan there being one best way of managing design information, manufacturingcompanieshavetoadjust themanagementofdesigninformationtothecontextofthe industrializationproject.Asa result, itwas foundnecessary to studymultiplecaseshighlightingthediversityofthephenomenonstudied.Inthis,itwaspossibletominimizetherisksoftenrelatedtoasingle‐casestudysuchasmisjudgingasingleevent,misinterpretationoftherepetitivenessofthecase,observerbiasesandlimitsintheabilitytogeneralizethecasestudyresults(Vossetal.,2002;Yin,2009).3.3.2 Real‐timeorretrospectivecasesTo study the managing of design information in the production system designprocess,eitherretrospectiveorreal‐timecasescanbeused.Theresultsofthethesisarebasedonacombinationofreal‐timeandretrospectivecases.Tworeal‐timecasestudies following two different industrialization projects were conducted. Byfollowing the twoprojects in detail, possible negative effects such as participantsnot recalling critical events were avoided, while at the same time the pre‐understanding of the problem studied increased (see, among others, Leonard‐Barton,1990;Vossetal.,2002).However,doingareal‐timecasestudyrequiresaconsiderableamountoftimebothfortheresearcherandtheparticipatingcompany,whichneedstoprovideaccessduringtheentireresearch.Therefore,concurrentlywith the two real‐time studies, two retrospective studies were conducted toincreasetheexternalvalidityofthefindings.Inaddition,thisreducedtheissuesoftime and access – two issues that are critical and pervasive in real‐time studies(Cross,2000;Robson,2002;Vossetal.,2002).


31

3.3.3 CaseselectionInordertoidentifyareferencepopulationofpossiblecasesthatcouldbeincludedinthedatacollection,atheoreticalsamplingwasused.Theobjectiveoftheoreticalsampling is to identify cases that replicate previous cases or extend emergenttheory(Eisenhardt,1989).Inthepresentresearch,fourdifferentsamplingcriteriawere used: 1) automotive industry, 2) responsibility for product and productionsystemdesign,3)stage‐gateproductdevelopmentmodel,and4)productionsystemtype.1. Theautomotive industryhas tohandle shrinkingproduct life cycles combined

with an increasing number of product models and variants, which has majorimplications on investment costs and cost per manufactured/sold product(Asnafi et al., 2008). Consequently, the automotive industry has to have theabilitytoeffectivelyandefficientlydesignproductionsystems.

2. The companies selected for the case studies should be responsible for bothproductandproductionsystemdevelopment.Thecriterionisrelatedtothehighinterdependency of these two functions, where product design engineersprocesslargeamountsofrelevantandnecessarydesigninformation.

3. Sincethefocusofthisresearchisthemanagementofdesigninformationratherthan the improving of the actual design process, it was important to selectcompaniesthathadanunderstandingoftheworkactivitiestobeperformedinthe design process. Another motive is that companies applying a stage‐gatemodelarealsoobligedtointegratetheworkactivitiesofdifferentfunctionsandthus themanagingofdesign informationbetweendifferent functions iscrucialtocarryouttheworkactivitiesofeachphase.

4. Designing a production system requires a holistic perspective. Therefore, theproduction system of the company had to include people and productionequipment of various degrees of automation, i.e. the focus was on semi‐automatedproductionsystems.

ThethreeselectedcompaniesandthecasesstudiedaresummarizedinTable2.Table2. Overviewofthecompaniesselectedandthecasesstudied CompanyI CompanyII CompanyIIICompanytype First‐tiersupplier Originalequipment

manufacturerOriginalequipmentmanufacturer

Studylocation Sweden England Sweden SwedenCasetype Real‐time Real‐time Retrospective RetrospectiveReferredtoas CaseB CaseP CaseE CaseT

Inaddition, to select cases,boundarieswereset related to the researchquestionsandtheresearchframeworktodeterminewheretheproblemoftheresearchcouldbestudied.Thefocuswasonfindingcasesthatallowedforcomparisonbetweenthecollecteddatabutstillallowedforgeneralizationofthefindings.ThisisinlinewiththerecommendationbyYin(2009)touseareplicationandnotasamplinglogicfor

RESEARCHMETHODOLOGY

32

theselectionofcases.Thegoalwastoselectcasesthatcapturedthewidestpossiblecoverage of activities, challenges, and strategies for the managing of designinformation,i.e.whatPettigrew(1990)callspolartypes,inwhichthephenomenonstudiedisclearlyobservable.Ifthenumberofcasesstudiedisrelativelysmall,itispreferable to choose cases that highlight the differences (Pettigrew, 1990).However, it isimportanttonotethatthecaseselectionwasalsoinfluencedbytheindustrializationprojects thatcouldbestudied, i.e. thereal‐timestudieshadto fitthetimelineoftheresearch.Basedonthechoicesofthecasesinvolvedinthedatacollection,thecasesrepresentdifferentattributesincludingscopeofthenewproductionsystem,investmentcosts,newness of the product, different divisions, and the location. As the researchprogressedandthedatacollectioncontinued,itbecameclearthatthecaseswereofpolar types. The differences were not only traceable to the different attributesdiscussed above, but also to other factors that became clear during the datacollectionsuchaspreviousdocumentationofdesign informationandorganizationoftheindustrializationprojects.3.3.4 UnitofanalysisThe unit of analysis refers to the fundamental problem of what the case is andshouldbe related to theway the initial researchquestionshavebeenposed (Yin,2009). In the research presented in this thesis a different unit of analysis wasselectedforthereal‐timecasestudiescomparedtotheretrospectivecasestudies.Thedesignofproductionsystemsisoftencarriedoutaspartoftheindustrializationprojectatmanufacturingcompaniesandthereforetheindustrializationprojectwaschosen as the unit of analysis for Case B and Case P. This means that theindustrializationprojectisusedtostudyallactivitiesthatareperformedtoachievetheobjectiveofdesigningaproductionsystem.Itisimportanttonotethatsincetheindustrialization project has a broader scope than the design of a productionsystem,thefocusofthecasestudywasontheearlyphasesoftheindustrializationproject. The context of the industrialization project is the company and the newproduct development project (see Figure 8). The numerous activities of theindustrializationprojectarecarriedoutbyrepresentativesfromdifferentfunctionsof the manufacturing company and even external equipment suppliers performpartsoftheprojecttasks.Theexternalequipmentsuppliersdonotnecessarilyneedtobeincludedintheprojectorganization;neverthelesstheactivitiescarriedoutbythe equipment supplier and those related to the acquisition of the productionequipmenthaveasubstantialimpactontheconceptualproductionsystemsolution.Sinceitispossibletohavemorethanonelevelofanalysisinacase(Yin,2009),theactivities carried out when designing the production system and acquiring theproductionequipmentareregardedassubunitsoftheanalysis.


33

Figure8. TheindustrializationprojectastheunitofanalysisofCaseBandCase

P.Theretrospectivecasestudiesarebasedonthecommonneedtoacquireproductionequipment from an external equipment supplier. Therefore, the productionequipment acquisition process is the unit of analysis for Case E and Case T. Thecontextoftheproductionequipmentacquisitionprojectisthecompany(seeFigure9).Althoughdifferentunitsofanalysiswereselected,thedataoftheretrospectivecasestudiescanbecomparedwiththedataofthereal‐timecasestudiessincethesubunitofanalysisinCaseBandCasePissimilartotheunitofanalysisinCaseEandCaseT.

Figure9. Theproductionequipmentacquisitionprojectastheunitofanalysisin

CaseEandCaseT.MixedmethodresearchIntheresearchpresentedinthisthesisthecasestudymethodwascombinedwithasurvey method, which provides several advantages and benefits. The surveymethodwasselectedtogatherfurtherinformationaboutthemanagementofdesigninformation concerning the external equipment supplier involved inindustrializationprojects and ishereafter referred to as Survey S. The case studymethod and the surveymethod can complement each other andwhen used in acombination can expand the scope and breadth of the research, giving a moredetailed understanding of a specific situation and leading to more precise andgeneralized results (Miles and Huberman, 1994). In a similar way, Yin (2009)arguesthattheuseofmixedmethodresearchcancontributetothecollectionofaricher and stronger array of evidence than can be accomplished by any singlemethodalone.Thus,bycombiningthecasestudymethodwiththesurveymethoditwaspossibletocombinetheperspectivesofboththemanufacturingcompanyandtheequipmentsupplier.

RESEARCHMETHODOLOGY

34

3.4 DATACOLLECTIONTheresearchpresentedinthisthesisisbasedonfourcasestudiesandonesurveywithdifferentcharacteristics:CaseBandCaseP–tworeal‐timestudies,CaseEandCase T – two retrospective studies, and Survey S. The studies partly overlapped;Figure10illustratesthedatacollectionprocessinachronologicalorder.Sincetheresearchcontextandstructureweredifferent,theprocessfordatacollectionvariedbetween thedifferentempirical studies.Therefore, thissectiondescribes indetailtheresearchcontextanddatacollection.Thedatacollectionforthereal‐timecasestudies,theretrospectivecasestudies,andthesurvey,respectively,isdescribedinmore detail in the following subsections (3.4.1‐3.4.3). In general, during the casestudiesdifferenttechniquesforcollectingdataweretriangulated.

Figure10. Datacollectionprocess.3.4.1 CaseBandCaseP:Real‐timecasestudiesBoth real‐time case studieswere carriedout atCompany I. Company I is a globalsupplier in theautomotive industryand isresponsible fordeliveryof the finishedproduct, product development, and continued technological renewal. Theindustrialization projects studied in Case B and Case P belonged to two differentdivisionsofthecasestudycompany.Ingeneral,eventsandactionswereobservedfirst‐handatthecompanyforatotalof37 days in Case B and 34 days in Case P. In Case B, data collection started inNovember 2009 and finished in August 2011; the entire industrialization projectwas studied but the focuswas on the preparatory and detail design phases.Datacollection in Case P reflected the preparatory design phase and took place fromFebruary2011toApril2011.Overall,beingatthecompanywasimportantforthedata collected. For example, on several occasions, project members discussedcritical aspects andpossibleproblem solutions at greater lengthduring the lunchbreakthaninprojectmeetings.Datawerecollectedfrommultiplesourcesofevidenceaimingatdatatriangulation(seeTable3forCaseBandTable4forCaseP).Inpractice,thismeansthatduringdata collection the same problem or fact was addressed by more than a singlesource of evidence. For example, one observation revolved around the fact thatdesigninformationwasdifferentlymanageddependingonwhethertheinformationwas shared among functions within the manufacturing company or with theequipment supplier. Inorder tounderstandwhy thedifferencesexistedand theirconsequences, the documented design information was studied, inquiries duringinterviewsweremade,andrelevantmeetingswereobserved.


35

Duringthedatacollection,adiarywaskeptwithobservationsandimpressionsaftereach day spent at themanufacturing site. The information recorded in the diaryincludedalsomethodologicalaspectssuchasthepersonsmetandthecontentandduration of the meetings. During data collection, the tentative results werecontinuouslydiscussedwiththetwosupervisors,whocouldreflectonthefindingsfromadistanceandgavevaluable input forthenextstepsthatshouldbetakeninthecollectionofdata.Bothcasestudiesstartedwithanintroductorymeetingtodiscusstheproposalandscopeoftheresearch.Atthemeeting,representativesfromdifferentorganizationallevelsofthecompanyaswellasoneofthesupervisorswerepresent,andsuitableindustrialization projects were selected. It was agreed to take a more holisticapproach inthe industrializationprojectsstudied.Astheresearchprogressed, thenatureofthedatacollectionchanged.AsdescribedinSection3.1,thecollaborationwiththemanufacturingcompanywascarriedoutinfourperiods:(1)introduction,(2) preliminary feedback, (3) participation in development activities, and (4)evaluationandfeedback.The introductionperiodaimedatbecoming familiarwitheachmanufacturingsiteand the studied industrialization project. Therefore, serial projectmeetingswereattended and numerous informal conversations with people from differentfunctions thatwere active in theNPDproject tookplace.The conversationswerestructured to the extent that the content of the discussion concerned the specificissuesofthefunctionintheNPDproject.Furthermore,inordertobecomefamiliarwiththeparticularprojectandtounderstandunderlyingassumptions,background,andscopeof the industrializationproject,documentswerestudied.Togetherwiththe employees responsible for the industrialization project, suitable respondentswhoaffectedorwereaffectedbytheindustrializationprojectwereidentified.Withthe respondents (ranging from top‐level management top operators) semi‐structuredinterviewswereconductedregarding Respondents’roleinthestudiedindustrializationproject Specificissuesandaspectsrelatedtothefunctionoftherespondent,which

shouldbeconsideredinthestudiedindustrializationproject AppliedworkmethodsandstandardsintheindustrializationprojectinNPD InformationhandlinginthestudiedprojectDuring the second period, i.e. the preliminary feedback period, findings from theintroductionperiodwerepresentedresultinginalistofissuestobeconsideredinthe current project. The feedback was used as a basis to identify suitabledevelopment activities and twodevelopment activitieswereestablished.The firstactivityconcernedthemethodsandstandardsusedintheindustrializationprojectandthesecondactivityrelatedtothedocumentationofexperiences.

RESEARCHMETHODOLOGY

36

Intheparticipationindevelopmentactivitiesperiod,theresearcherbecamepartofa group that aimed at improving the quality of the technical requirementsspecification.5Thus,thedevelopmentactivitieswererelatedtooneofthestandardsusedintheindustrializationproject.Asagroupmember,theresearchertookpartinthediscussionsconcerningthestructureandoverallcontent/aspectsthatshouldbeincluded in the technical requirements specification,while technical expertswereresponsible for the specification and detailed content of the requirements. Theresearcher reviewed and gave feedback on the proposed revised technicalrequirements specification. The second development activity was carried outwithouttheinvolvementoftheresearcher.Forexample,inCaseBaso‐called‘whitebook’ was developed, a report that documented experiences and lessons learnedfromtheindustrializationproject.In addition, as part of a development activity, semi‐structured interviews wereconductedandprojectmeetingswerefollowed.Theinterviewsincludedquestionsregardingthedevelopmentactivities,progress,andchangesintheindustrializationprojectandinformationhandling.Thefourthperiod–evaluationandfeedback–consistedmainlyofreflectionsonthecollected data. Less time was spent on‐site but contact continued by mail andtelephone to follow up the progress in the industrialization project. Further, theresultswerepresentedtothecompanyatfeedbackseminars,andinCaseBafinalevaluationoftheindustrializationprojectfourmonthsafterthestartofproductionwasmade. The ability of dissociating oneself from the studied research objectivewasimportantforreflectionanddrawingacademicconclusions.

5Thetechnicalrequirementsspecificationisanimportanttoolwhendesigningproductionsystems;itcontainsadetaileddescriptionofrequirementsthatshouldbemetinthedesignandsubsequentbuildingoftheproductionequipment.

MANA

GEME

NTOF

DESIG

NINFOR

MATIO

NINT

HEPR

ODUCTIO

NSYSTEMDESIG

NPRO

CESS

37

Table3.

Overv

iewofthed

atacollecte

dforCaseBd

uring37

daysbetween

Novem

ber2009a

ndAu

gust2

011

Techn

ique

No.

Duration

(minu

tes)

Typesofprim

aryda

tatha

twere

collecte

d

Passiveobservations:

NPDg

roup

930‐45

Status

andp

rogres

softh

eNPDproje

ct,inform

ationsharingbe

tween

thed

ifferen

tfunctions.

Indust

rializa

tiong

roup

1130‐90

Status

andp

rogres

softh

eindustrializationp

roject

,coord

inationb

etween

logisticsand

production

Equip

mentsup

plier

690‐60

0Techn

icaleq

uipme

ntdevel

opme

nt,identificat

ionofinfor

mationtob

einclu

dedinthetechn

ical

specification

Variousme

etings

1130‐18

0Inform

ationab

outth

eprod

uct,work

ingroutines,p

roject

status

Participantobservation:

Works

hop

475‐12

0Ma

nufactur

ingstr

ategy,id

entificat

ionoforderw

inners

,orderqualifiers

,ando

rderlo

sers

Lesso

nslea

rned,f

eed‐forward

5

60Pre

limina

ryfindin

gs/pro

position

s,continuouswo

rkTechn

icalsp

ecificationrevie

w5

40‐60

Feedback,po

tentialim

provem

ent

Face‐to‐faceinterviews

2440‐90

Roles,re

sponsi

bilitie

s,prob

lems,a

ndrequireme

ntsthatsho

uldbe

inclu

dedinthefutur

esyst

em

Respo

ndents:M

anage

rProd

uction

Engine

ering,P

roject

Manage

rIndust

rializa

tion,

Project

Manager

Advanced

Engine

ering,P

rogram

Manage

r,Opera

tions

Manager

,Facility

Manage

r,Quality

Assur

ance

Engin

eer,P

urchaser

,LogisticsM

anage

r,Ma

intenance

Engin

eer,V

ice

Presid

ent,P

roduct

Manage

r,Prod

uction

Engin

eer,W

orksho

pManage

r,Team

Leader,As

sembly

Opera

tor

Inform

alconversation

Daily

Widerange,e

.g.opinions,s

tatus,on

‐going

activitie

sDocum

ents

Fulla

ccess

RESEARCH

METHO

DOLOGY

38

Table4.

Overv

iewofthed

atacollecte

dforCasePd

uring34

daysbetween

Febru

ary20

11an

dApri

l2011

Techn

ique

No.

Duration

(minu

tes)

Typesofprim

aryda

tatha

twere

collecte

d

Passiveobservations:

NPDg

roup

545‐75

Status

andp

rogres

softh

eNPDproje

ct,inform

ationsharingbe

tween

thed

ifferen

tfunctions

Manufactur

ingen

gineer

inggroup

920‐90

Status

andp

rogres

softh

eprod

uction

devel

opme

ntpro

ject

Variousme

etings

1030‐30

0Inform

ationab

outth

eprod

uct,work

ingroutines,p

roject

status

Participantobservation:

Preparation

forw

orksho

p1

180

Struct

urean

dconten

tofaproduct

ionsyste

mdesignm

odel

Works

hop

1120

Struct

urean

dconten

tofaproduct

ionsyste

mdesignm

odel

Lesso

nslea

rned,f

eed‐forward

4

50‐70

Prelim

inaryfindin

gs/pro

position

s,continuouswo

rkTechn

icalsp

ecificationrevie

w3

35‐60

Feedback,po

tentialim

provem

ent

Face‐to‐faceinterviews

1430‐80

Roles,re

sponsi

bilitie

s,prob

lems,a

ndrequireme

ntsthatsho

uldbe

inclu

dedinthefutur

esyst

em

Respo

ndents:Pr

oduct

ionEn

gineer

ingManage

r,Advanced

Engin

eering

Proje

ctLeader,Qu

ality

Manager

/Prog

ramManage

r,Busi

nessAreaMa

nager

,SalesDirec

tor,Operation

sDirecto

r,Glob

alEngin

eering

Direc

tor,JigandTo

olDesigner,S

eniorDe

velopme

ntEngin

eer,Pr

ototyp

eOperator

Inform

alconversation

Daily

Widerange,e

.g.opinions,s

tatus,on

‐going

activitie

sDocum

ents

Fulla

ccess


39

3.4.2 CaseEandCaseT:RetrospectivecasestudiesThetworetrospectivecasesCaseEandCaseTwerebasedonthecommonneedtoacquire production equipment in industrialization projects. Particular emphasiswasplacedonhowtofacilitatetheacquisitionofmoreenvironmentallysustainableproductionequipmentbyformulatingenvironmentallysustainablerequirementsinthe technical requirements specification, which is used for the acquisition ofproductionequipment.6Thedatacollectedweresomewhatdissimilarwithrespecttotheamountofdatacollectedineachcompany(seeTable5andTable6).Thedatacollectedprimarilyderivedfromopen‐ended,semi‐structuredinterviewsbutweresupplementedbydocumentstudies.Mostoftheinterviewshadtworespondentsatthe same time due to time restrictions. During Case T, one of the environmentalmanagersalsoobservedtheremaininginterviews.Theinterviewguidewasdividedintoseveralareas: Overallacquisitionprocess Contentanduseofthetechnicalrequirementsspecificationinproduction

systemdevelopmentprojects Respondents’responsibilitiesandinvolvementintheacquisitionandthework

withthetechnicalrequirementsspecification Considerationofenvironmentallysustainablerequirementsinthetechnical

requirementsspecification Environmentallysustainableproductionsystems KeyperformanceindicatorsThedata analysed in the present researchmainly relate to the first three bullets,while theremainingaspectswerenot the focusof theanalysisbutprovidedsomevaluable background and context. All interviewswere face‐to‐face interviews andincludedseveralfollow‐upquestionsdependingonthecontext‐specificanswersoftherespondents.Further,theinterviewsweredonebytwointerviewers,butitwasagreedthatonepersonshouldconducttheinterview,whiletheotherobservedandtooknotes.Allinterviewswererecordedandtranscribed.Table5. Dataforthesemi‐structuredinterviewsduringCaseENo.ofrespondents Positionofrespondent(s) Interviewdate Duration(minutes)2 EnvironmentalManagerQuality

Environment,SafetyManager20Sept2010 70

2 TwoMaintenanceEngineers 21Sept2010 941 ProductionEngineeringManager 28Sept2010 622 ProjectLeaderProduction

DevelopmentStrategicPurchaser29Sept2010 88

1 TeamManagerIndustrialization 29Sept2010 39

6Theacquisitionincludedhardwareandsoftwareand,inthefollowing,‘productionequipment’willrefertoboth.

RESEARCHMETHODOLOGY

40

Table6. Dataforthesemi‐structuredinterviewsduringCaseTNo.ofrespondents Positionofrespondent(s) Interviewdate Duration(minutes)2 TwoEnvironmentalManagers 5Oct2010 882 EngineeringManager

EnvironmentalCoordinator5Oct2010 62

2 MachinePurchasingExpertEnvironmentalCoordinator

5Oct2010 64

3.4.3 SurveySThe focus of the survey method was to investigate the management of designbetweenequipmentsuppliersandmanufacturingcompaniesandtoidentifysuccessfactorsfacilitatingproductionequipmentacquisition.ThepreliminaryresultsfromCaseBwerethefoundationfortheformulationofaquestionnairetobeansweredby equipment suppliers. The random sample included 46 equipment suppliers.Twenty‐eight respondents returned the survey, of which three answers wereinvalid. It is important tonote that the surveywasof adescriptive characterandwasbasedonanon‐probabilisticsamplingof thecompanies.Adescriptivesurveytendstouseasignificantamountofqualitativedataandaimsatidentifyingtrendsratherthanmakingstatisticalinference(Forza,2009;Tanner,2002).The questionnaire contained 13 questions concerning design information and itsmanagementaswellas19potentialsuccessfactors.Theimportanceofthesuccessfactors was measured on a seven‐point Likert scale with 1 indicating that thedimension was not at all important for the performance, i.e. ‘disagree’, and 7indicating that it was of crucial importance, i.e. ‘strongly agree’. The fiveintermediatevalues representeda sequence including theoption ‘undecided’.Therespondents could also choose the alternative ‘don’t know’. Further, therespondents were asked to rank the fivemost important factors and could stateothersuccessfactorsforasmoothproductionequipmentacquisitionthatwerenotincludedinthesurvey.3.5 ANALYSISOFTHEEMPIRICALDATAThe analysis of the collected data followed the guidelines provided byMiles andHuberman(1994),whodefinedataanalysisasconsistingofthreeconcurrentflowsofactivity: Data reduction, which consists of selecting, focusing, simplifying, abstracting,

andtransformingthedatathatappearinwritten‐upfieldnotesortransactions Datadisplay,which is anorganized, compressed assemblyof information that

permitsconclusiondrawing Conclusiondrawing/verification.Thedataanalysisprocesscanbedividedintosixdifferentstepsandischaracterizedby a continuous interplay between theory and data, see Figure 11. The startingpoint for theresearch(Step1)was thereviewof theacademic literaturerelevantforthepresentthesisresultinginanoverallanalysismodel,whichisbasedonthe


41

assumption that the management of design information consists of the threedimensionsacquiring,sharing,andusingashasbeenarguedinFigure4.

Figure11. The data reduction and analysis process based on a continuous

interactionbetweentheoryanddata(basedonRichtnér,2004).Duringdatacollectiondataweredocumentedandcoded ina casestudydatabase(step2).Inthepresentresearch,datacollectedduringthecasestudieswereintwocategories: (1) case study notes, i.e. expanded notes of problems, challenges, andreflectionsdocumentedinfieldnotesandadiaryassoonaspossibleafterthecasevisit and (2) case study documents, i.e. the data collected in connectionwith theinterviewsandobservationsaswellasinternalcompanydocuments.Thecoding,i.e.thecategorizingofthecollecteddataintokeyvariablesderivedfromthetheoreticalframework, isvital for case research (MilesandHuberman,1994).Asa first step,data were categorized according to the two research questions, required designinformation (design information categories) and management of designinformation.Step three concerns the within‐case analysis, which aimed at becoming familiarwith each case and to identify unique patterns for each case (Eisenhardt, 1989).Therefore,adetaileddescriptionofeachcasewasdeveloped,whichmeans thatatotaloffourcasewrite‐upswerecreated.Thecasewrite‐upsfollowedtheresearchvariables design information categories, acquiring, sharing and using designinformation,andenabledtoputthesamefactsfromdifferentsourcestogether.Astheanalysisofthecollecteddataadvancedforeachcasestudy,recurrentpatternsofthemanagementof design information started to emerge for each case.Basedonthe emergence of these patterns new subcategories had to be added to thecategoriesoriginallyusedduringthecodingofthematerial.Further,theidentifiedpatterns guided the creation ofmatrices about the design information needed indesignactivitiesandhowthisinformationwasmanaged.

RESEARCHMETHODOLOGY

42

Inthenextphaseacross‐caseanalysiswasdone,i.e.generalpatternsacrosscaseswere searched for (step4) in order to gobeyond initial impressions (Eisenhardt,1989). The cross‐case analysis used several techniques such as clustering,comparisons,andnotingrelationsbetweenvariables(MilesandHuberman,1994)andbuiltonthecase‐writeupsofthesingle‐caseanalysis.First,eachpartidentifiedas belonging to an information category or one of the three categories acquiring,sharing, and usingwas coded into the relevant category and then further brokendown into subcategories. All together the four main categories had 14subcategories.Second,thefindingswerematchedagainstthedifferentphasesoftheproduction system design process. Based on these two steps it was possible tocompare the findings across the cases and to identify patterns. It is relevant toemphasizethatthedataofthedifferentcasestudieswerenotsimplyagglomerated,rathertheanalysiswasdoneunderacommoncategoryinordertoensurethateachcase kept its interdependence in the analysis. This is in line with Miles andHuberman (1994),who argue that it is important to understand each case by itsparticulardynamics.Inthefifthstep,theresultsfromtheanalysiswerecomparedwithandrelatedtotheexisting theory, i.e. enfolding the literature. Eisenhardt (1989) points out thatreviewingliteratureinvolvesaskingwhatissimilar,whatisdifferent,andwhy.Thecomparisonwith literature allowed identifying literature that concurredwith theresults of the empirical studies but also identifying literature that was ratherconflicting.Inthesixthandfinalstageconclusionsweredrawn.3.6 JUDGINGTHECREDIBILITYOFTHERESEARCHEventhoughthedifferentmethodsappliedwerechosenconsciously,thereisaneedtoquestionanddiscussthequalityoftheresearchinordertoestablishcredibilityintheresearchresults.Thequalityof thecasestudyresearchcanbe judgedon fourcriteria: construct validity, internal validity, external validity, and reliability (Yin,2009).3.6.1 ConstructvalidityConstruct validity is about establishing operationalmeasures that are reasonablemeasuresfortheconceptstudied.However,sinceresearchersoftenfailtoestablishoperationalmeasuresandcasestudyresearchoftenrestsonsubjectivejudgements,toestablishconstructvalidityisespeciallyproblematicincasestudyresearch(Yin,2009).Theuseofmultiplesourcesofevidence,i.e.datatriangulation,increasesthevalidityoftheresearch(Vossetal.,2002;Yin,2009).Aspreviouslydescribed,inallfour casesmultiple sources of evidencewereused (seeTables2‐5).Although theretrospectivecases(CaseEandCaseT)mainlyusedsemi‐structuredinterviewsasameanstocollectinformationalsocompanydocumentswereused.Further,inCaseEand CaseT, the interviewswere generally conductedwith two respondents, thusminimizingtheriskthatquestionsremainedunanswered.Duringeachcase,morethanonepersonwas interviewedfromdifferent functionsand levels of the organization. Besides analysing the documents provided by thecompanies,businesspressandofficialcompanydocuments(e.g.companywebsites,annualreports,etc.)werealsoreviewedbefore,during,andafterthecasestudies.Thusitwaspossibletogetcomplementaryviewsontheresearchvariables.Further,


43

questions to ask and the choice of suitable respondents were modified duringprogress in Case B and Case P when the understanding of the design processincreased.Toavoidoratleastminimizetheimpactofthepersonalviewoftheresearcher,thetheoretical framework guided the construction of the operational measures.However, problems regarding the applicability during data collection wereencountered. Although constructs seemed to be valid, they were not necessarilyusablebytherespondents.Forexample,informationcanbeprovidedinavarietyofways, such as visual, auditory, kinaesthetic, or tactile (Vincent and Ross, 2001).Many respondents found it challenging to indicate how different types of designinformation were provided. Often different types of information were providedsimultaneously using different means, which made it impossible to constructabsolutescales.However,theuseofmultiple‐casestudieswithdualtypesofcasesprovidedabetteropportunitytoconstructvalidity,asthestabilityofconstructscanbevalidatedbeyondtheimmediatecase(Leonard‐Barton,1990).Inthereal‐timecasestudiesonlynotesweretakenduringtheinterviewsincreasingtheriskofmisinterpretation.However,theprojectdescriptionsofthereal‐timecasestudieswere reviewedby insiders. Further, since the real‐time case studieswereperformedoverseveralmonths,therewasthepossibilitytogobacktotherelevantemployeesoncemoretorechecknotes.Inaddition,theresultswerepresentedanddiscussed,whichincreasedtheunderstanding,andthefeedbackreceivedwasusedtoupdateresults.3.6.2 InternalvalidityInternalvaliditymeanstheprocessofestablishingcausalrelationships(Vossetal.,2002). Internal validity is strengthened by using real‐time case studies as theyenable the researcher to trace the cause‐and‐effect relationship (Leonard‐Barton,1990).InCaseBandCaseP,datawerecollectedfirst‐handandwerenotdependenton participants recalling critical events correctly after the event had happened.Further, the problem of post‐rationalizationwasminimized; the interpretation ofreconstructed events may differ from that made by participants at the time theevent occurred (Voss et al., 2002). Reflecting on the observations as soon aspossibleafterthefieldworkmadeiteasiertorecallandremembercriticalaspectsobserved.Therewasalsothepossibilitytovalidateanddiscardthecausalrelationsfoundduetothereceivedfeedbackonthepropositions.During analysis attention was paid to identifying differences when comparingempirical data with theory, which is in line with Christensen’s (2006)recommendation.Thediscoveryofdifferencesisaprerequisiteforlessambiguityindescriptions and improved theory. Further, the enfolding of literature was animportant step in the analysis and thus increased the validity of the findings(Eisenhardt,1989;Vossetal.,2002).Ithasalsobeendiscussedthatinternalvalidityrelates to the credibility and authenticity of the research results (Miles andHuberman,1994).InterviewingmorethanonerespondentatthesametimeinCaseEandCaseTcouldreducetheinternalvalidityastherespondentsmightinfluenceeach other when answering. However, all case studies are based on datatriangulationandtheresultshavebeendiscussedwithresearchcolleagues.Further,at the beginning of each interview, the perspective chosen, background, and

RESEARCHMETHODOLOGY

44

assumptions were explained to the participants. These activities are in line withmeans suggested by Merriam (1994) and should thus strengthen the internalvalidity.3.6.3 ExternalvalidityExternalvalidityisaboutwhethertheconclusionsdrawnfromthedataandcontextof the research study can be trusted and applied in different contexts as well(Christensen,2006).Thus,externalvalidityisaboutgeneralizationoutsidethecase.Incasestudyresearch,externalvaliditycannotbeestablishedthroughtestsusingdifferentdatasets(statisticalgeneralization);ratherincasestudyresearchexternalvaliditycanonlybeestablishedthroughanalyticalgeneralization(Yin,2009).Retrospective cases strengthen external validity (Leonard‐Barton, 1990) as theyallow for a more controlled case selection according to critical parameters of astudy (Voss et al., 2002). Two retrospective cases were included in the researchpresented in this thesis. Since particular focus was on how to facilitate theacquisition of environmentally sustainable equipment, Case E and Case T wereselected based on their general interest in accomplishing more environmentallysustainableproductionsystems.Inaddition,themanagementofdesigninformationin production system design is not something unique to the cases examined.Althoughtheunitofanalysiswastheindustrializationproject,itcanbearguedthatfew industrializationprojectsarecompletelyunique. Instead, the industrializationprojectmaybeconsideredasarepetitivemeanswithsimilaractivitieseventhoughthe specific production system is unique. Therefore, parallels might be drawn tocomparableorganizationsinsimilarsituations,particularlyasthecasestudieswereselectedbasedonthereplication logicassuggestedbyEisenhardt(1989)andYin(2009). Further, by expanding the research to include equipment suppliersrepresenting a relatively high variety of contexts andbackgroundsbymeans of asurvey, theresearchwasable tooffset someof the lackofexternalvalidityof thecasestudies.3.6.4 ReliabilityReliabilityreferstotheabilitywithwhichanotherresearchercanreplicatethecasestudy and arrive at similar findings and conclusions (Yin, 2009). Miles andHuberman (1994) mention dependability and auditability as related to thereliability of case study research. Thus it is important to describe the researchprocessinnecessarydetail,whichisrequiredinorderforthereadertofollowtheprocess. In the present research, the data collected arewell documented in fieldnotes,diary,anddocumentstudies.Further,allinterviewsinCaseEandCaseThavebeentranscribed.Tominimizetheriskofmemoryfailure,inCaseBandCasePdatawere documented continuously in a diary (e.g. Patel and Davidson, 2003). Thisthesis has an explicitly stated research objective and description of the datacollected and decisions made during the case studies in order to increase thereliability. Further, six papers are appended in the thesis to provide a richdescriptionof thedata collected in the case studies and the survey. In case studyresearch,asmuchdataaspossibleshouldbepresented(Yin,2009).


45

While the previous measures mainly increase the reliability with regard to datacollection, Säfsten (2002) highlights also the need to look separately at the dataanalysisandathowtheanalysismayaffectthereliability.Aspreviouslydescribed,the data analysis followed the three flows of activities suggested by Miles andHuberman(1994)andtheaimwastobeasexplicitaspossibleabouttheanalysisprocessandthestepstakenintheprocessoftheinterplaybetweentheoryanddata.Thusitshouldbepossibleforotherresearcherstorepeattheanalysisofthedata.Finally, theresearchhasbeencarefullyplannedandthemeansusedforcollectingandanalysingthedatahavebeentestedandutilizedinpreviouscasestudies.Casestudy research is not informal or causal, rather it relies on careful planning andrealization (McCutcheon and Meredith, 1993). It should, however, be noted thatdespite all efforts of providing a detailed description and being rigorous in datacollection and analysis, individual judgements and backgrounds affected theresearchprocess.Insum,tostrengthenthevalidityandreliability,differentmethodshavebeenused(see Table 7). It should, however, be pointed out thatmost of the empirical datacollection,dataanalysis,andconclusiondrawinghasbeencarriedoutindividually,implyingthattheresearchprocessisstilltosomeextentablackbox.Undoubtedly,the rich data collected during Case B and Case P facilitated understanding butimpliedalsothelargestchallengeofmaintainingsufficientdistancetotheresearchsubjects. Thus, the need for reflection and discussion with academic supervisorsand colleagues cannot be underestimated. Overall, it is believed that the detaileddescriptionoftheresearchmethodologyincreasedthecredibilityoftheconclusionspresentedinthisthesis.Table7. The appliedmethods for strengthening validity and reliability (based

onOlausson,2009)*Test Methodsused CaseB CaseP CaseE CaseTConstructvalidity

Useofdatatriangulation √ √ √ √Keyconstructsrelying onpreviousstudies (√) (√) √ √Combiningreal‐timewithretrospectivestudies

√ √ √ √Reviewofcasestudydescriptionsbyinsiders

√ √ Internalvalidity

Real‐timestudies √ √ Identificationofuniqueandcross‐casepatterns

√ √ √ √Enfoldingliterature √ √ √ √Useofdatatriangulation √ √ √ √Discussionwithresearchcolleagues √ √ √ √Clearcommunicationofassumptionsmade √ √ √ √

Externalvalidity

Retrospectivecases √ √Replicationlogic (√) √ √ √

Reliability Detailedandcontinuousdatadocumentation

√ √ √ √ Explicitdescriptions √ √ √ √ Systematicworkprocedures √ √ √ √

*Bracketsdenotethatthemethod wasusedtoalimitedextent.

RESEARCHMETHODOLOGY

46

3.7 CONTRIBUTIONTOTHEPAPERSDuringtheprocess,papershavebeenwritten,highlightingdifferentaspectsofthemanaging of design information; they are appended to this thesis. Theprocess ofwriting up the papers was important to be able to prioritize and select relevantdata.Table8clarifiesthesemanticsusedinthethesisandtheattachedpapersandsummarizes thecontributions fromthedifferentauthorsof thepapers.Allpaperswereinitiatedandtheconferencepaperspresentedbytheauthorofthisthesis.Table8. Case study denotations in the thesis and papers and the authors’

contributionstothepapersPaper Cases,

SurveyDenotationinthepapers

First‐authorcontribution Contributionofco‐author(s)

I ‐ ‐ Theoryreview,datacollection,dataanalysisandwriting

Review,qualityassurance

II CaseB CompanyA Theoryreview,datacollection,dataanalysisandwriting

Theoryreview,datacollection,dataanalysis,writing

III CaseB,SurveyS

CaseStudy,Survey

Theoryreview,datacollection,dataanalysisandwriting


IV CaseB CaseStudy Theoryreview,datacollection,dataanalysisandwriting


V CaseB,CaseP

CaseA,CaseB Theoryreview,datacollection,dataanalysisandwriting


VI CaseB,CaseP

CaseA,CaseB Theoryreview,datacollection,dataanalysisandwriting



47

CHAPTER4

:EMPIRICALFINDINGS

CHAPTERINTRODUCTIONIn thischapterof thethesis, theempirical findings fromthefourcasestudiesandthe survey are presented in more detail. Separate case descriptions are given toillustrate the specific situation and preconditionswhen designing the productionsystem,followedbyadescriptionofthemanagementofdesigninformation.

4.1 CASEBThecasestudyreferredtoasCaseBrelatestotheintroductionofanewproductinthe case study company, which had 15 years of product development andproductionexperience.Thecasestudycompanywasaminorplayerinthestudiedproduct segment on a market with rather tough market conditions. The newproductdevelopment(NPD)projectwasinitiatedtostrengthenthemarketpositionandto lead toapositivecash flow in theproductsegment.Therefore, thisprojectwas deemed as strategically important by themanagement and received a greatdealofmanagementattention.TheNPDprojectwas initiated in2008 in cooperationwith a customerandhadatimeframeofthreeyearswithaspecifieddateforstartofproduction.Tosupportthevariousworkactivities, theprojectwasdivided into threeworkpackages: (1)product development, (2) industrialization, and (3) improvements after start ofproduction.CaseBtargetedworkpackagetwo,whichcoveredtheindustrializationfromconceptgeneration to serialproductionand thus included the responsibilityfor the design of the production system and the acquisition of the productionequipment.Thenewproductwashighlycomplex,inlinewithpreviousproductsonthe same platform. However, the product could not be assembled in the existingproduction system,which required thedevelopmentof a newproduction system.Thedifferentstagesofproductandproductiondevelopmentwerecarriedoutinthesamebuilding.TheNPDprojectfollowedastage‐gateprocesswithcriticalgo/nogodecisionpointsandincludedmembersfromdifferentfunctions.Althoughproductionsystemdesignissues were considered in the stage‐gate model, the model was created from aproductperspective.Asaresult, the focusofgravity in themodelwasonproductissues,whilemany important issues of the production systemdesign process felloutside of the used stage‐gate model. A formal process coordinating the work

EMPIRICALFINDINGS

48

activities required in the industrializationproject separately from theNPDmodelwasnotavailable.Ingeneral,theprojectwascarriedoutunderseveretimepressureasthetimeplanforproductdesignwasnotadheredtoandtheresourcesallocatedforworkpackagetwowereinadequateatthebeginning.AtthestartoftheNPDprojectitwasmainlytheproductionengineeringmanagerwhogavevaluableinputtoworkpackageoneandworkedwithaproductionsystemconcept.Theproductionengineermanagerhad experience from several production system design projects and he had alsobeen responsible for the design of the existing production system for the sameproduct platform. From November 2009, i.e. after one and a half years, extraresources were allocated to work with work package two. One of the resourcesprovided was the appointment of a separate industrialization project managerresponsible for the industrialization of the product from concept until serialproduction.The appointed industrializationprojectmanager came fromaneutralfunction,i.e.hewasanexpertinprojectmanagementissuesbutnotinproductionissues.Duetothehighpriorityoftheproject,theindustrializationprojectmanagerreported jointly with the advanced engineering projectmanager (responsible forworkpackageone)directlytothesteeringgrouponaweeklybasis.Anotherpersonassignedtoworkfulltimewiththeindustrializationprojectwastheoneresponsibleforthedesignofthematerialsupplysystem.Forthefirsteightmonthsanexternalconsultantwasassignedthetaskofdesigningthematerialsupplysystem,whileaninternal resource was appointed for the remaining time of the industrializationproject.Since there was little room for concept iterations, the case study companycommissionedone equipment supplier to also design a concept solutionbetweenNovemberandDecember2009.Inparallel,twointernalsolutionshadbeencreatedbased on the earlier ideas of the production engineering manager. The threedifferent concepts were evaluated and synthesized into one final solution, whichfocused on production equipment and material supply aspects. Overall, thegenerationoftheconceptualsolutionwasinfluencedbythefinancialandtemporalscopeoftheproject.Althoughproductionsystemdesignwasnotanewtasktothecasestudycompany,the collection and documentation of design information prior to the studiedindustrialization projectmainly concerned production equipment information, i.e.information that focusedon functions,properties,andcapabilitiesof the technicalsubsystem. Since the production equipment was acquired from an externalequipment supplier, it was necessary to work with technical information ratherthaninformationregardingmaterialsupply,human,andcontrolsubsystems.Thesetypesof informationhadnotbeenspecified indetail inpreviousprojects,while itwasanaturalsteptodocumentinformationconcerningtheproductionequipmentacquisition. For example, technical specifications and guidelines from previousprojects were accessible. Other design information than the one related to theproductionequipmentacquisitionwasusually transferredorallyorwas stored inthe mind of the system designer and was thus based on experience rather thandocumented facts. Thus, the ability to reuse previous experiences that were notrelatedtotheacquisitionoftheproductionequipmenttendedtobelimited.


49

As a holistic perspectivewas emphasized in the studied industrialization project,specific focus was placed on various production system parameters. That meansthat not only design information necessary for the acquisition of the productionequipment from an external equipment supplier was included but also designinformation for internal use such as information about work organization andmaterial supply. In general, information concerning material supply, human, andcontrol systems played a minor role in the concept generation but becameincreasingly important in thedetaileddesignphase.However,byhavingaholisticperspective when designing production systems, the information included in thetechnical requirement specification was modified. Relevant design informationaboutmaterialsupply,human,andcontrolsubsystemswasincluded.Forexample,information regarding workplace design (ergonomic aspects) was included butinformationaboutworkorganizationwasleftout,asthisinformationwasdeemedirrelevant to the equipment supplier. Further, to ensure a better fit between theequipment ordered and future customer demands, design information about themanufacturingandproductstrategy,leanproduction,andproductionperformancewas included. Another aspect was the need for environmental awareness. Onpreviousoccasions,theonlyreferencewasthattheenvironmentalimpactshouldbereducedasmuchaspossible,whichisopentointerpretationanddifficulttoverify.In the studied industrialization project, the information gathered resulted in 17environmental requirements with clear verification measures. The specifiedenvironmental requirements dealt mainly with energy and material concerns aswellastheavoidanceofwastebyimprovingtheproductionprocessitself.To collect relevant information, the existingproduction systemwas analysed andevaluatedtoidentifygoodsolutionstobeincludedinthenewproductionsystem.Inorder to gain an understanding of assembly challenges, the new product wasanalysed through various iterations of assembly and disassembly. In addition,similar products of the main competitors were studied in detail. Further, at thebeginningoftheproductionsystemdesignprocessmucheffortwasalsoplacedonacquiring information about market opportunities, objectives, and strategies.Market distinctionsweremade in order to identifymarket potential, competitiveadvantages on different markets, and market characteristics. Potential futurecustomers and markets were cross‐functionally analysed and documented. Thisanalysisresultedinabetterunderstandingoffuturedemandsfromnewcustomersandmarkets both geographically and in terms of new product areas. The resultswereusedtoidentifyfuturedemandsontheproductionsystem.In order to improve the flow of information among the different functions, theindustrialization project manager gradually introduced two key initiatives:documentation and a cross‐functional industrialization project team. The firstinitiative involved a detailed documentation of relevant and necessary designinformationinordertoimproveaccessibilityofinformationwithinbutalsobeyondtheprojectboundaries.Thedetaileddocumentationalso led tosomestandardsofhow information should be documented. The second initiative related to thecomposition and start of a cross‐functional industrialization team. Relevantfunctions including production engineering, material supply, and qualitymanagementmet in separateprojectmeetings todiscuss thesituationanddecideon future actions. Further, the workshopmanager and operators were invited if

EMPIRICALFINDINGS

50

needed and thus much earlier included in the project compared to previousprojects. The result was that relevant design information was frequently sharedwithintheindustrializationproject.Onemajorimplicationwasthatitwaseasierfortheindustrializationmanagertoprioritizeworkactivities,clarifyexpectations,andadvance the interests and needs of production against the interests of otherfunctions.Basedontheneedtoshare informationwiththeexternalequipmentsupplierandon the tightschedule,attentionwasplacedonminimizingmisunderstandingsanddifferent interpretations of the specified design information. In order to avoidunnecessarydiscussionandconflicts,averificationplanwasdeveloped.Thatis, inthestudied industrializationproject,agreatdealofeffortwasspentonwhenandhow to verify the specified requirements. A comprehensive verification planwasdeveloped, which was adjusted to the request of quotation for the productionequipment.Inordertocontroltheprogressoftheproductionequipmentdesignanddevelopmentaswellas fulfilmentofstatedrequirements, thecasestudycompanydefined different verification occasions. Figure 12 illustrates the structure of theverificationplan.

Figure12. IllustrationoftheverificationplaninCaseB.


51

AsillustratedinFigure12,foreachverificationoccasion,itwasclearlystatedwhatverificationtypeshouldbeused,howeachrequirementshouldbeverified,andthecourse of verification. A total of 22 different types of verification including bothtestinganddocumentationwereidentified.Theidentifiedtypeswerethencoupledtothedifferentrequirementsstatedinthetechnicalrequirementsspecification.Inaddition, the document included a detailed description of underwhat conditionseachrequirementwouldbeapproved.Anumberofrequirementswereverifiedonmorethanoneoccasiontoensureprogressaswellasprovidethepossibilitytoactattheappropriatetime.4.2 CASEPCasePwascarriedoutatabusinessunitwithadifferentproductsegmentfromthatin Case B, which implied different prerequisites for the creation of a productionsystem.Inthestudiedproductsegment,thecasestudycompanywasthetechnologyleader and one of the world’s leading suppliers with a long track record ofsuccessful innovationandNPD.Marketsharewas30to40percentdependingonproductvariantandmarket.The NPD project was very challenging since it was the most complex productdevelopedbythesitesofarandthusalsohadmajorimplicationsontheproductioncomplexity.Astandardproducthadabout15components,while thenewproducthad around 60 components. Further, the product included a new raw material,whichneededtobehandledwithspecialequipment,influencedthebalancingintheassembly process, and required new competencies from the assembly operators.Theproductconceptpresentedin2008receivedpositivefeedbackfromcustomers.Therefore,theproductwasintroducedtothemarketearlierthanoriginallyplannedresulting in a high time pressure.Despite the challenges, theprojectwas seen asstrategically important for themanufacturingcompany,sinceasuccessfulproductintroduction could imply a large market potential. Although the work was notdivided into different work packages, the production engineering manager wasassignedasresponsiblefortheindustrializationofthenewproduct.Hewaslocatedtogetherwiththeproductengineersinoneoffice.Althoughtheproductdesignwasnotcompletedat therequiredgate, thedeadlinefor market launch was unchanged. The fixed date for start of production wasprescribed by the customer, who did not allow any delays. The customer alsoimposed tough requirements regarding the installation of the productionequipmentandthepreproductionseries,whichmadethetimescheduletight.TraditionallytheAdvancedProductQualityPlanning(APQP)7frameworkwasusedtofacilitatenewproductdevelopment.However,duetotheneedtomanagemore7 The APQP is a framework of procedures and techniques used to develop products in industry,particularlytheautomotive industry.The frameworkaimsatensuringacollaborativeproductandproductionsystemdesign, i.e. itspurpose is to facilitatecommunicationwitheveryone involved intheprojectincludinginternalandexternalsuppliers(Thisse,1998).APQPconsistsoffiveconcurrentandcollaborativephases:(1)programplanninganddefinition,(2)productdesignanddevelopmentverification, (3) process design and development verification, (4) product and process validation,and(5)feedback,assessment,andcorrectiveaction.Themethodissupportedbyavarietyoftoolsthatprovidearoadmaptofollow.

EMPIRICALFINDINGS

52

and largerNPDprojects and increased customerexpectations, a stage‐gatemodelwith a series ofmapped‐out activities from inspection to launchwas introduced.Even though the stage‐gate model encompassed APQP elements and wasbenchmarked with the models used by the other business divisions of the casestudy company, the list of activities and linked standards and routines was lesscomprehensive.ThecasestudiedwasoneofthefirstNPDprojectstobemanagedbyusing thenewmodelwhichhadnotbeen testedpreviously.Asa consequence,routinesstillneededtobeestablishedandthedetailsconcerningproductionsystemdesignwereevenlesscomprehensivecomparedwithotherstage‐gatemodelsusedatthecasestudycompany.Because theproductionengineeringmanagerwasnew inhis job, the instructionsgivenintheNPDstage‐gatemodelwerenotsufficientenoughtosupporthisworkactivities.Inordertosupporthiminhisworkactivitiesandhelphimbecomemoreeffective in future production system development projects, the managementrecognizedtheneed tocreateaproductionsystemdevelopmentmodel thatcouldbe integratedwith the NPD projectmodel. The developedmodel with a detailedemphasisontheproductionsystemdesignactivitiesissummarizedinFigure13.Toensure a high level of recognition, the production system development modelconsisted also of stages and gates. For each activity listed below, a detaileddiscussionof thecontentof theactivitywasreportedinaseparate,moredetaileddocument. The detailed description gave an overview of the required inputinformation for each activity, and each stage incorporated the sharing of newinformationbetweenmultiplefunctions.

MANA

GEME

NTOF

DESIG

NINFOR

MATIO

NINT

HEPR

ODUCTIO

NSYSTEMDESIG

NPRO

CESS

53

Figure13.D

etailedlistofwo

rkact

ivitiesth

atsho

uldbe

accom

plishe

dintheproduct

ionsyste

mdesignp

rocess

.

EMPIRICALFINDINGS

54

In order to acquire relevant information, the production engineering managerfrequently met and discussed critical issues with key persons such as the globalengineeringdirector, the operations director and the businessareamanager. Thebusiness areamanagerwas tobe responsible for theproductionafterproductionramp‐upandprovidedvaluable feedback fromtheoperationsperspectiveearly intheindustrializationproject.Theoperationsdirectorgaveinputregardingobjectiveand strategy, while the global engineering director was an important source forinformation about the production equipment. In addition, the productionengineeringmanagerwaspartoftheNPDteam.Eventhoughtheprojectwaschallengingandrathercomplexaccordingtotheteammembers, itwasnot theonlyproject forwhich theproductionengineermanagerwas responsible. Concurrently with the studied project, the case study companyworkedwithfiveother“equallyimportant”projectswhichallhadthesamedateforthestartofproduction.Theseprojectsalsorequiredthecreationofnewproductionsystems but were less complex as similar products are manufactured at the sitetoday. In addition, the production engineering manager was also responsible foroperationalproductionengineeringissues.Theemployeesworkingwiththedesignoftheproductionsystemweredependentontheproductinformationbutalsoondirectivesfromthemanagementwithregardto financial restrictions and strategy. However, the uncertainty concerning finalcustomerdesignandfuturemarketsmadethecommunicationofdirectivesdifficult.For example, thedifferences in theproductdesign8were so extensive that itwasdifficult to judge whether the products for two different customers could beassembledon the sameproduction system.As a result, itwasdifficult toplan forand design an appropriate production system as the decision about one or twoproductionsystemshadmajor implications for thesolution.Anotherdecisionthatcontainedimportantinformationfortheproductionsystemdesignbutthathadnotyet been finalizedwas themake‐or‐buy decision, i.e.what components should bemadein‐houseandwhatshouldbeacquiredfromsuppliers.Thecomplexityoftheproductanditsrelatedproductionchallengesaswellastheabsence of previous knowledgemade equipment suppliers and the consultant animportantsourceofinformationforcreatingaconceptualsolution.Theconsultantalso supported the production engineering manager in his work of creating aconceptual system solution. Further, through the contact with the equipmentsupplier, there was the possibility to study production systems of othermanufacturing companies that had met the challenge of balancing with theparticularrawmaterial.Assembly of prototypes gave valuable product information. Further, product andproduction system designers were placed in the same office, which facilitatedcommunication between the people involved. One of themajor challenges duringconcept generation was that the product information, i.e. information about thefinal design of the product that was going to be manufactured, was highlyambiguous. The customer was also a source of information with regard to the8 In the studied product segment, there is no generic product design; instead each application issignificantlyadjustedtofitthedemandsofthecustomer’sproduct.


55

production ramp‐up, i.e. the customer determined, for example,whenproductionequipmenthadtobeinplaceorhowtheproductionequipmentshouldbetested.In order to keep the deadline, much effort was devoted to finding technologicalsolutionstotheproblem.Asnosimilarproducthadbeenmanufacturedpreviouslyby the division, therewas no possibility to study the existing production system.Overall, the focus was on the technical subsystem, i.e. the production equipmentrequired. The case study company worked intensively with the request forquotation,i.e.whatinformationshouldbeincludedandhowtheinformationshouldbe structured in order to improveunderstandability for the equipment suppliers.For example, the content of the request forquotationwasbenchmarkedwith therequestforquotationofothermanufacturingcompaniesintheautomotiveindustry.Even though it was agreed to take a more holistic approach, information aboutmaterial supply, human, and control subsystemswas not prioritized early in theconcept generation phase and had only been documented to a limited extent inearlierindustrializationprojects.Themajorityofinformationrelatedtothesethreesubsystemswas transferred duringmeetings without any links to standards androutines.Further,onemajorchallengewastogetemployeesinvolvedintheprojectas things were not going smoothly and numerous problems had to be solved.Nevertheless,therewereseveralfactorsindicatingthattheinitiativestakenbytheemployeesresultedinamoreholisticapproach.Forexample,theearlyinclusionofthe business area manager in the project made it more explicit to also considerinformation that concerned operational production issues. In addition, focus wasalso placed on developing support for the production engineering manager thatincludednotonlyproductionequipmentissuesbutallaspectsaffectingthecreationoftheproductionsystem.4.3 CASEEOverall,theproductionequipmentacquisitionwascarriedoutinaprojectfollowinga standardized process including a comprehensive technical requirementspecificationaspartoftherequestforquotations.Thecasestudycompanydidnothave a long tradition of major production equipment acquisition projects, butduring the previous five years almost the entire production equipment had beenreplaced.At thebeginning, the acquisitionof production equipmentdidnotworksatisfactorily.Forexample,thedialoguewindowoftheproductionequipmenthadadifferentappearancealthough ithad the samecontrol system,which implied thatthe operator had to learn the handling of the dialoguewindow each time he/sheusedanewtypeofequipment.Asaresult,majoradjustmentsandmodificationsinthetechnicalrequirementspecificationweremade.Oneofthemajorproblemswasnotthatthecontentofthetechnicalrequirementspecificationwaswrongbutthatdirectiveswereunclearanddifficulttounderstand.Thechangesmadeledtoclearerdirectives, which ensured better consistency in the equipment. The requirementspecification was deemed as a useful document but it was also highlighted thatusually not all requirements stated can be fulfilled in the acquisition projects. Inaddition to the technical requirement specification, project‐specific requirements,company‐specific documents and information about the product are usuallytransferred.

EMPIRICALFINDINGS

56

Today, the technical requirement specification is updated everyoneor twoyearsdependingon theavailable time.Themodifications followastandardprocessandare often based on feedback received from equipment suppliers or experiencegainedwiththeexistingequipment.Theequipmentsuppliers’feedbackisseenasavaluable input, as it is aligned to changes in the context. Before changes in thetechnicalrequirementspecificationareimplemented,revisionmeetingstakeplacein which suggestions are discussed, and it is mutually agreed whether thesuggestions should be included or not. Themajor benefits of a standardized anddetailedtechnicalrequirementspecificationcamefromtheincreasedconsistencyintheproductionequipment,whichminimizedtheneedforsparepartsandreducedthe support required with regard to, for example, the software. The technicalrequirement specification included eight environmentally sustainablerequirements.Therewereseveralobstacles toamoreholisticperspectiveof thespecificationofthe requirements included in the technical requirement specification, i.e.havingamore comprehensive technical requirement specification. First, the functionresponsibleforthetechnicalrequirementspecificationwasproductionengineering,which had its competency in issues related to the production equipment, i.e. thetechnicalsubsystem.Second,thecontentofthetechnicalrequirementspecificationwas based on experience from operations and previous acquisition projects.However, production engineering had limited insight into problems of the othersubsystems. Another factor that affected the specification of other requirementswas the absence of clear objectives related to the environment. For example,environmental sustainability has been a buzzword for a long time but itwas notimplemented in theoperationalorganization.Environmental sustainability tendedto be in the periphery and not in the focus of the business. Finally, although thecompanyhadadetailedandsophisticated technicalrequirement specification, thefinal production equipment depended also on the background and interest of theperson responsible for the acquisition of the production equipment. This impliedthat if theperson in chargeof the acquisitionof theproduction equipmenthad ageneralinterestinmaterialsupply,therewasamuchstrongerfocusontheseissueswhenacquiringtheproductionequipment.4.4 CASETThecompanyinCaseThadalongtraditionofacquiringproductionequipmentandhad worked intensively on a process to be used when acquiring productionequipment;theacquisitionwascarriedoutintheformofaproject.Theprocesshadlinks to standards and routines, of which the technical requirement specificationwas one critical document. Other information transferred related among otherthingstothespecificproject,theproductstobemanufactured,andpolicies.Therequirementspecificationisadjustedonanannualbasis.Updatesaremadebyexperts in the specificarea.Forexample, themaintenance function is responsibleforthespecificationofthemaintenancerequirementsthatshouldbeincludedinthetechnical requirement specification. Changes in the technical requirementspecificationwerederivedfromexperiencesinoperation,butalsofromtheinputofthe equipment suppliers. The general parts of the requirement specificationwerewritten by senior staff, i.e. employees who had detailed knowledge about


57

productionequipmentandtheacquisitionofproductionequipment.Theability toensure that all requirements were fulfilled and that the process of acquiringproduction equipment was less dependent on individuals was mentioned as themajoradvantageofthestandardizedtechnicalrequirementspecification.Thus,thecompany also allocated the responsibility for the acquisition of the productionequipment to employees with limited knowledge and yet a high degree ofconsistencyintheirproductionequipment.Ontheotherhand,thedetailedlevelofspecifyingrequirementsmadeitdifficultforalessexperiencedandknowledgeableperson to follow up all requirements. In addition, the high level of detail causedextracosts,asthecompany’sstandardswerenotnecessarilysimilartothoseoftheequipmentsupplier.Ingeneral,environmentallysustainablerequirementswerenotlistedasaseparatecategorybutincludedinthesections,wherejudgedasrelevant.Itwasarguedthatacategoryonlyrelatedtoenvironmentallysustainablerequirementswouldrequireachangeintheworkwiththeequipmentsuppliers.Equipmentsuppliersdidnothaveenvironmentally sustainable experts employed who could work exclusively withthese requirements; the equipment supplier rather concentrated on findingtechnological solutions to the specified demands. In addition, it was emphasizedthatitwasimportanttoconsiderfirsthowtheproductionprocessactuallyworkedandthenspecifyhowthevariousproductionprocessesinfluencedtheenvironment.Although a wide variety of areas were considered, three causes that made thespecification of other requirements a challenging task were identified. First,requirementsneededtobespecifiedonsuchagenericlevelthattheywerevalidtoall acquisition projects. Second, the employees working with the requirementspecificationwerenoexpertsinotherareasthatmightaffecttheproductionsystemsuchasmaterialsupplyissues,whichalsomadeitproblematictounderstandtheirimpact.Third,notallaspectshadthesamepriorityastechnologicalissues.Thiscanbeillustratedbytheexampleofenvironmentalrequirements.Itwasconcludedthattechnological problems limited themanufacturing company in its efforts tomakemore products, to produce according to drawings, etc., while environmentallysustainablerequirementswerenotalwaysseenasanobstacletoachievinghigherlevelsofperformanceandthusnotasdetailedasthetechnicalrequirements.4.5 SURVEYSWith regard to the design information requiredby equipment suppliers it canbesaid that various categories of design information are applied by equipmentsuppliers, see Table 9. The equipment suppliers quoted the requirementspecification as themost important document for theirwork activities,while theverificationplanplayedonlyaminorrolefortheworkoftheequipmentsuppliers.Furthermore, the design information exchanged between the equipment supplierand the buyer, i.e. the manufacturing company, was classified into hard designinformation and soft design information. Overall, equipment suppliers tended torely more on hard design information (median 70 per cent) than soft designinformation(median30percent),seeFigure14.Itshouldbeemphasized,though,thatall equipment suppliers stated that theyuseda combinationofhardandsoftdesign information and did not solely rely on one kind of design information.Examples of typical harddesign informationwere cycle time, production volume,

EMPIRICALFINDINGS

58

price, and capacity requirements, while soft design information related toergonomics, safety, and maintainability. However, the equipment suppliers alsostressed that performance information, such as tact time, production volume, orquality requirements stated by the manufacturing company could be furtherenhanced.Table9. Categoriesofinformationusedbytheequipmentsuppliersandranked

byrelevancetothesuppliersInformationcategories Quoted(%)Technical information – information about general technical requirements andproject‐specifictechnicalrequirements

96%Product information – information about the products that are going to bemanufacturedinthenewproductionequipment

80%

Strategicinformation–informationaboutneedsandwants 64%Projectinformation–informationabouttiming,scope,termsofpurchase,etc. 56%Contextinformation–informationaboutthespecificbackgroundandcontextoftheactualproductionequipmentacquisitionproject

48%

Financialinformation–informationabouttheinvestmentbudget 28%*Verification information – information about how and when to verify statedrequirements

8%

*Note:Financialinformationwasnotaproposedcategory ofinformationinthequestionnairebutwas quoted by seven equipment suppliers as a relevant category of information. Thus, therelevance might have been different if financial information had been included as a proposedalternative.

0

1

2

3

4

5

6

7

8

9

10 % 20 % 30 % 40 % 50 % 60 % 70 % 80 % 90 % 100 %

Num

ber o

f eq

uipm

ent

supp

liers

Figure14. The proportion of hard design information used by equipment

suppliersinrelationtototalinformation.Inaddition, thedata fromthesurvey indicated thatequipmentsuppliersobtainedthemajority of the requireddesign information from their customers. In total 21outof25equipmentsuppliersstatedthat50percentormoreoftheneededdesigninformation was provided by the customer. For the whole sample, however, themedianwas50percent,which indicates that thedesign informationprovidedbythe manufacturing company was incomplete. At the time a lack of designinformationwasidentifiedbytheemployeesoftheequipmentsupplier,theyhadto


59

actontheirowninitiativetoobtaintherequireddesigninformation.Inthesecasesequipmentsuppliersgotaccesstodesigninformationbygettingincontactwiththecustomer, preferably an appointed contact person. Usually telephone or e‐mailcommunicationwasusedtoobtainthemissingdesigninformation,butsometimestheequipmentsuppliermadeastudyvisit.Anotheraspectofobtaininginformationisthatthepreferredcommunicationmediadependsonthetypeofinformationtobeshared.The results showed thatdocumentswereusually thepreferredchoice forobtaininginformation,butinformationabouttheobjectivewaspreferablyobtainedbypersonalinteraction.Further, the differences in the perception of relevant and necessary designinformation between the manufacturing company and the equipment supplierbecameclearwhentheequipmentsupplierhadtoexplicitlyindicatetheproportionof relevance of the perceived design information, i.e. the percentage of thetransferred design information needed by equipment suppliers in their workactivities.Basedontheanswersoftheequipmentsuppliersitwasidentifiedthatthemajority of the transferred design information was needed by the equipmentsuppliers(median80percent).However,onlyfourequipmentsuppliersstatedthattheprovideddesigninformationwas100percentneeded.The results also highlighted the need to appoint a skilled contact person. Theequipment suppliers stated that the manufacturing companies should carefullyselect a skilled contact person for effective collaboration.The contact personwasrankedasthemost importantfactorthatcontributedtoaneffectiveacquisitionofproduction equipment, seeTable 10. The possibility to openly discuss alternativeoptions with the customer was considered very important by the equipmentsuppliers.Table10summarizesthosefactorsthatwereidentifiedbytheequipmentsuppliers as being themost important to achieve an effective production systemacquisition process. Overall, it can be concluded that formal coordinationmechanismsweremoreimportantthaninformalcoordinationmechanisms.Table10. Thefourmostimportantfactorscontributingtoaneffectiveproduction

equipmentacquisitionprocessasidentifiedby25equipmentsuppliersinthesurvey

Factor

No.ofrespondentsthatrankedfactoras

importanttoalargeextent

oneofthefivemostimportant

factors*

ofcrucialimportance

mostimportantfactor*

Appointedcontactperson

9 18 13 8

Clear/formalhand‐over

9 8 9 3

Continuouscommunication

16 17 6 5

Commonobjective

16 15 5 5

*Note:Theequipmentsupplierswereaskedtochooseandrankthefivemostimportantfactorsoutof19statedfactors.Therespondentsrankedthemostimportantfactoras1,thesecondmostimportantfactoras2,etc.

EMPIRICALFINDINGS

60

Tosummarize,inthischapterpartsofthematerialgatheredduringthecasestudyandthesurveyhavebeenbrieflydescribed.Thedescriptionprovidesanoverviewof the case‐specific prerequisites and how design informationwasmanaged. Thematerial presented in this chapter will be used in the analysis. Discussing theanalysisoftheempiricalfindingsisthesubjectofthenextchapter.


61

CHAPTER5

:CONTENTANDSTRUCTUREOF

DESIGNINFORMATION

CHAPTERINTRODUCTIONThepresentchapteranalysestheempiricalfindingsstartingwithacategorizingofthe design information followed by an analysis of the management of designinformation. In order to clarify the relationship between the management ofinformation and the process of designing the production system, testablepropositionsareformulated.

5.1 CATEGORIZINGDESIGNINFORMATIONIn the following, the design information applied when designing a productionsystem will be categorized. Based on the empirical findings and the frame ofreference, ten categories of design information to be used when designing aproduction system are identified. The different categories of design informationidentifiedarebasedondifferencesintheinformationcontent.The first and most common category of information is production systeminformation,whichregards informationabout theelementsandsubsystemsof theproduction system. Based on Groover’s (2008) classification of the productionsysteminto foursubsystems(seeChapter2), theempirical findings indicatedthatinformation regarding the technical subsystem had the highest priority whendesigningtheproductionsystem.Inaddition, thefindingsfromCaseBandCasePindicated that information about the other three subsystems was only of minorrelevanceintheearlyphasesbutbecameincreasinglyimportantinthefinalstages.The production system information was either developed internally from therepresentativesofthefunctions9that influencedthedesignoftheirsubsystemsorexternallyfromconsultantsandequipmentsuppliers.InCaseBandCaseP, thedesignof theproductionsystempartlyoverlappedwiththat of the product to be manufactured. Previous research has pointed out thatconcurrentdevelopmentprojectsofproductspresuppose interactionbetween thedesign activities. As a consequence, it has been argued that the manufacturingfunction needs to receive sufficient information about the product design,9Asdescribed inChapter2 in this thesis the term function isused todenote responsibilities andworkareasrequiredtodesigntheproductionsystem.

CONTENTANDSTRUCTUREOFDESIGNINFORMATION

62

particularlywhen thedesignof theproductand theproduction systemshouldbecarriedoutinparallel(GerwinandBarrowman,2002;HaandPorteus,1995).Theempirical findings showed that product information, i.e. all information about theproducts to be manufactured, was also an important category of informationrequired for the design of the production system. Amore detailed description ofproduct information is provided inPaper II. Theproduct informationwasmainlydeveloped by the representatives of the design engineering and verificationlaboratoryfunctions.InCasePknowledgeofhowthenewmaterialshouldbehandledinoperationwaslimited. Therefore, opportunitieswereused to studyproduction systemsof othermanufacturing companies that had solved similar problems. Studying otherproduction systems is a type of benchmarking, resulting in a reference point(SandkullandJohansson,1996)andcanbeasourceofinspiration(Säfsten,2002).InCaseEandCaseTitwasalsopointedoutthatgoodrelationswiththeequipmentsupplierprovidedinsightsintohowcompetitorshavesolvedsimilarproblems.Asaconsequence, competitor information is identified as an important category ofdesign informationwhendesigning theproductionsystem. It is important tonotethat competitor information in this context refers to the study of suitableproductionsystemsolutionsoutsidethecompanyratherthantoinformationaboutcompetitors’products.The empirical studies revealed that thedesignof theproduction systemwasalsoaffectedbythemanufacturingstrategyandleanproductionphilosophy.InCaseTithasbeenemphasized that theproduction system solutionhas tobebasedon theproductionphilosophytofacilitateoperationalexcellence.Beforethedesignoftheproduction system started in Case B and Case P, it was agreed on what leanimplementation level had to be accomplished. Company I had different levels ofimplementing leanproduction,whereeach levelplaceddifferentrequirementson,forexample,keyperformanceindicators,overallequipmenteffectiveness,operatormaintenance, etc. Because of the different possibilities to design a productionsystem,informationabouttheoverallobjectiveshastobeclearlydefined.Asstatedbyoneequipmentsupplierinthesurvey,theabsenceofacommonobjectiveattheircustomerscouldleadtomisunderstandingsandmiscommunicationwhendesigningtheproductionsystem.Onlywithinformationabouttheobjectivesisitpossibletoachieve a fit between the expectedproduction systemand theproduction systemsolution (Wu, 1994). Therefore, strategic information can be identified as aninformationcategoryrequiredwhendesigningtheproductionsystem.Anothercategory identified iscustomer information,whichrefersto theparticulardemands placed by a customer. The signing of a customer in Case B and Case Pmade it necessary to clearly specify the demands of the final customer and toconsider them in the design of the production system. When designing theproduction system, customer information related among other things to totalmarket demand, product packaging, and test specifications and thus did not dealwith the functionality of the designed product. Having information about specificcustomer characteristics has also been expressed as important by Ruffini (1999).This information categorymight bemore uncertainwhen no customer has beensignedtotheproject.


63

The design of the production systems was closely aligned to the marketrequirementsinCaseBandCaseP.Duringtheinitialphasestherewereanumberofdiscussionstodevelopanunderstandingoftheproductmarketandthecustomersinordertounderstandtheconsequencesfortheproductionsystemsolution.Thisisinlinewithearlierresearchintheproductionsystemdesignfieldwhicharguesthatthemarket strategy should be consideredwhen designing the production system(Bennett andForrester,1993;Ruffini,1999).Becauseof thedifferentdemandsofdifferent markets, it is necessary to includemarket informationwhen designingproductionsystems.Theinformationinthiscategorycamefromtherepresentativeof the market function or from outside the company and provided data aboutcurrentandpotentialcustomersandmarkets.In all four case studies, the design of the production system was influenced bylegislation, which can differ from country to country. Work organization andenvironmentalsustainabilityissuesaretypicalexamples.InCaseB,thepatentofacompetitor also influenced the production system solution and disregarding itwouldhavehadsevereconsequencessuchaspenaltiesorrework.Theinformationthat determines the legal freedom of decisions is categorized as regulatoryinformation.Itincludesrules,regulations,patents,etc.andismadeavailablebytheauthorities.The prerequisites for carrying out the design of the production system differedbetweenCaseBandCaseP.However,despitethedifferencesitwasobviousthattheprocess required somekindof guidance. For instance, inorder tomake sure thateveryone was updated and that the design of the production system wascoordinated, project meetings were continuously initiated and a timeline wasdeveloped.Meetingswerenotonlyusedtokeepmembersinformed,buttheyalsoofferedtheproject leaderapossibilitytodeterminewhethertherewasaneedformembers to support each other and take on other activities. The value ofinformation thatcontributes toastructuredprocesshasalsobeenpointedoutbythe equipment suppliers in the survey. The need to have information about themanagement and control of the project is crucial to the management ofmanufacturingprojects(O'Sullivan,1994).Theinformationthatguidestheprocessof designing the production system is categorized as project managementinformation,anditwasmainlytheresponsibilityoftheprojectmanagertoprovidethis information. However, also engineering design is an important sources forprojectcentricinformation(PaperII).The design information identified in the empirical studies and categorized asfinancialinformationdefinedthebudgetandcostsandwasprovidedbythesteeringcommittee. Accordingly, the financial information determined the scope forinvestmentsandresourceallocationinbothCaseBandCasePandinvestmentsinproduction equipment in Case E and Case T. Several equipment suppliers in thequestionnairepointedoutthat itwouldbegenerallybeneficialtoreceivefinancialinformation from their customers to define the scope of the project. To avoidgreater risk of costly rework, equipment suppliers wanted to have informationabout the budget to be able to present a quote in linewith expectations of theircustomers. The importance of financial considerations is also confirmed by theresearchpresentedbyBellgran(1998).


64

Tofollowtheworkprogressattheequipmentsupplierandtohavethepossibilitytocheck if all requirements specified in the request for quotationwere fulfilledwasseenasanimportantaspectinCaseB.Therefore,theemployeesputgreatemphasisonthedevelopmentofaverificationplan,whichinvolvedaprocessforverifyingthespecified requirements at different points in time and by different means. TocontinuouslyfollowupthestatusofeachindividualrequirementspecifiedisakeyissueinadevelopmentprocessasdiscussedbyAndersson(2003).Thus,verificationinformation about how and when to verify specified requirements is therefore acategory that needs to be included in the design of the production system. Theemployees who specified the requirements were responsible for the verificationinformation.Although the content of the design informationwill be dependent on the specificcontext, the ten categories of design informationdescribed above (seeFigure15)should be viewed as a guiding principle around which the content of designinformation should be individually discussed. The identification of the designinformationcategoriescreatedafoundationforwhatdesigninformationneededtobehandledwhendesigning theproductionsystem,whichwillbediscussed in thenextsection.

Figure15 Tenidentifiedcategoriesofdesigninformationrequiredforthedesign

oftheproductionsystem.5.2 STRUCTUREFORHANDLINGDESIGNINFORMATIONIn linewithprior research (Frishammar,2005;FrishammarandYlinenpää,2007;Ottum and Moore, 1997) discussed in Chapter 2, the management of designinformationcanbedividedintothethreedimensionsacquiring,sharing,andusingdesigninformation.Dividingthemanagementofdesigninformationintothethreedimensions facilitates structuring and analysis of the management of designinformation(seealsoPaperV).


65

5.2.1 AcquiringofdesigninformationInformationtypeAnexaminationofthetenidentifieddesigninformationcategoriesrevealsthatthecontentofthedesigninformationhadtocoverfouraspectsimportantforthedesignoftheproductionsystem. First, it was necessary to acquire design information about all elements and

subsystems of the production system in order to minimize the risk ofsuboptimization,whichbuildsontheideaofapplyingaholisticperspective(e.g.Bennett,1986;Groover,2008).

Second,ithasbeenarguedthattheexternalcontextneedstobeanalysedwhendesigningtheproductionsystemasnoteverythingofimportanceoccurswithinthemanufacturingcompany(BennettandForrester,1993;Ruffini,1999).Thus,there isaneedtoacquiredesign information that isdevelopedexternal to themanufacturingcompany.

Third,when designing the production system, it is also necessary to considerinternal requirements, i.e. information that was developed internal by themanufacturing company but external to the industrialization project. Forexample, the design of the production system has to be in line with themanufacturing strategy and the production philosophy (Cochran et al.,2001/2002;Duda,2000).

Fourth, information about project organization and progress needed to beacquiredinordertoprovidesufficientconstraintsandcontrolmechanisms.Thisis in linewitharguingthatthesuccessofaprojectdependsontheplanningoftheprocess(Bellgran,1998).

Basedontheempiricalfindings,thefollowingconclusioncouldbedrawn:Acquiringabroadvarietyofdesigninformationfacilitatestheprocessofdesigningtheproductionsystem.

Accordingtotheory,informationcanalsobedividedintohardandsoftinformationand any kind of information is almost always a combination of soft and hardinformation (Frishammar, 2003; Häckner, 1988). During the design of theproductionsystem,bothhardandsoftinformationwasusedtoaccomplishthetaskathand.Nevertheless, theempirical findingshighlighted thatagreateramountofhardinformationwasacquiredwhendesigningtheproductionsystem.Onepossiblereason for the reliance on hard information is the possibility to easily processgreater amounts of information (Häckner, 1988). Another explanation is that theprocessing of hard information is not tied to an individual person (Shrivastava,1985).However, one shouldnotunderestimate thevalueof soft information. Softinformationwasneededtoplacehardinformationinacontextandtoimprovetheunderstanding of how to apply the acquired hard information. Based on theempiricalfindings,thefollowingconclusioncouldbedrawn:Acquiringacombinationofsoftandharddesigninformationfacilitatestheprocessofdesigningtheproductionsystem,butthelargerpartshouldbehardinformation.


66

SourceofinformationToacquireinformationrelatedtoeachidentifieddesigninformationcategorymadeitnecessarytoacquiredesigninformationfrombothinternalandexternalsourceswhendesigningtheproductionsystem.ThisissimilartothefindingspresentedbyZahayetal.(2004),whopointoutthatinformationiseitherdevelopedinternallyororiginatesfromoutsidethecompany.Internallydevelopeddesigninformationwasacquired from employees of the company ranging from operators to top‐levelmanagement.Typically,externalsourcesrelevantfortheproductionsystemdesignprojectinthecasesstudiedwereequipmentsuppliers,consultants,customers,andregulatory bodies. Table 11 classifies the ten identified categories of designinformation (see Figure 15) according to where the content of the designinformationwasdeveloped.Table11. The ten design information categories classified according to theiroriginInternallydevelopeddesigninformation

Internallyandexternallydevelopeddesigninformation

Externally developeddesigninformation

1.Product 6.Market 8.Regulatory2.Strategic 7.Productionsystem: 9.Competitor3.Projectmanagement • Technical 10.Customer4.Financial • Materialsupply5.Verification • Human • ControlThe empirical findings showed a pattern where internal sources of designinformation seemed to be preferred over external ones, i.e. whenever possible,design information was acquired from an internal source. One explanation forfavouring internal sources is the perceived accessibility of sources, which ispositivelyrelatedtothefrequencyofusage(Sawyerretal.,2000).Theabsenceofaclear strategy of how to acquire information from external sourceswas found inCase B and Case P. In general, the design information that was acquired fromexternalsourceswasusedtogetamorecompleteunderstanding,tosolveproblemsinanappropriatewayandto limit thescopeofpossibleactions.For instance, thedesign of the technical subsystem benefited from acquiring additional designinformation from equipment suppliers who were experts in this area. Thus, theacquiringofdesigninformationfromexternalsourcesprovidedavaluableinputtothedesignoftheproductionsystem.Basedontheempiricalfindings,thefollowingconclusioncouldbedrawn:Acquiring design information from internal and external sources facilitates theprocess of designing the production system, but internal sources are the preferredchoice.

RecallingthediscussioninChapter2,sourcescanalsobedividedintopersonalandimpersonalsources.FromtheanalysisofCaseBandCasePitisclearthatpersonalsources, i.e. “other people”, were the preferred source for the acquiring ofinformation.Continualmeetingsprovidedanefficientpossibilitytoacquirevaluableinformation from other project members. In addition, the empirical findings onimpersonal sources reveal that the process, i.e. the operationalized productionsystem, was hardly applied as a source. Seeing production systems that were in


67

operation as a sourcewas only appliedwhen the production system of the sameproductplatformhadbeenstudiedandanalysedinCaseB.Documents,ontheotherhand, were considered as an important source of information, particularly whendesigningthetechnicalsubsystemoftheproductionsystem.ThefindingsaboutpreferringpersonalsourcesisinlinewiththefindingsofDaftetal. (1988), who emphasize that personal sources may be needed to allow forenactment and clarification of ambiguous situations. The process of designing aproduction system is often associated with a high degree of uncertainty withrepresentatives from different functions differing in training and background.Anotherexplanationfortheuseofpersonalratherthanimpersonalsourcesmaybethe limited documents available. Based on the empirical findings, the followingconclusioncouldbedrawn:Acquiring design information from personal and impersonal sources facilitates theprocess of designing the production system, but personal sources are the preferredchoice.

The process of designing the production system evolved over a considerableamountof timeandconsistedofseveraldistinctphasesasdiscussed inChapter2and highlighted in Chapter 4. Although the phases are partly overlapping, eachphaserequirestheaccomplishmentofdifferentworkactivities,seeFigure13,andthus the required information varied between the different phases of theproduction system design process. Based on the findings in Case B and Case P,Figure 16 summarizes the categories of design information thatwere required ineach phase of the design process. The analysis was based on the phases andactivitiesthatweredescribedinCasePinChapter4.In the concept study phase in Figure 16 there was a need to attain almost allcategoriesofdesigninformation,whileinthescopingphaseonlytwocategoriesofdesigninformationwereattainedaccordingtotheempiricalfindings.Inaddition,inthefirsttwophasesmucheffortwasplacedonacquiringinformationthatensuredalignmentwiththeexternalcontext.Inthelaterthreephases,ontheotherhand,theacquireddesign informationdealt toa largerextentwithaccomplishingasuitabledesignofthefoursubsystems,i.e.thetechnical,materialsupply,human,andcontrolsystems of the production system. The findings from Case B and Case P alsoindicatedthatsoftinformationwasmorerelevantearlyintheprocessofdesigningthe production system, and design information of a hard character becamemoreimportant in the later phases. Based on the empirical findings, the followingconclusioncouldbedrawn:Acquiringdifferentdesigninformationcategoriesatdifferentpointsintimefacilitatestheprocessofdesigningtheproductionsystem.Ingeneral,softinformationshouldbeemphasizedintheearlydesignphases,whilehardinformationbecomesmorerelevantinthelaterdesignphases.


68

Figure16 Categoriesofdesigninformationacquiredinthedifferentphasesofthe

production system design process. The different grey scales indicatetheoriginoftheinformation.

5.2.2 SharingofdesigninformationCommunicationmediumThe results of the empirical studies showed that in order to accomplish theinterdependent work activities, design information was shared among therepresentativesfromthevariousfunctionsinvolvedinthedesignoftheproductionsystem.Thesharingofdesigninformationtookplacebyeitheralessrichmediumsuch as documents or a richmedium such as face‐to‐face interaction. During theprogressoftheindustrializationprojectsinCaseBandCaseP,continualmeetingsindifferentconstellationstookplace.Inadditiontotheweeklymeetingswiththenewproduct development team, separate meetings concerning the design of theproduction system took place on a frequent basis. The cross‐functional meetingsfocusedoncriticalissuesthatneededtobeaddressedandusuallyresultedinactionlists thatwere followedupat thenextmeeting.Among theprojectmemberswhowereinvolvedinthedesignoftheproductionsystem,thesharingofinformationbyface‐to‐faceinteractionwasacknowledgedasinformativeandparticularlysuitablefor solving unclear issues in both Case B and Case P. This is in line with otherresearchadvocating that that a richmedium is appropriate forresolving complexissuessinceitallowsfor immediatefeedbackandenactment(e.g.DaftandLengel,1986;Frishammaretal.,2011).Although face‐to‐face contacts and meetings were the dominant strategy for thesharing of information in the cases studied, documents were also used to sharerelevant information.The resultsof the survey revealed thatdocumentswere thepreferredwayofobtaininginformationfromtheircustomers.However,aspointed


69

out in Case E documents are a suitablemediumonlywhen information has beencarefullydocumented,i.e.thereisaneedtodevotegreatcaretothestructureandcontentofdocumentstominimizetheriskthatinformationwillbemisunderstoodorinterpretedincorrectly.OnethingworthnotingisthatinCaseBandCasePdocumentsweremostsuitablewhenacommonperspectiveandterminologyhadbeenestablished.Forinstance,inorder to share the design informationwith the equipment supplier, it wasmoreimportanttohavedevicesinplacethatfacilitatedthesharingofasufficientamountoftherequireddesigninformationthantomeetfrequentlytodiscusscriticalissues.Once a common understanding and perspective has been established among thefunctionsinvolved,documentscanbeusedtocoordinatethetaskathand(DaftandLengel,1984).Thepeopleinvolvedinthedesignofthetechnicalsubsystemhadthesamebackgroundandtrainingastherepresentativesoftheequipmentsuppliersinthecasesstudied.However,oneshouldrememberthattherewasagenerallackofdocumentationinearlierprojects,whichmakesitdifficulttoshareinformationbymeans of documents. Therefore, it would be beneficial to improve the overalldocumentation in order to base the choice of the communicationmediumon theinformationrichnessandnotontheonlymediumavailable.Basedontheempiricalfindings,thefollowingconclusioncouldbedrawn:Sharingdesigninformationthroughdocumentsandmeetingsfacilitatestheprocessofdesigning the production system butmeetings should be emphasized in ambiguoussituations, while documents are more suitable for the sharing of well understoodmessages.

FormalizationAchieving a holistic view required the sharing of information between severalfunctionsinthemanufacturingcompanyandwithexternalpartnersinCaseBandCaseP.Figure17summarizestheflowofinformationbetweentheindustrializationproject manager and production engineering manager with the internal andexternal functions involved in thedesignof theproductionsystem inCaseB. It isimportanttonotethatthefigureisschematicandincludesonlyanindicationoftheinformation flow on a general level. However, it highlights the informationdependencyamongthefunctionsinvolvedintheproductionsystemdesignproject.That is, if design information was not adequately shared among the functionsinvolved, it caused difficulties in carrying out the task at hand without avoidingsuboptimization.Challengesandconsequencesofthesharingofdesigninformationbetween representatives from design engineering and the representatives fromproductionengineeringarediscussedindetailinPaperII.ComparingtheindustrializationprojectinCaseBwiththatinCasePindicatesthatthesharingofdesigninformationbetweenspecializedfunctionswasfacilitatedbyorganizing the design of the production system in a separate industrializationproject includingaproject teamandadesignatedprojectmanager.This is in linewiththereasoningpresentedbyLove(1996)arguingthataprocessofdesigningaproduction system is organizationally complex requiring a project team withmembersfromdifferentspecializedfunctions.Projectteamsfacilitatethesharingofinformationacrossfunctionalboundaries(LawrenceandLorsch,1967).Inaddition,it has been argued that a successful project requires a knowledgeable project


70

managerwhocandevotesufficient time toplan,manage,andmonitor theproject(Mabert et al., 1992). In Case B the project manager worked full time with thestudied industrialization project focusing on structuring the work and keepingdeadlines.Thisincludedamongotherthingstoensurethattherequiredinformationwas shared. Based on the empirical findings, the following conclusion could bedrawn:Sharingdesigninformationinadedicated,cross‐functionalproductionsystemdesignteamfacilitatestheprocessofdesigningtheproductionsystem.

Figure 17. Design information received from and sent to the industrialization

projectmanagerandproductionengineeringmanagerwhendesigningthepreproductionsysteminCaseB.

AsindicatedinChapter2,toeffectivelyintegratecross‐functionalactivitiesrequiresthesharingof information.The findings inCaseBandCasePdisplayed,however,differences in how design information was shared with the external equipmentsupplier compared to the sharing of design information among the internalfunctionsofthemanufacturingcompany,(seealsoPaperIV).Sharing of design informationwith the external equipment supplier. Based on theempirical findings from the four case studies, a process for acquiring production


71

equipment couldbe identified (seeFigure18). In eachphase sharing informationbetween the equipment supplier and their customers was required. First, therequest forquotationwassent toanumberofpotentialequipmentsuppliers.Theselectionofpotential supplierswasbasedonpreviousexperience, timelineof theproject,complexity,etc.Inthesecondphasetheequipmentsupplierswereinvitedtopresenttheirquotesatthemanufacturingcompany.Thereafter,atechnicalandcommercial evaluation of the quotes was made at the manufacturing companybeforeoneequipmentsupplierwasselectedforfinalnegotiation.Finallyacontractwascreatedbasedonthetechnicalsolutionandthefinancialdetails.Overall,itcanbeconcludedthattheequipmentsupplierselectionfollowsaformalizedprocess.

Figure18. Process for selecting a production equipment supplier based on the

fourcasestudies.Tobeable to integrate theactivities carriedoutby theequipmentsupplier,plansandscheduleswereestablished includinga timeplan fromsendingtherequestofquotationtotheinstallationoftheequipmentandstart‐upofproduction.Thetimeplan provided a guideline regarding the distribution of design information thatneeded to be shared in the different phases. Further, the use of standardizeddocumentsensuredthatpreviousexperienceswerereusedandthattheinformationwas easy to understand for the equipment supplier. In Case E and Case T, thetechnical requirements specification was updated on a frequent basis to reflectchangesintheexternalcontextorlessonslearnedincompletedacquisitionprojects.Sharingofdesigninformationamongspecializedfunctions.ThestandardthatguidedthedesignoftheproductionsysteminCaseBandCasePwasthestage‐gatemodelused in new product development. However, the applied stage‐gate model wascreatedfromaproductperspectiveandthusdidnotincludedetailedstrategiesforthe design of the production system. For instance, the available schedule lackedsomecrucialdatesrelevantwhendesigningtheproductionsystemsuchaswhentoorder the production equipment. In addition, the production system designactivities that were included in the product development stage‐gate model wereassigned to individual functions. As a result, each function handled the activitiesseparatelyasfaraspossibleevenifitwouldhavebeenusefultoshareinformationwithother functionsworkingwith thedesignof theproduction system.Basedonthe findings in Case B and Case P, it can be concluded that the sharing of designinformation among specialized functions within themanufacturing companywas


72

less formalized than the sharing of design information with the equipmentsuppliers. As a result, the empirical findings suggest that formalization facilitatesthe sharing of information among functions. The findings are in accordancewithpreviousstudies,whichemphasizethatanintangibleandnon‐standardizedcontentof the design process hampers the sharing of information (Moenaert and Souder,1990),whileastage‐gateprocesswithcriticalgo/no‐godecisionsatvariouspointsprovidesprocedures for improved informationsharing (CooperandKleinschmidt,1991; Griffin and Hauser, 1996). Based on the empirical findings, the followingconclusioncouldbedrawn:Sharingdesign information in a formalized processwith different stagesandgatesfacilitatestheprocessofdesigningtheproductionsystem.

ThesharingofdesigninformationinCaseBandCasePwasalsoaffectedbymoreinformal coordination mechanisms that promoted the transfer of designinformation among specialized functions. When the specialized functions werelocated close to each other, information was directly shared when unanticipatedproblems and challenges arose during the design of the production system. Thisreducedtheriskthatdesigninformationwasdelayed,lost,oraltered.Furthermore,more spontaneous sharingofdesign information tookplacewhenemployeesmeteachother in thehallsoraround the lunchandcoffeebreaks.On theotherhand,longer distances between the functions involved reduced the frequency ofspontaneous information sharing and made face‐to‐face communicationinconvenient.Thisobservation is in linewith thatofAllen(1977),whoconcludedthat the probability of interaction between people is rapidly reduced with thephysicaldistancebetweentheirworklocations.Projectmembersoftheproductionsystemdesignprojectneedtofindwaystoreducethephysicaldistancebetweenthespecialized functions to facilitate understanding each other (Allen, 1970;VandeveldeandVanDierdonck,2003).Inaddition,informalpersonalcontactswithotheremployeesnotdirectlyinvolvedin the design of the production system were established in Case B. Theindustrialization project leader contacted other employeeswho he thought couldcontributeto theproblemsolving.These formlessrelationssupportedthesharingofadditionalinformation.Theimportanceofformlessrelationshasbeendiscussedby, among others, Griffin and Hauser (1996), who emphasize that a formlessrelationincreasescoordinationintheprocessanddecreasesprojectuncertainties.Basedontheempiricalfindings,thefollowingconclusioncouldbedrawn:Sharingdesigninformationthroughinformalcoordinationmechanismsincludingco‐locationofspecialized functionsandthebuildingof formlessrelations facilitatestheprocessofdesigningtheproductionsystem.

5.2.3 UsingdesigninformationIn Paper I it is argued that the use of design information was influenced by theinformationqualityandwhetherinformationwaspragmaticornot.Informationquality Four information quality levels are distinguished: relevant information, soundinformation, optimized process, and reliable infrastructure, where each level is


73

determined by four criteria (see Chapter 2 for a further description). In thefollowing,imperfectionsinthefourinformationqualitylevelswillbeanalysedanddiscussedindetail.Relevant information refers to whether information is comprehensive, accurate,clear, and easily applicable (Eppler, 2006). In Case B and Case P therewere fewdocuments available supporting the design of the production system at thebeginning of the industrialization projects. When the members in theindustrializationprojectsstudiedrealizedthatadditional informationwasneeded,theywerenotalwayssureofthedetailofinformationrequired.SimilarresultswerefoundinCaseE,wheretheabsenceofdocumentationinthetechnicalrequirementsspecificationresultedinunwanteddifferentsolutionsbetweenproductionsystemswhentheystarted toreplace theexistingproductionsystemfiveyearsago. In theprojects studied in Case B and Case P, efforts were made to improve thedocumentation for the currentproject, but also for generaldocumentation,whichcan be reused in future projects. For instance, in Case B, the process of makingproduct design changeswas standardized and formalized by the industrializationproject manager in order to ensure that the information could be applied andunderstoodbythevariousfunctions.However, alsodocumented information sometimes causedproblemsregarding itsrelevance. The findings from the survey revealed that the technical requirementsspecification often contained too detailed information, while at the same timeproject‐specific information dealingwith the context, background, and scopewasmissing. It was argued by the suppliers that too detailed requirementsspecifications were unnecessary at an early stage. Overall, the empirical findingssuggestthatthereweredifficultiesinjudgingtherelevancyoftheinformation.Soundinformationreferstowhetherinformationisconcise,consistent,correct,andcurrent (Eppler, 2006). During Case B and Case P, information was presenteddifferently depending on the competence and experience of the author of theinformation. It could be found that some informationwas extensively presented,such as the information concerning the technical subsystem, while informationabout the human subsystem was hardly even labelled. Further, as most of theinformation came from one personal source, there was often no possibility tovalidatethereceivedinformationandthustherewasariskthattheinformationwasbiasedandcontainederrors.OneofthemajorchallengesidentifiedinCaseBandCasePwastheneedtooverlapthedesignof theproduction systemwith theproductdesign.Theproductdesignproject did not follow the mandatory timeline, and information about the finaldesign of the product that was supposed to be manufactured was delayedconsiderably. Consequently,much of thework of designing the actual productionsystemwasbasedonpreliminaryinformationthatchangedduringtheprogressofthenewproductdevelopmentproject.Theuseofpreliminaryinformationhadtwoconsequences. First, theworkwith the design of the production systemwas alsodelayed as the employees waited for more complete information. Second, somerework was required when product design changes were made. This is alsodiscussedbyTerwieschetal.(2002).Theempiricalfindingssuggestthatthereweredeficienciesregardingthesoundinformationqualitylevel.


74

Optimizedprocessreferstowhethertheprocessbywhichtheinformationiscreatedanddistributedisconvenientandwhetheritisprovidedinatimely,traceable,andinteractivemanner(Eppler,2006).Theactualwayofhandlingdesign informationin the industrialization projects studied did not facilitate obtaining the requireddesign information. The search for additional design information was time‐consumingasitwasnotalwaysclearwheretofindtherequiredinformation.Often,there was a need to contact several persons who were assumed to have therequireddesigninformation.Further,itwasdifficultaftersometimehadpassedtotrace the information,as informationwasoften transferredorally.Anotheraspectthatclearly influencedtheoptimizedprocess levelwas the latereleaseofproductinformation in the industrialization projects studied. As the product informationbecame available late, the ability of influencing the product design was reduced,whichlimitedtheavailableoptionsforthedesignoftheproductionsystem.Reliable infrastructure refers to theprocess throughwhich information isactuallyprovided and considers accessibility, security,maintainability, and speed (Eppler,2006). The empirical findings from Case B and Case P and the survey revealedissueswiththeaccessibilityoftherequireddesigninformationwhendesigningtheproductionsystem.Although informationwasavailableonthe intranet, itwasnotalwayseasy toknowwhere to find it,or informationwaskept ina localdatabasethat was not accessible to other functions. Another factor that hindered theaccessibilitywasthefactthatnotalldesigninformationwasavailablewhenneeded,suchaslatedecisionsontheproductdesign.Asthedesigninformationconcerningtheproductionsystemwasnothandledinanykind of formalized information system, personal contacts were often used toretrieve the required information. However, there was no guarantee that thecontactedpersonpossessedtheinformationandevenifhe/shehadrelevantdesigninformation, the information was often incomplete, which required involvingadditional people in the cases studied. Thus, retrieving design information wassometimes found frustrating and time‐consuming by the project team memberswhen designing the production system. In addition, the limited documentationreducedtheabilitytomaintaintheinformationovertimeandthecontentofdesigninformationwasnot restricted,which increased the risk forunwantedchanges intheinformationcontent.Accordingtotheempiricalfindings,itcanbeassumedthatthereweredeficienciesregardingthereliableinfrastructurelevel.Asummaryoftheanalysisoftheempiricalfindingsindicatesthattheacquiringandsharingofdesigninformationoftencausedsevereproblemswiththequalityoftheinformation,andthisaffectedtheutilizationofdesigninformationwhendesigningtheproductionsystem.Thereby,theempiricalresultssupportpriorresearchstatingthat information with few deficiencies in quality is likely to be used to a higherextent than informationwithmajor imperfections in informationquality(O'Reilly,1982).Furtherithasbeenpointedoutthatcomplementingapproacheshavetobeappliedtodealwithshortcomingsininformationquality(Eppler,2006;Johansson,2009).Basedontheempiricalfindings,thefollowingconclusioncouldbedrawn:Using design information with high information quality facilitates the process ofdesigning the production system, but different complementing approaches arerequiredtoachievehighinformationquality.


75

PragmaticInformationAnother aspect that clearly influenced the use of information was the degree ofnewnessofitscontent.Informationneedstobeacombinationofconfirmationandnovelty, i.e.pragmatic inorder tobeuseful to the receiver (Fjällström,2007;VonWeizsäcker, 1974). Comparing the industrializationproject in CasePwith that inCaseB suggests that thedegreeof noveltyhas tobe carefully balanced to ensurethattheinformationisusable.InCaseP,theemployeeswereforcedtotakeadvicefromexperts(bothconsultantsandequipmentsuppliers)tounderstandthecontentand implications of the provided information as the amount of novelty of theinformation was overwhelming. On the other hand, in Case B, the design of theproductionsystemwaslesstroublesomeastheemployeescouldrelateinformationto experiences gained from the assembling of the previous product generation.Anotherexampleshowingthe importanceofcombiningconfirmationwithnoveltyistheeffortsofspecifyingenvironmentallysustainablerequirementsinCaseEandCase T. Although the importance of considering environmentally sustainablerequirements in the technical requirements specificationwasnotdisputed, itwasdifficult to actually specify this kind of requirements. Specifying environmentalrequirementswasnotemphasizedearlier,whichmadeitdifficulttounderstandtheprovided information and transfer it into concrete requirements. Based on theempiricalfindings,thefollowingconclusioncouldbedrawn:Usingpragmaticdesigninformationfacilitatestheprocessofdesigningtheproductionsystem.

Tosummarize,theanalysisoftheempiricalfindingsshowsthatthemanagementofdesign information is influenced by a number of circumstances and it may bedifficult to understand how the circumstances affect the management of designinformation.Therefore, inChapter6theanalysisistakenonestepfurtherandthefindingsare synthesized.Asdiscussed inChapter3 the researchhasbeencarriedout within the field of applied research and thus should also have industrialusefulness.This implies, to someextent, thatChapter6 ismoreprescriptive thanthe thesis has been so far, suggesting a framework that should contribute to aneffectivemanagementofdesigninformation.


77

CHAPTER6

:TOWARDSEFFECTIVE

MANAGEMENTOFDESIGNINFORMATION

CHAPTERINTRODUCTIONThischapteranswersthetworesearchquestionsandsynthesizesthemintoadesigninformation management framework. The framework is a suggestion for how aneffective management of design information can be achieved when designing theproductionsystem.

6.1 REQUIREDDESIGNINFORMATIONWHENDESIGNINGPRODUCTIONSYSTEMS

Research question 1 addresses the design information required when designingproduction systems. The research in this thesis identifies and describes tencategories of such information (see Section 5.1). The identified design informationcategoriescoverabroadvarietyofaspectsconcerningtheproductionsysteminitselfbutalsothewaythedesignactivitiesshouldbecarriedout.Thedesigninformationcan be further divided into information that avoids suboptimization of individualelements and subsystems of the production system, and design information thatensuresthattheproductionsystemisconsistentwiththedemandsfromtheinternalandexternalcontext.ThisissimilartowhatMiller(1992)andRuffini(1999)callaninternalandexternal fit.Althoughtheresearchdoesnotdisputethereasoningthatthere is not one production system design representing the best fit (Bozarth andMcDermott,1998;VandeVenandDrazin,1985),thefindingsshowthatthedesigninformationneeds to supporta conceptual solution that is consistentwith internalandexternaldemands.Thus,thedesigninformationcategoriescanbeclassifiedintofourmaingroups:1. Design information that minimizes the risk of suboptimization of individual

elementsandsubsystemsoftheproductionsystem.2. Design information that contributes to the alignment of the production system

solutionwiththerequirementsplacedbytheexternalcontext.3. Design information that contributes to the alignment of the production system

solutionwiththerequirementsplacedbytheinternalcontext.

TOWARDSEFFECTIVEMANAGEMENTOFDESIGNINFORMATION

78

4. Designinformationthatfacilitatesadvancementsinthedesignwork.The first group, i.e. design information that ensures that suboptimization ofindividualelementsandsubsystemsoftheproductionsystemisavoided,minimizesthe riskof slackboth in termsofwasteand in termsof resources. Inaproductionsystem where design information about all four subsystems, technical, materialsupply, human, and control systems, isnot integrated, the interdependenceamongthesubsystemsisdisregarded,whichincreasestheriskthatthesubsystemsdonotsupport each other in operation. In theworst case, the created production systemcouldimplythatthefoursubsystemsareirreconcilable.Theneedtoincludedesigninformation about all four subsystems can be traced back to the need to have aholisticperspectivewhendesigningtheproductionsystems(Bennett,1986;Groover,2008). Taking design information about all four subsystems into account shouldensure a groundingofdesigndecisions in the context of thedemandsof theothersubsystems.The second group is design information that ensures that the production systemsolution is aligned with the requirements placed by the external context. Not alldesign information required is developed within the manufacturing company.Factorsexternaltothemanufacturingcompanyshouldhaveanimpactonthescopeof decisions when designing the production but can also provide opportunities.Models emphasizing the external context such as, for example, that presented byBennett and Forrester (1993), acknowledge the value of analysing the externalcontextwhendesigningtheproductionsystem.Takingdesigninformationaboutthecontext of the manufacturing company into account should contribute to aproduction system that is consistent with demands placed by customers,competitors,regulatorybodies,etc.The third group is design information that ensures that the production systemsolution is aligned with the requirements placed by the internal context. Thisinformation facilitates that the production system solution provides the capacitiesandcapabilitiesrequiredtobesuccessfulinthemarkets.Forexample,byincludingdesign information about the manufacturing strategy in the production systemdesignprocess,theperformanceoftheproductionsystemshouldalsobeco‐alignedwith the corporate strategy (e.g. Hayes and Wheelwright, 1984; Skinner, 1969).Takingdesigninformationabouttheinternalcontextoftheindustrializationprojectinto account should facilitate that the production system solution satisfies theobjectivesofthemanufacturingcompany.The fourth group, i.e. the design information that facilitates advancements in thedesignworkisbasedontheneedfortransparencyandguidanceforthetaskathand.This is in congruence with what Petersson and Petersson (1992) call operationalinformation.Thepeopleinvolvedinthedesignoftheproductionsystemneedtobesupportedintheirworkactivitiesbyinformationthatensuresasmoothprogressofthedesignwork.Itisimportanttoplanthedesignprocessandtohaveastructuredandsystematicworkingprocess (Bellgran,1998).Takingdesign informationabouttheproductionsystemdesignprocessintoaccountshouldpromoteaneffectiveandefficientdesignprocess.


79

6.2 THECHARACTERISTICSOFDESIGNINFORMATIONMANAGEMENTWHENDESIGNINGPRODUCTIONSYSTEMS

The second research question addresses the characteristics of themanagement ofdesign information when designing production systems. A characteristic can bereferred to as a distinguishing trait10 meaning that the management of designinformationwillbedependentontheparticularsituation.Thus,themanagementofdesigninformationhascertaincharacteristics.As discussed in Chapter 2, design information is not incorporated in the designprocess,because the information isnotacquired, sharedorused.Therefore, in theproduct development literature it has been suggested that the management ofinformation has to be considered as a multidimensional concept consisting of thethreedimensionsacquiring, sharing,andusingofdesign information(Frishammar,2005; Frishammar and Ylinenpää, 2007; Ottum and Moore, 1997). The presentresearchshowsthatthisconceptisalsovalidwhendesigningtheproductionsystem.Inaddition,thefindingsofthecurrentresearchsuggestthatacquiring,sharing,andusingofdesigninformationarehighlyinterrelatedwitheachdimensionaffectingtheother two, i.e. themanagementofdesign information is an iterativeprocesswhichshouldbeviewedasaloopingofacquiring,sharing,andusingofdesigninformation,seeFigure19.

Figure19. Themanagementofdesigninformationshouldbeviewedasaloopingof

acquiring, sharing, and using of design information with a highdependencybetweenthethreedimensions.

As a result, the acquiring, sharing, and using of design information has to beperformed continuously throughout the production system design process andshould not be considered as a singular event. Figure 20 below illustrates that themanagementofdesigninformationconsistsofnumerousloopsofacquiring,sharing,andusingdesigninformationineachphaseoftheproductionsystemdesignprocess.

10www.m‐w.com,”characteristic”,accessed12January2012.


80

Figure20. The management of design information is a continuous looping of

acquiring,sharing,andusingofdesigninformationthroughallphasesoftheproductionsystemdesignprocess.

Further,theanalysisconductedinChapter5revealedthatthemanagementofdesigninformationwas influencedby several characteristics (see alsoPaperV).Basedonthe empirical findings six characteristics could be identified namely: informationtype, source of information, communication medium, formalization, informationquality and pragmatic information. In Chapter 5, each characteristic is discussedunder the informationmanagementdimension it influences.Figure21 illustratesasummaryofthesecharacteristics.

Figure21. Overview of the identified characteristics affecting themanagement of

designinformation.In the following text a more detailed description of the meaning of eachcharacteristicisprovided.


81

The first characteristic is the information type that needs to be acquired whendesigning the production system. Facilitating the acquiring of a broad variety ofdesign information is of crucial importance, as the generated production systemsolutionshouldrelyonacomprehensiveviewgoingbeyondthatofemphasizingthetechnicalsubsystem.Theimportanceofhavingaholisticperspectiveandensuringafit to the internal andexternal contexthasbeendiscussedpreviously (e.g.Bennettand Forrester, 1993; Ruffini, 1999). Further, it is important to acquire also designinformationabout theprojectorganization,whichgivesguidanceonadvancing theproject within a predefined scope. However, one should be aware that having acomprehensive view of the design of the production system implies that atremendous amount of design information needs to be acquired. This involves theriskofinformationoverload,whileatthesametimetheacquiredinformationneedstobeunderstood.Therefore,itisimportanttoacquireacombinationofhardandsoftdesign information, of which hard information facilitates the handling of largeamounts of design information and soft information provides the contextualdescription.Thesecondcharacteristicregardsthesourceoftherequireddesigninformation.Theempirical findings suggest that the acquiring of the required design informationshould not be limited to the own organization. Design information should beacquired from both internal and external sources. Although a combination ofinformation sources were applied in the empirical studies, internal and personalsourceswere by far the preferred choice for acquiringdesign information, as theywereeasilyaccessible.However, external sourcesmayhelp tohandle the trade‐offbetween limited resources and the need to minimize uncertainty (Frishammar,2003).Thus,thereisaneedforaclearstrategythatsupportstheacquiringofdesigninformationalsofromexternalsources.The third characteristic refers to the communicationmedium appliedwhen designinformation isshared.Sharingdesign informationthroughpersonal interactionhasseveral positive consequences such as the ability to process rich information,interpretunclearissues,andallowforenactmentandclarification(DaftandLengel,1986).Thereare,however,risksinvolvedinrelyingheavilyonpersonalinteractionfor the sharing of design information. Since not all project members are alwaysinvolved indirect interaction, it is challenging toensure that information is sharedwith all functions that would benefit from the information. Another risk is thatinformationisnotcomprehensivelydocumented,whichreducestheabilitytoreuseexperiencesinfutureprojects.Basedonthereasoningabove,itisarguedthatthereis a need to ensure that all relevant functions can acquaint themselves with therelevantdesign information.Onestrategywouldbe to support thedevelopmentofstandard documents. These documents are valuable for the sharing of well‐understooddesigninformationthatdoesnotneedanyfurtherclarification.Thefourthcharacteristicisbasedontheformalizationoftheprocessofdesigningtheproduction system. The degree of formalization seems to have a key role for thesharingofdesigninformation(PaperIV).Themoreformalizedtheprocesswas,themore structuredwas the sharing of design information. Therefore, it is suggestedthat itwouldbebeneficial to createaproduction systemdesignprocess similar tothe stage‐gate model used in new product development with dedicated resourcesand a clear process ownership structure. Performing the work activities in each


82

phaserequiresthesharingofdesigninformationbothinsideandoutsidetheprojectteam.ThisissupportedbyCooperandKleinschmidt(1991),whofoundthatastage‐gate process led to more information exchange and multifunctional discussions.Further, by dedicating resources to the process, the employees could committhemselvestotheworkofdesigningtheproductionsystem.Processownershipwascrucial for coordinating the sharing of design information among the functionsinvolved. However, it is important to also acknowledge the value of informalcoordinationmechanismssuchasco‐locationandformlessrelations.Thesharingofdesign information can be improved by formal as well as informal coordinationmechanisms(FrishammarandYlinenpää,2007).Informalcoordinationmechanismsare particularly valuablewhen employeeswith different backgrounds and trainingneedtoshareinformation.Thefifthcharacteristicidentifiedisinformationquality,whichinfluenceswhetherthedesign information is used or not when designing the production system. Majorimperfectionsininformationqualityhavenumerousnegativeconsequencessuchasconfusion, distraction, and delays. On the other hand, high information qualityfacilitates that thedesign information isactuallyused in the taskathandand thusprovidescomprehensivejustificationsforeachdecision.Hence,effortsareneededtoaccomplishhigh information contentquality (relevant and sound information) andhighinformationmediaquality(optimizedprocessandreliablestructure).However,as each imperfection has different causes, several but complementary approachesarerequiredtoobtainhighinformationquality(Eppler,2006).The finalandsixthcharacteristicrefersto theneedfor informationcontenttobeacombinationofconfirmationandnovelty, i.e.pragmatic information. Studies in thisresearch(CaseBandCaseP) togetherwithFjällström’s(2007) researchshowthatpragmatic information is importantwhen using information since the informationhastobeunderstandable,whileatthesametimeitshouldbedifferentfrompreviousknowledge.Theneedtohandleabroadvarietyofdesign information increasestherisk that the novelty parts dominate the information content, which may preventunderstandingofthedesigninformation.Asaresult,aprojectteamresponsibleforthe design of the production system should include representatives from differentfunctionswithdifferentbackgroundsandtraininginordertoensurethatthedesigninformation can be understood and thus utilized in the design of the productionsystem.6.3 DESIGNINFORMATIONMANAGEMENTFRAMEWORKThe objective of the research presented in this thesis is to develop knowledge tocontribute to an effective management of design information when designingproduction systems. The analysis of the empirical findings reveal that themanagement of design information is complex and requires different initiatives indifferent phases of the production system design process. As a result, it may bedifficult for practitioners to assimilate and utilize the developed knowledge.Therefore,thissubsectionsoutlinesadesigninformationmanagementframeworktosummarizeandvisualizetheknowledgegained.Thedevelopedframeworkprovidesastructurethatshouldcontributetolong‐termknowledgeontheissue.TheframeworkisanelaborationoftheresultfromResearchQuestion2.Itisbuiltonthethreedimensionsofmanagingdesigninformation,i.e.theacquiring,sharing,and


83

usingofdesign information (Frishammar,2005;FrishammarandYlinenpää,2007)and thesixcharacteristicsbrieflydescribedabove.Thus, theframeworkcomprisesthreeparts: The first part – acquiring of design information – comprises the type of design

informationrequiredandthesourceofthedesigninformation. Thesecondpart–sharingofdesigninformation–comprisesthecommunication

mediumappliedandthedegreeofformalizationinthedesignprocess. Thethirdpart–usingofdesigninformation–comprisesinformationqualityand

pragmaticinformation.InChapter5and inSection6.2, a setof critical factorsweredescribedundereachcharacteristic, such as design information categories required or whether designinformation shouldbe acquired from internal or external sources. InFigure22, anoverview of the identified critical factors relevant to consider for an effectivemanagementofdesigninformationispresented.Thereisalwaysaneedtodealwiththesefactorswhendesigningtheproductionsystem.However,eachfactorneedstobeadjustedtothespecificsetting.Forexample,whetherdesigninformationshouldbe sharedbymeansofdocumentsordirect interactiondependsamongstotheronthe background and training of the team members. The overview should beconsidered as a structure around which the factors of the management of designinformationcanbediscussedwhendesigningproductionsystems.

Figure22. Overview of the key factors influencing the management of design

informationwhendesigningproductionsystems.


84

The purpose of developing the design information management framework is tocreateavaluable tool that canbeapplied toeffectivelymanagedesign informationwhendesigningtheproductionsystem.Thechallengeformanufacturingcompaniesis to acknowledge that the design of production systems has a number of phases,eachwithdifferentactivities thatneed tobeperformed.Thevariousactivities thatneedtobecarriedoutmakeitimpossibletomanagedesigninformationinthesameway throughout theentiredesignprocess.By separating thevariousactivities intodifferentphases,aspresentedinFigure13,itwaspossibletopointoutapproachestobe used for an effectivemanagement of design information. The structure for thedesigninformationmanagementframeworkevolved,seeFigure23.Theframeworkincorporates the factors relevant to reflect upon for an effective management ofdesign information and provide guidelines of how these factors should beconsidered. The design information management framework should be used byproject managers and project members involved in the design of the productionsystem.Based on the findings of the present study the framework highlights that thedemands on the management of design information at the beginning of theproduction systemdesignprocessweredifferent from those in the laterphasesoftheprocess.Atthebeginningitisimportanttoreduceanyequivocalitysurroundingtheprocessofdesigningtheproductionsystem,whichpoints towardstheneedforpersonalcommitment,astheemphasisshouldbeonpersonalinteractionasoftenaspossiblefacilitatingextensivediscussions.Asequivocalityisreduced,itispossibletorely increasingly on alternatives that are independent of personal interaction. Theframeworkalsoillustratesthatsomefactorsareofrelevanceinseveralorallphasesof the production system design process. One should also note that each workactivityidentifiedinFigure13offersmoredetailedguidanceontherequireddesigninformationineachphase.

MANA

GEME

NTOF

DESIG

NINFOR

MATIO

NINT

HEPR

ODUCTIO

NSYSTEMDESIG

NPRO

CESS

85

Figure23

Thed

esigninfor

mationm

anage

mentframe

worktobeusedwh

ende

signin

gprod

uction

syste

ms.


87

CHAPTER7

:DISCUSSIONANDCONCLUSIONS

CHAPTERINTRODUCTIONInthefinalchapterof thepresentthesisthefocus isondiscussingtheresultsandthe research methodology and also to outline the scientific and industrialcontributionsaswellastosuggestdirectionsforfurtherresearch.Thediscussionofthe results promotes ideas and should be seen as a startingpoint to advance thefindingsofthepresentresearch.Thechapterendswithconcludingremarks.

7.1 GENERALDISCUSSIONTheanalysis identified ten categoriesof information that shouldbe appliedwhendesigning the production system. Overall, the identified categories should be ofgeneral value for the design of production systems as their importance has alsobeen discussed in previous literature. However, it is important to note that thedesign information categories that need to be handled by the manufacturingcompanyshouldbeamatteroftheapproachtakentothedesignoftheproductionsystem.AsdescribedinChapter2,itispossibletodifferentiatebetweenaconcept‐generating approach, a concept‐driven approach, and a supplier‐driven approach(Säfsten, 2002). Case B and Case P can be categorized by a concept‐generatingapproach, which implies that the industrialization projectwas responsible for alldesignsteps.Asaresult,alldesigninformationcategoriesshouldbehandledbythemanufacturingcompany.Thismightbedifferent in, forexample,asupplier‐drivenapproach, where the equipment supplier is responsible for creating a detaileddesignsolution.In general, the management of design information when designing productionsystemswas challenging.Oneexplanation fordifficulties is related to theneed toapplyaholisticperspectiveintheconceptgenerationoftheproductionsystem.Theempirical findings indicate thatmuch effortwas placed on acquiring informationabout the technical subsystem, while acquisition of design information regardingthe human,material supply, and control subsystemswas limited. The employeesinvolvedintheproductionsystemdesignprocessdidnotdisputethesignificanceofacquiring design information about all subsystems of the production system butfounditdifficulttogatherconcretefactsthatgobeyondlawsandprovisions.Several reasons may lie behind the discrepancy concerning the amount ofinformationacquired.First,thetechnicalinformationistheunderlyinginformationused for the creation of production equipment, which is often acquired from anexternal equipment supplier. External equipment suppliers need information to

DISCUSSIONANDCONCLUSIONS

88

undertaketheirworkactivitiesofdesigningandbuildingtheproductionequipmentatagivenpointintime.Withoutinformation,equipmentsupplierscannotfulfiltheobjective,andtheiraccesstoadditionalinformationcanonlybesatisfiedbygettingin contact with people employed at the manufacturing company. Second, theproduction equipment represents a large part of the production systemdevelopmentprojectcosts,andthecostof failing isevident,whilethecost fortheother subsystems might not be as apparent. Another argument is that theproductionsystemisoftenimplementedinanalreadyexistingplantwithavailablestructures regarding the material supply, human, and control subsystems.Therefore,thedesigninformationthatneedstobeacquiredconcerningthesethreesubsystems relates to operations, which may conceal the benefits of acquiringinformation about the subsystems as they were already predefined from thebeginning.Themanagement of design information is also affected if an external equipmentsupplier is responsible for the design and subsequent building of the productionequipment.Equipmentsuppliersaresourcesofmajorinnovationsinmanufacturingtechnology, for which the incentives are greater and adopted by the larger userfirms(Hutchesonetal.,1996;ReichsteinandSalter,2006).Furthermore,Utterback(1994) argues thatmajor innovations have a tendency to come from unexpecteddirections. However, if major process innovations originate from outside thecompany,thereisaneedtomonitortheexternalenvironmentofthecompany.Therole of a gatekeeper, i.e. a person that helps to overcome barriers raised byorganizational boundaries, has received a fair share of attention in new productdevelopment (NPD) theory. Gatekeepers can overcome barriers based ondifferences such as terminology, norms, and values (Allen, 1977; Tushman andScanlan, 1981) and can be described as key communicators (Davis and Wilkof,1988)whoarestronglylinkedtotheinternalandexternalorganization(TushmanandScanlan,1981).Consequently,gatekeepersat themanufacturingcompanycanprovide a link between the organization and its environment by collecting andtranslating relevant information (see Paper III). Therefore, engaging gatekeepersalso in theproduction systemdesignprocess couldbeof value for increasing theinformation flow between the manufacturing company and the externalenvironment.Further,overa longtimethefocusintheproductionsystemareahasnotbeenonthe design process; rather a considerable amount of work has been devoted toproductionflexibility,i.e.thecreationof“the”systemthathas“theabilitytochangeor react with little penalty in time, effort, cost, or performance” (De Toni andTonchia, 1998, p. 1591). Since flexibility costs money, it is necessary to define adynamicsetofrequirementstowhichtheproductionsystemwillreactwithnoorminimalstructuralintervention(Slack,1987).However,concernshavebeenraisedthat the high level of uncertainty of today’s competitive environment makes itimpossible to specify the production flexibility required. Further, it has beenpointedoutthatmanufacturingcompaniesfindtheconceptofflexibilitydifficulttodealwithdespiteitsbeingrecognizedascentraltodesignefforts(Slack,2000).InCaseBandCaseP,theneedtopreparetheproductionsystemforfutureproductvariants was emphasized. With regard to design production systems that couldhandle future product requirements, information about the life length of the


89

product and potential future markets and customers had been collected andanalysed.Thisinformationwasusedtodeterminetherequiredflexibilityandlengthoflifeoftheproductionsystem.However,thecasestudiesdemonstratethathavinga long‐term perspective of the design of the production system was not easilyachieved. An important conclusion drawn in Paper VI was that in order toaccomplishamorelong‐termdesignofaproductionsystemitwasnotsufficienttoinclude only design information about the current product generation, but alsodesign information that was related to the introduction of future generations ofproductandproductvariantswasneeded.However,thefindingsalsorevealedthatit was difficult to identify, structure, and document the information in anappropriateway thatwould allow for conscious planning of a production systemthatispreparedforbothtoday’sandtomorrow’schallenges.Asaresult,thereisaneedtolimitthedemandforflexibilitybyusingstrategiesthatcancontributetoamoresystematicandstructuredwayofplanningforchangesinthe product design, whichmay lead to changes in the production system design.Productionsystemsneedtobedesignedinawaythattheycanhandleawiderangeofproductvariantsandupdateswithoutrequiringanexcessivelevelofflexibility,asthiswouldbenoteconomicallyjustifiable.Thus,fromamoregeneralperspective,tosynchronizeonlythedesignoftheproductionsystemwiththeactualproductstobemanufactured is insufficient. Rather, the introduction of future generations ofproducts and product variants must be taken into account when designingproductionsystems.Onepossibilitywouldbetocreateaproductportfoliooffutureproductsandaco‐portfolioofproductionsystems.Thisproactivewayof thinkingleads to aportfolio of a numberof tentativeproducts and aportfolio of tentativeproduction systems that match each other (see also Paper VI). Therefore, it isargued that manufacturing companies could accelerate their design process byusing a portfolio approach in which the product and the production systemportfolioareplannedsimultaneously.Aneffectiveportfoliomanagementisaboutmakingstrategicchoices(Cooperetal.,1999), which necessitates systematic and careful planning tomake sure that theproducts will fit well with the company’s overall objective. However, applying aproductionsystemportfolioapproachplacesdifferentdemandsonthemanagementofdesigninformation.Tobeabletoincludeevenfuturedemandsandalsoinfluencefuture product design calls for design information. The creation of a productionsystem portfolio alongside the product portfolio requires applying preliminaryinformationinthedesignoftheproductionsystemtoalargerextentthanrequiredwhenonlyconsideringtheactualproductgeneration.Theempiricalfindingsofthecurrentresearchandpriorresearch(HauptmanandHirji,1996; Johansson,2009)has,however,shownthat theuseofpreliminary informationmayhaveanegativeimpactonwhetherornotdesigninformationisacquired,shared,andused.7.2 CONCLUSIONSToday’smanufacturing industry ismoreglobal thanever facingnewchallengesoflong‐term viability. Designing production systems in an efficient and effectivemannerwas initiallypointedoutas critical to creatingcompetitiveadvantages. Inthe researchpresented in this thesis, it is argued that the capability to effectivelymanagetherequireddesigninformationisonewaytofacilitatethedesignprocess.


90

The objective of the research has been to develop knowledge to contribute to aneffectivemanagementofdesigninformationwhendesigningproductionsystems.Inordertofulfiltheresearchobjective,tworesearchquestionshavebeenformulatedandanswered.The answer to the first research question reveals the wide variety of designinformation thatneeds tobeconsidered in theproductionsystemdesignprocess.Based on the results of the investigation, it can be concluded that designinformation required when designing the production system is a combination ofinformationthatminimizestheriskofsuboptimization,ensuresanalignmentwithexternalandinternaldemands,andsupportsadvancementsintheresearchprocess.Theanswertothefirstresearchquestioncontributestoachievingtheobjectivebyincreasing the understanding of what design information is required during theprocess of designing the production system and thus identifies what designinformationneedstobemanaged.TheanswertothesecondresearchquestionconfirmspriorresearchfromtheNPDfieldseeingthemanagementofdesigninformationasamultidimensionalconstructconsistingofthethreedimensionsacquiring,sharing,andusingdesigninformationalso when designing the production system. Further, it is argued that themanagement of design information is affected by six characteristics that can beattributed to the three dimensions. These characteristics are type of information,sourceof information,communicationmedium, formalization, informationquality,andpragmaticinformation.Theanswertothesecondresearchquestioncontributestoachievingtheobjectivebyincreasingtheunderstandingofwhatcharacterizesthemanagement of design information and thus identifies how the process can beshapedtobestfitthecontextoftheparticularproductionsystemdesignproject.The answers to the two research questions were synthesized into a designinformationmanagement framework, see Figures 22 and 23. The findings are animportant first step in an area that has received only limited attention fromacademics and practitioners alike. The framework visualizes the complexity ofmanaging design information and presents a structured approach of how designinformation can be managed when designing the production system. In the longterm the framework may even facilitate handling production system designactivitiesmoreproactively.Aproactivewayofworkingimpliesthateventsaredealtwithbeforeorwhentheyoccurinordertocreatefavourableoutcomes.AshasbeenillustratedinFigure5,proactivehandlingofworkactivitiesdependsonwhetheraneventisplannedand/orwhetherthehandlingoftheeventisknown.Therefore,aneffectivemanagementofdesigninformationshouldleadtoimprovedpossibilitiestore‐use previous experiences and to plan for different activities that need to becarriedoutintheproductionsystemdesignprocess.7.3 METHODOLOGICALDISCUSSIONAlthoughit isbelievedthatthecurrentresearchcontributestoknowledgeaboutamoreeffectivemanagementofdesign informationwhendesigningtheproductionsystem, it is important toacknowledge that theselectedresearchmethodand theresearchdesigninfluencetheconclusionsthatcanbedrawnintheresearch.


91

Thechoiceofthecasestudymethodforansweringtheresearchquestionsresultedinverydetailedandrichdataabout themanagementofdesign informationwhendesigningtheproductionsystem.Therefore,achallengeperceivedintheuseofthecase studymethodwas handling the overwhelming amount of data. As Pettigrew(1990,p.281)concludes,thevolumeofdatabringsaboutthedangerof“deathbydata asphyxiation”.While in statistical analyses the researcher is guided by rulesand formulas, support in the analysis of case study data has been insufficientlydeveloped (Yin, 2009). To minimize the risk that the amount of data wasoverwhelming, the collected data were continuously analysed and each datacollection occasion was prepared in order to avoid what Voss et al. (2002) call“industrial tourism”, i.e. visiting numerous organizations without being certainabouttheobjectiveoftheresearch.Further,theanalysiswasbasedonamodelthatcontributed to a structured and systematic analysis of the data. It is necessary toconsider analytic approaches in the research design to avoid ending up in asituationofnotknowingwhattodowiththecollecteddata(Yin,2009).Theresearchdesignrestsontworeal‐timeandtworetrospectivecasestudiesandadescriptive survey. The case studies were performed in the automotive industrybecause of the dynamic conditions of handling shrinking product life cyclescombinedwithan increasingnumberofproductmodelsandvariants.The surveywas answered by representatives of equipment suppliers, since equipmentsuppliers are of great importance for the design of production systems. Theresearch design was perceived as suitable as it revealed similarities but alsohighlightedindividualdifferences.An industrializationproject isoften carriedoutover severalmonthsoryearsandincludesdifferentfunctions,whichmadeitdifficulttocoordinatethedatacollection.However, a considerable amountof timewas spent at the case study company inCaseBandCasePandthuseventscouldbeobservedatfirsthandinsteadofrelyingonretrospectivenarratives.InCaseB,theindustrializationprojectwasstudiedforoneandahalfyears,whichmadeitimpossibletospendeverydayatthecompany.Therefore,tightcoordinationbetweentheresearcherandtheprojectmemberswasusedtomakesurethatthatnocriticalissuesweremissed.Inaddition,tominimizethe risk of relying on the retelling of individual employees, various kinds of datatriangulationwereapplied.It is evident that the outcome of the research was influenced by the selectedresearch method. The question is whether an explanatory survey method couldhave been an alternative method for studying the management of designinformation.OnepossibilitywouldhavebeentotakethepropositionsmadeinpriorNPDresearchandtesttheirvalidationintheproductionsystemdesignprocess.Itislikely that suchanapproachwouldhave led tomoreanalyticalgeneralizabilityofthe findings. Nevertheless, as noted byMeredith (1998), the increased reliabilityandvalidityassociatedwithamorerationalistapproachcouldonlybeobtainedatthe expense of the contextual and temporal richness that case studies offer. As aresult, the findingswouldnothavebeenasdetailedas theonespresented in thisthesis and there would have been the risk that critical issues particular to theproductionsystemdesignprocesswouldhavebeenlost.


92

7.4 DISCUSSIONOFTHECONTRIBUTIONSOFTHERESEARCH

7.4.1 ScientificcontributionsTheproduction systemdesignprocess refers to theprocessof creatingadetaileddescriptionof theproposedproductionsystem.What this thesishasadded to thefield of the design of production systems is new knowledge about the designinformationrequired in thedesignprocessby identifying tencategoriesofdesigninformation(seeFigure15)thatcanbedividedintofourmaingroups.Further,theresearchpresentedinthethesisoutlineswhatcategoriesofdesigninformationarerequiredinthedifferentphasesoftheproductionsystemdesignprocessandwherethedesigninformationisdeveloped,asillustratedinFigure16.Anothercontributionofthisthesisistheidentificationofthesixcharacteristicsthataffectthemanagementofdesigninformation.Thecharacteristicscanbeassignedtothethreedimensionsacquiring,sharing,andusingdesigninformation.Whilepriorresearchdevelopedtoolsthatsupportedtheoverallproductionsystemdesign,thecurrent research outlines a design information management framework thatprovidesimportantinsightsregardingtheactualmanagementofdesigninformationwhen designing production systems. Further, the empirical findings presented inthis thesis increase the overall awareness of the importance of an effectivemanagement of design information as a critical factor for successfulindustrializationprojects.7.4.2 IndustrialcontributionsIn the production system design process a wide variety of design information isrequired,whichhas tobe acquiredandused to create aproduction system.Eventhough the studied companieshada routineofdesigningproduction systems, themanagement of design informationwas challenging. By using the outlined designinformationmanagementframework,challengesrelatedtothemanagingofdesigninformationcanbereduced.Forinstance,itisindicatedwheretherelevantdesigninformationisgenerated,whichshouldmakeiteasiertoaccesstherequireddesigninformation.Further, a common way of working in NPD projects is to regard the productionsystemdesignprocessasoneofseveralsub‐activities.However,onthebasisofthepresentresults,itispossibletoarguethatthekeytosuccessliesintheabilityofamanufacturingcompanyanditsmanagerstounderstandthatitmaynotbeenoughto consider the production system design process as just one of many activitiesrequired to introduce a product to the market. The production system designprocessrequiresstructureandtransparencyforthepeopleinvolvedintheprocess.Based on the analysis in Chapter 5, it is possible to see the potential benefits offormalandinformalcoordinationmechanisms.Ahighdegreeofformalizationintheproduction system design process should improve the sharing of informationbetweenspecializedfunctionsandalsoincreasetheunderstandingofthedifferentneedsofthevariousfunctionsinvolvedinthedesignprocess.Inaddition,ahigherdegree of formalization contributes to more documentation. Documentation iscrucialtolearningfrompreviousmistakesandtotransferringknowledgebetweendifferentproductionsystemdesignprojects.


93

Thefindingsofthecurrentthesisshowthatthemanagementofdesigninformationisheavilydependentonthepersonsinvolvedinthedesignprocess.Thismeansthatinformation is preferably acquired from personal sources and shared throughhuman interaction.This increases the risk that informationgets lostwhenpeoplechangetheirworkingplacesorarenotinvolvedinfutureprojects.Further,althoughthecurrentresearchdoesnotoutlineanykindofformalizedinformationsystemtobe used when designing a production system, it highlights that manufacturingcompaniesshouldpayattentiontodevelopinganinformationsystemtobeusedindesignprocesses. Special attention shouldbegiven tomeasures that facilitate thetransferofrichinformation,i.e.informationthathelpstoreduceequivocality.Thiskindofinformationisoftendifficulttocaptureandmanagebutofhighvaluewhendesigningtheproductionsystem.Onepossibilitywouldbetoexplorethepotentialofvideo‐andaudiotapes.Asmoreandmoreprojectsarecarriedoutininternationalenvironments where functions are located at different geographical places,companies are in urgent need of information systems that support them inmanagingdisparatekindsofdesigninformationandinmakingdesigninformationavailableacrossfunctionsandlocations.7.5 SUGGESTIONSFORFURTHERRESEARCHSubsequentand future research in theareaofmanagementofdesign informationwhendesigningtheproductionsystemcantakeanumberofdirections.Whileacasestudydesigngenerallyallowsforretainingtheholisticandmeaningfulviewofthephenomenonstudied,thegeneralizationofthefindingshastobedonewith caution. The case studies have been carried out in the automotive industry.Hence, more research is clearly needed to evaluate and test the value of theproposed framework. One possibility would be to replicate the empirical datacollectionindifferenttypesofindustries.Anotherpossibilitywouldbetoturntoafixedresearchapproachsuchasthesurveymethodtoinvestigatethepossibilityofstatistical generalization. As Karlsson (2009) has remarked, in order to developknowledge in a field, there is a need to go throughdifferent phases startingwithexploring before being able to describe a field of knowledge, to know thecomponentsbeforeunderstanding the relations, and toknow the relationsbeforepredictingtheeffects.Further research is also required concerning the identified characteristics (seeFigure 21) and the design informationmanagement framework. Although the sixcharacteristicsseemplausiblesincethesignificanceoftheidentifiedcharacteristicshasbeen identified inprior research, it is likely that the current researchhasnotidentified the complete range of characteristics that affect the management ofdesign information. Further, the identified six characteristics are probablyinterrelated.Forexample,thesourceofinformationhasprobablyconsequencesforthecommunicationmediumapplied.Accordingly,furtherresearchshouldfocusonhow these characteristics are connected and their consequences for themanagementofdesigninformation.AnotherlimitationoftheresearchwasthatthesuggestionsmadeinFigure23areonlyphase‐specifictoalimitedextent.However,itwouldbeofinteresttostudyifthemanagementofdesigninformationhastobemoreadjustedtoeachphaseduetothedifferentworkactivitiesconductedineach


94

phaseofthedesignprocessandifsomecharacteristicsaremoreorlessrelevantinthedifferentphases.Finally,alsothesettingsofthecurrentresearchinfluencedtheconclusionsdrawn.Inthecasesstudied,theinternalfunctionsinvolvedintheproductionsystemdesignprocesswerelocatedatthesamesite.Today,thereisnoguaranteethatallfunctionsare located at the same site, and one general conclusion on studies of dispersedsettings is that geographical distance influences the flow of information. Thus, itwouldbeof interesttosee ifandhowthemanagementofdesigninformationwillchangewhenkeyfunctionsarelocatedatdifferentplaces.Anotherimportantaspectthat turns into a suggestion for further research is the fact that the frameworkpresentedinthecurrentresearchistoalargeextentlimitedtothepresentproductgeneration, i.e. the design of the production system was aligned to the actualproduct developed. However, in an ever‐changing environment with reducedproduct life cycles, the issue of design information concerning future productgenerations isofhighrelevance.Tobeable todesignproductionsystemsthatareeasilyadjustabletofutureproductgenerationsrequireshavinganefficientflowofinformationbetweenproductdesignandmanufacturingfunctionsindicatingfutureproductdemandsbutalsomanufacturingcapabilities.Assuggested,onepossibilitywouldbe tobase futureresearchon the ideaofusingaproductionsystemdesignportfolioapproachwhendesigningproductionsystems.Research in the field of production system design and particularly on themanagement of design information is valuable in the effort to create the bestpossible production system. Since an effective and efficient production systemdesignprocesscanplayacriticalroleingaininganedge,themotivationforfurtherresearchintheareashouldbehigh.


95

REFERENCES

Abernathy, W.J. and Wayne, K. (1974), "Limits of the Learning Curve", HarvardBusinessReview,Vol.52,No.5,pp.109‐119.

Adler,P.S.(1995),"InterdepartmentalInterdependenceandCoordination:TheCaseof theDesign/Manufacturing interface",Organization Science, Vol. 6, No. 2,pp.147‐167.

Aguilar,F.J.(1967),ScanningtheBusinessEnvironment,NewYork,NY:Macmillan.Aitken,J.,Childerhouse,P.andTowill,D.(2003),"Theimpactofproductlifecycleon

supplychainstrategy",InternationalJournalofProductionEconomics,Vol.85,No.2,pp.127‐140.

Allen, T.J. (1970), "Communication networks in R & D Laboratories", R&DManagement,Vol.1,No.1,pp.14‐21.

Allen, T.J. (1977),Managing the Flow ofTechnology:TechnologyTransfer and theDissemination of Technological Information Within the R&D Organization,Cambridge,MA:MITPress.

Alm, I. and Fjällström, S., (2003), "Dealing with Production Disturbances: TheSignificance of Change and Pragmatic Information", in Automated SystemsBased onHuman Skill andKnowledge, Proceedings of the 8th InternationalFederation of Automatic Control Symposium, Gothenburg, Sweden, 22‐24September,2003.

Andersson, F. (2003), The Dynamics of Requirements and Product ConceptManagement: A Product Modelling Approach, PhD Thesis, ChalmersUniversityofTechnology,Gothenburg,Sweden.

Andreasen, M.M. and Hein, L. (1987), Integrated product development, Berlin,Germany:Springer‐Verlag.

Arbnor,I.andBjerke,B.(2009),Methodology forCreatingBusinessKnowledge,3rded.,London,UK:SAGE.

Asnafi, N.,Hjelm, S., Holmgren, B.,Malmsköld, L., Granström,M., Svenningstorp, J.andMellgren,H.(2008),2015SustainableManufacturingSystemsCapableofProducing Innovative Environmentally Friendly and Safe Products, R&DprogramproposaltosecurecompetitivevehicleandpowertrainproductioninSweden,

Barney, J. (1991), "FirmResourcesandSustainedCompetitiveAdvantage", JournalofManagement,Vol.17,No.1,pp.99‐120.

Barney, J.,Wright,M.andKetchenJr.,D.J.(2001),"Theresource‐basedviewofthefirm:Tenyearsafter1991",JournalofManagement,Vol.27,No.6,pp.625‐641.

Barton,J.A.,Love,D.M.andTaylor,G.D.(2001),"Designdetermines70%ofcost?Areviewofimplicationsfordesignevaluation",JournalofEngineeringDesign,Vol.12,No.1,pp.47‐58.

Baxter,M.(1995),ProductDesign:Practicalmethodsforthesystematicdevelopmentofnewproducts,London,UK:Chapman&Hall.

Bellgran,M.(1998),SystematicDesignofAssemblySystems:PreconditionsandDesignProcessPlanning,PhDThesis,LinköpingUniversity,Linköping,Sweden.

REFERENCES

96

Bellgran,M.andSäfsten,K.(2010),ProductionDevelopment:DesignandOperationofProductionSystems,London,UK:Springer‐Verlag.

Bennett, D. (1986), Production Systems Design, London, UK: Butterworth‐Heinemann.

Bennett,D.J.andForrester,P.L.(1993),Market‐FocusedProductionSystems:DesignandImplementation,NewYork,NYandLondon,UK:PrenticeHall.

Berglund,M., Skog, E. andÖjmertz, B. (2001),Driftsättningavmonteringssystem ‐Erfarenheterochsynpunkterfrånsvenskaindustriföretag,IVF‐rapport01009[inSwedish],IVF,Mölndal,Sweden.

Bergman, B. and Klefsjö, B. (2010), Quality from customer needs to customersatisfaction,3rded.,Lund:Studentlitteratur.

Bozarth,C.andMcDermott,C.(1998),"Configurationsinmanufacturingstrategy:areview and directions for future research", Journal of OperationsManagement,Vol.16,No.4,pp.427‐439.

Bruch, J. (2009), Information Requirements in a Proactive AssemblyWork Setting,LicentiateThesis,ChalmersUniversityofTechnology,Gothenburg,Sweden.

Bruch,J.andKarltun,J.,(2009),"InformationRequirementsinaProactiveAssemblyWork Setting", Proceedings of the 3rd International Conference onChangeable, Agile, Reconfigurable and Virtual Production (CARV 2009),Munich,Germany,5‐7October,2009.

Bruzelius,L.H.andSkärvad,P.‐H.(1989),Integreradorganisationslära[inSwedish],6thed.,Lund:Studentlitteratur.

Butcher,H.(1998),MeetingManagers'InformationNeeds,London,UK:Aslib.Byström, K. (1999), Task Complexity, Information Types and Information Sources:

Examination of Relationships, Doctoral Thesis, University of Tampere,Tampere,Finnland.

Chapanis, A. (1996),Human Factors in Systems Engineering, New York, NY: JohnWiley&Sons.

Checkland, P. (1999), Systems Thinking, Systems Practice: Includes a 30‐yearretrospective,8thed.,Chichester,UK:JohnWiley&Sons.

Chisholm,A.W.J.(1990),"Nomenclatureanddefinitionsformanufacturingsystems",CIRPAnnals,Vol.39,No.2,pp.735‐742.

Choudhari, S.C., Adil, G.K. and Ananthakumar, U. (2010), "Congruence ofmanufacturing decision areas in a production system: a researchframework",InternationalJournalofProductionResearch,Vol.48,No.20,pp.5963‐5989.

Christensen,C.M.(2006),"TheOngoingProcessofBuildingaTheoryofDisruption",JournalofProductInnovationManagement,Vol.23,No.1,pp.39‐55.

Chryssolouris,G.(2006),ManufacturingSystems:TheoryandPractice,2nded.,NewYork,NY:Springer.

Churchman, C.W. (1978), Systemanalys [in Swedish] (Original title: The SystemsApproach,translatedbyBjörk,L.),Stockholm:Rabén&Sjögren.

Clark, K.B. and Fujimoto, T. (1991), Product Development Performance: Strategy,Organization, and Management in the World Auto Industry, Boston, MA:HarvardBusinessSchoolPress.

Cochran,D.S., Arinez, J.F., Duda, J.W. andLinck, J. (2001/2002), "Adecompositionapproach for manufacturing system design", Journal of ManufacturingSystems,Vol.20,No.6,pp.371‐389.


97

Cooper, R.G. (1975), "Why new industrial products fail", Industrial MarketingManagement,Vol.4,No.6,pp.315‐326.

Cooper, R.G. (1999), "From Experience: The Invisible Success Factors in ProductInnovation", Journal ofProduct InnovationManagement, Vol. 16, No. 2, pp.115‐133.

Cooper, R.G. (2008), "Perspective: The Stage Gate® Idea‐to‐Launch Process—Update, What's New, and NexGen Systems", Journal of Product InnovationManagement,Vol.25,No.3,pp.213‐232.

Cooper, R.G. and Kleinschmidt, E.J. (1991), "New product processes at leadingindustrial firms", IndustrialMarketingManagement,Vol.20,No.2,pp.137‐147.

Cooper, R.G., Edgett, S.J. and Kleinschmidt, E.J. (1999), "New Product PortfolioManagement: Practices and Performance", Journal of Product InnovationManagement,Vol.16,No.4,pp.333‐351.

Cross,N.(2000),EngineeringDesignMethods:StrategiesforProductDesign,3rded.,Chichester,UK:JohnWiley&Sons.

Daft, R.L. and Lengel, R.H. (1984), "Information Richness: A New Approach toManagerial Information Processing and Organization Design", Research inOrganizationalBehavior,Vol.6,pp.191‐233.

Daft, R.L. and Weick, K.E. (1984), "Toward a Model of Organizations asInterpretation Systems",AcademyofManagementReview, Vol. 9,No. 2, pp.284‐295.

Daft,R.L.andLengel,R.H.(1986),"OrganizationalInformationRequirements,MediaRichnessandStructuralDesign",ManagementScience,Vol.32,No.5,pp.554‐571.

Daft, R.L., Sormunen, J. and Parks, D. (1988), "Chief executive scanning,environmental characteristics, and company performance: An empiricalstudy",StrategicManagementJournal,Vol.9,No.2,pp.123‐139.

Dankbaar,B.(2007),"GlobalSourcingandInnovation:TheConsequencesofLosingbothOrganizationalandGeographicalProximity",EuropeanPlanningStudies,Vol.15,No.2,pp.271‐288.

Davenport, T.H. (1997), Information Ecology: Mastering the Information andKnowledgeEnvironment,NewYork,NY:OxfordUniversityPress.

Davenport, T.H. and Prusak, L. (1998), Working knowledge: How organizationsmanagewhattheyknow,Boston,MA:HarvardBusinessSchoolPress.

Davis, P. andWilkof,M. (1988), "Scientific and technical information transfer forhightechnology:keepingthefigureinitsground",R&DManagement,Vol.18,No.1,pp.45‐58.

DeToni,A.andTonchia,S.(1998),"Manufacturingflexibility:Aliteraturereview",InternationalJournalofProductionResearch,Vol.36,No.6,pp.1587‐1617.

Deshpande, R. and Zaltman, G. (1982), "Factors Affecting the Use of MarketResearch Information:APathAnalysis", JournalofMarketingResearch,Vol.19,No.1,pp.14‐31.

Draaijer, D. andBoer,H. (1995), "Designingmarket‐oriented production systems:theoryandpractice",IntegratedManufacturingSystems,Vol.6,No.4,pp.4‐15.

REFERENCES

98

Duda, J.W. (2000), A Decomposition‐Based Approach to Linking Strategy,PerformanceMeasurement, andManufacturing System Design, PhD Thesis,MassachusettsInstituteofTechnology,Cambridge,MA.

Edmunds, A. and Morris, A. (2000), "The problem of information overload inbusiness organisations: a review of the literature", International Journal ofInformationManagement,Vol.20,No.1,pp.17‐28.

Eisenhardt,K.M.(1989),"BuildingTheoriesfromCaseStudyResearch",AcademyofManagementReview,Vol.14,No.4,pp.532‐550.

ElMaraghy,H.A.(2009),"ChangingandEvolvingProductsandSystems‐ModelsandEnablers", in ElMaraghy, H.A. (Ed.), Changeable and ReconfigurableManufacturingSystems,London:Springer‐Verlag,pp.25‐45.

ElMaraghy, H. and Wiendahl, H.‐P. (2009), "Changeability ‐ An Introduction", inElMaraghy, H.A. (Ed.), Changeable and Reconfigurable ManufacturingSystems,London,UK:Springer‐Verlag,pp.3‐24.

Engström, T., Jonsson, D. and Medbo, L. (1998), "The Volvo Uddevalla plant andinterpretations of industrial design processes", Integrated ManufacturingSystems,Vol.9,No.5,pp.279‐295.

Eppler, M.J. (2006), Managing Information Quality: Increasing the Value ofInformation inKnowledge‐intensiveProducts andProcesses, 2nd ed., Berlin,Germany:Springer‐Verlag.

Fjällström,S.(2007),TheRoleofInformationinProductionRamp‐upSituations,PhDThesis,ChalmersUniversityofTechnology,Gothenburg,Sweden.

Fjällström, S., Säfsten, K., Harlin, U. and Stahre, J. (2009), "Information enablingproductionramp‐up",JournalofManufacturingTechnologyManagement,Vol.20,No.2,pp.178‐196.

Forza, C. (2009), "Surveys", in Karlsson, C. (Ed.), Researching OperationsManagement,NewYork,NY:Routledge,pp.84‐161.

Frishammar,J.(2003),"Informationuseinstrategicdecisionmaking",ManagementDecision,Vol.41,No.4,pp.318‐326.

Frishammar, J. (2005),TowardsaTheoryofManaging Information inNewProductDevelopment,PhDThesis,LuleåUniversityofTechnology,Luleå.

Frishammar, J. and Hörte, S.Å. (2005), "Managing External Information inManufacturing Firms: The Impact on Innovation Performance", Journal ofProductInnovationManagement,Vol.22,No.3,pp.251‐266.

Frishammar, J. and Ylinenpää, H. (2007), "Managing Information in New ProductDevelopment: A Conceptual Review, Research Propositions and TentativeModel", International Journalof InnovationManagement, Vol. 11,No. 4, pp.441‐467.

Frishammar, J., Florén, H. andWincent, J. (2011), "Beyondmanaging uncertainty:insights from studying equivocality in the fuzzy front end of product andprocessinnovationprojects",IEEETransactionsonEngineeringManagement,Vol.58,No.3,pp.551‐563.

Galbraith, J.R. (1973), Designing Complex Organizations, Reading, MA: Addison‐Wesley.

Galliers,R.(1987),"OntheNatureofInformation:Introduction",inGalliers,R.(Ed.),Information Analysis: Selected Readings,Wokingham, UK: Addison‐Wesley,pp.3‐6.


99

Gerwin,D. andBarrowman,N.J. (2002), "AnEvaluationofResearchon IntegratedProductDevelopment",ManagementScience,Vol.48,No.7,pp.938‐953.

Girotra, K., Terwiesch, C. and Ulrich, K.T. (2007), "Valuing R&D Projects in aPortfolio:EvidencefromthePharmaceuticalIndustry",ManagementScience,Vol.53,No.9,pp.1452‐1466.

Griffin, A. (1997), "PDMA Research on New Product Development Practices:Updating Trends and Benchmarking Best Practices", Journal of ProductInnovationManagement,Vol.14,No.6,pp.429‐458.

Griffin,A.andHauser, J.R. (1996), "IntegratingR&DandMarketing:AReviewandAnalysis of theLiterature", JournalofProduct InnovationManagement, Vol.13,No.3,pp.191‐215.

Groover, M.P. (2008), Automation, Production Systems, and Computer‐IntegratedManufacturing,3rded.,UpperSaddleRiver,NJ:Prentice‐Hall.

Gupta, A.K., Raj, S.P. and Wilemon, D. (1987), "Managing the R&D‐marketinginterface",ResearchManagement,Vol.30,No.2,pp.38‐43.

Ha,A.Y.andPorteus,E.L.(1995),"OptimalTimingofReviewsinConcurrentDesignforManufacturability",ManagementScience,Vol.41,No.9,pp.1431‐1447.

Harlin, U. (2000), Towards Strategic Disturbance Management in AdvancedManufacturing Systems, Licentiate Thesis, Chalmers University ofTechnology,Gothenburg,Sweden.

Hauptman, O. and Hirji, K.K. (1996), "The influence of process concurrency onproject outcomes in product development: an empirical study of cross‐functional teams", IEEE Transactions on EngineeringManagement, Vol. 43,No.2,pp.153‐164.

Hayes, R.H. and Wheelwright, S.C. (1984), Restoring our Competitive Edge:CompetingThroughManufacturing,NewYork,NY:JohnWiley&Sons.

Hayes,R.,Pisano,G.,Upton,D.andWheelwright,S.(2005),Operations,Strategy,andTechnology:PursuingtheCompetitiveEdge,Hoboken,NJ:JohnWiley&Sons.

Hendricks,K.B.andSinghal,V.R.(2008),"TheEffectofProductIntroductionDelaysonOperatingPerformance",ManagementScience,Vol.54,No.5,pp.878‐892.

Hertzum, M. and Pejtersen, A.M. (2000), "The information‐seeking practices ofengineers: searching for documents as well as for people", InformationProcessing&Management,Vol.36,No.5,pp.761‐778.

Hobday, M. (1998), "Product complexity, innovation and industrial organisation",ResearchPolicy,Vol.26,No.6,pp.689‐710.

Hobday, M. (2000), "The project‐based organisation: an ideal form for managingcomplexproductsandsystems?",ResearchPolicy,Vol.29,No.7‐8,pp.871‐893.

Hobday, M., Prencipe, A. and Davies, A. (2005), "Introduction", in Prencipe, A.,Davies, A. andHobday,M. (Eds.),TheBusinessof Systems Integration,NewYork,NY:OxfordUniversityPress,pp.1‐14.

Hoopes,D.G.,Madsen,T.L.andWalker,G.(2003),"Guesteditors'introductiontothespecial issue: why is there a resource‐based view? Toward a theory ofcompetitive heterogeneity", StrategicManagement Journal, Vol. 24, No. 10,pp.889‐902.

Hubka, V. and Eder, W.E. (1988), Theory of Technical Systems: A Total ConceptTheoryforEngineeringDesign,Berlin,Germany:Springer‐Verlag.

REFERENCES

100

Hutcheson,P.,Pearson,A.W.andBall,D.F. (1995), "Innovation inProcessPlant:ACaseStudyofEthylene",JournalofProductInnovationManagement,Vol.12,No.5,pp.415‐430.

Hutcheson,P.,Pearson,A.W.andBall,D.F.(1996),"Sourcesoftechnicalinnovationin the network of companies providing chemical process plant andequipment",ResearchPolicy,Vol.25,No.1,pp.25‐41.

Häckner, E. (1988), "Strategic development and information use", ScandinavianJournalofManagement,Vol.4,No.1‐2,pp.45‐61.

Jacobs, P.F. (1996), Stereolithography and Other RP&M Technologies: From RapidPrototypingtoRapidTooling,NewYork,NY:ASMEPress.

Johansson,A.(1999),OnOptimizingManufacturingSystems:AUserPerspective,PhDThesis,ChalmersUniversityofTechnology,Gothenburg,Sweden.

Johansson, B. and Nord, C. (1999),Nyanskaffning av produktionssystem –mer änbara inköp [in Swedish], Mölndal, Sweden: Institutet för verkstadstekniskforskning(IVF).

Johansson, E. (2009), "Information management for materials supply systemsdesign",InternationalJournalofProductionResearch,Vol.47,No.8,pp.2217‐2229.

Kahn, K.B. (1996), "Interdepartmental Integration: A Definitionwith Implicationsfor Product Development Performance", Journal of Product InnovationManagement,Vol.13,No.2,pp.137‐151.

Kahn,B.K., Strong,D.M. andWang,R.Y. (2002), "InformationQualityBenchmarks:ProductandServicePerformance",CommunicationsoftheACM,Vol.45,No.4,pp.184‐192.

Karlsson, C. (2009), "Researching Operations Management", in Karlsson, C. (Ed.),ResearchingOperationsManagement,NewYork,NY:Routledge,pp.6‐41.

Keen,P.G.W.andKnapp,E.M.(1996),EveryManager'sGuidetoBusinessProcesses:AGlossary ofKeyTerms&Concepts forToday'sBusiness Leader, Boston,MA:HarvardBusinessSchoolPress.

Kehoe, D.F., Little, D. and Lyons, A.C., (1992), "Measuring a company IQ",Proceedings of the Third International Conference on Factory 2000 –Competitive Performance Through Advanced Technology, London, England,27‐29July1992,pp.173‐178.

Keller,K.L.andStaelin,R.(1987),"EffectsofQualityandQuantityofInformationonDecisionEffectiveness",JournalofConsumerResearch,Vol.14,No.2,pp.200‐213.

Krishnan,V.andUlrich,K.T.(2001),"ProductDevelopmentDecisions:AReviewoftheLiterature",ManagementScience,Vol.47,No.1,pp.1‐21.

Lager,T.andFrishammar,J.(2010),"Equipmentsupplier/usercollaborationintheprocessindustries:insearchofenhancedoperatingperformance",JournalofManufacturingTechnologyManagement,Vol.21,No.6,pp.698‐720.

Larsson, A.C. (2006), "Interactive Research – Methods and conditions för jointanalysis", in Nielsen, K.A. and Svensson, L. (Eds.), Action and InteractiveResearch. Beyond Practice and Theory, Maastricht, Netherlands: ShakerPublishing,pp.241‐258.

Lawrence,P.R.andLorsch,J.W.(1967),"DifferentiationandIntegrationinComplexOrganizations",AdministrativeScienceQuarterly,Vol.12,No.1,pp.1‐47.


101

Lawrence, P.R. and Lorsch, J.W. (1986),OrganizationandEnvironment:ManagingDifferentiationandIntegration,Boston,MA:HarvardBusinessSchoolPress.

Leonard‐Barton,D.(1990),"ADualMethodologyforCaseStudies:SynergisticUseofa Longitudinal Single Site with Replicated Multiple Sites", OrganizationScience,Vol.1,No.3,pp.248‐266.

Lesca,H. andLesca,E. (1995),Gestionde l'information:Qualitéde l'informationetperformancesdel'entreprise[inFrench],Paris:Litec.

Lewis,M.,Brandon‐Jones,A.,Slack,N.andHoward,M.(2010),"Competingthroughoperations and supply: The role of classic and extended resource‐basedadvantage", International Journal of Operations& ProductionManagement,Vol.30,No.10,pp.1032‐1058.

Liker, J.K. (2004), The ToyotaWay: 14 Management Principles from theWorld'sGreatestManufacturer,NewYork,NY:McGraw‐Hill.

Love, D. (1996), "The Design of Manufacturing Systems", in Warner, M. (Ed.),InternationalEncyclopaediaofBusinessandManagement,Vol.4,London,UK:Routledge,pp.3154‐3174.

Mabert, V.A., Muth, J.F. and Schmenner, R.W. (1992), "Collapsing New ProductDevelopment Times: Six Case Studies", Journal of Product InnovationManagement,Vol.9,No.3,pp.200‐212.

MacGregor,S.P.,Thomson,A.I.andJuster,N.P.,(2001),"InformationSharingWithina Distributed, Collaborative Design Process: A Case Study", Proceedings ofASME 2001 Design Engineering Technical Conferences and Computers andInformation inEngineeringConference,Pittsburgh,PA,USA9‐12September,2001.

Magrab, E.B., Gupta, S.K., McCluskey, F.P. and Sandborn, P. (2010), IntegratedProduct and Process Design and Development: The Product RealizationProcess,2nded.,BocaRaton,FL.:CRCPress.

Makadok, R. (2001), "Toward a synthesis of the resource‐based and dynamic‐capabilityviewsofrentcreation",StrategicManagementJournal,Vol.22,No.5,pp.387‐401.

Martinez, J.I. and Jarillo, J.C. (1989), "The Evolution of Research on CoordinationMechanismsinMultinationalCorporations",JournalofInternationalBusinessStudies,Vol.20,No.3,pp.489‐514.

Mata, F.J., Fuerst, W.L. and Barney, J.B. (1995), "Information Technology andSustained Competitive Advantage: A Resource‐Based Analysis", MISQuarterly,Vol.19,No.4,pp.487‐505.

McCutcheon, D.M. and Meredith, J.R. (1993), "Conducting case study research inoperationsmanagement", JournalofOperationsManagement,Vol.11,No.3,pp.239‐256.

Meredith,J.(1998),"Buildingoperationsmanagementtheorythroughcaseandfieldresearch",JournalofOperationsManagement,Vol.16,No.4,pp.441‐454.

Merriam, S.B. (1994), Fallstudien som forskningsmetod [in Swedish], Lund:Studentlitteratur.

Miles, M.B. and Huberman, A.M. (1994), Qualitative Data Analysis: An ExpandedSourcebook,2nded.,London,UK:SAGE.

Miller,D.(1992),"EnvironmentalFitversusInternalFit",OrganizationScience,Vol.3,No.2,pp.159‐178.

REFERENCES

102

Miller,D.andFriesen,P.H.(1982),"Innovationinconservativeandentrepreneurialfirms:Twomodelsof strategicmomentum",StrategicManagement Journal,Vol.3,No.1,pp.1‐25.

Miltenburg, J. (2005),ManufacturingStrategy:How toFormulateand ImplementaWinningPlan,2nded.,NewYork,NY:ProductivityPress.

Moenaert, R.K. and Souder, W.E. (1990), "An Information Transfer Model forIntegrating Marketing and R&D Personnel in New Product DevelopmentProjects", Journal ofProduct InnovationManagement, Vol. 7, No. 2, pp. 91‐107.

Moenaert, R.K., De Meyer, A., Souder, W.E. and Deschoolmeester, D. (1995),"R&D/marketing communication during the fuzzy front‐end", IEEETransactionsonEngineeringManagement,Vol.42,No.3,pp.243‐258.

Mortensen, M. and Hinds, P. (2002), "Fuzzy Teams: Boundary Disagreement inDistributed and Collocated Teams", in Hinds, P.J. and Kiesler, S. (Eds.),Distributedwork,Cambridge,MA:MITPress,pp.283–308.

Olausson, D. (2009), Facing interface challenges in complex product development,PhDThesis,LinköpingUniversity,Linköping,Sweden.

Olausson,D.andMagnusson,T. (2011),"Manufacturingcompetencies inhigh‐techNPD: on the impact of vertical integration and coordination", InternationalJournal ofManufacturing Technology andManagement, Vol. 22, No. 3, pp.287‐300.

Omar, A.R., Harding, J.A. and Popplewell, K. (1999), "Design for customersatisfaction:an informationmodellingapproach", IntegratedManufacturingSystems,Vol.10,No.4,pp.199‐209.

O'Reilly, C.A. (1982), "Variations in DecisionMakers' Use of Information Sources:The Impact of Quality and Accessibility of Information", Academy ofManagementJournal,Vol.25,No.4,pp.756‐771.

O'Sullivan, D. (1994), Manufacturing Systems Redesign: Creating the IntegratedManufacturingEnvironment,EnglewoodCliffs,NJ:PrenticeHallPTR.

Ottum, B.D. and Moore, W.L. (1997), "The Role of Market Information in NewProductSuccess/Failure",JournalofProductInnovationManagement,Vol.14,No.4,pp.258‐273.

Pahl,G. andBeitz,W. (1996),EngineeringDesign:ASystematicApproach, 2nded.,London:SpringerVerlag.

Patel, R. and Davidson, B. (2003), Forskningsmetodikens grunder: Att planera,genomföra och rapportera en undersökning [in Swedish], 3rd ed., Lund:Studentlitteratur.

Peteraf, M.A. (1993), "The Cornerstones of Competitive Advantage: A Resource‐BasedView",StrategicManagementJournal,Vol.14,No.3,pp.179‐191.

Petersson,S.andPetersson,K.(1992),Internt:ominformationochkommunikationiarbetslivet[inSwedish],Bromma,Sweden:Sakprosa.

Pettigrew, A.M. (1990), "Longitudinal Field Research on Change: Theory andPractice",OrganizationScience,Vol.1,No.3,pp.267‐292.

Pisano, G.P. (1997), The Development Factory: Unlocking the Potential of ProcessInnovation,Boston,MA:HarvardBusinessSchoolPress.

Powell, T.C. and Dent‐Micallef, A. (1997), "Information technology as competitiveadvantage:Theroleofhuman,business,andtechnologyresources",StrategicManagementJournal,Vol.18,No.5,pp.375‐405.


103

Rauterberg,M.andUlich,E.,(1996),"Informationprocessingforlearningsystems:an action theoretical approach", in Information, Intelligence and Systems,Proceedings of the IEEE International Conference on Systems, Man, andCybernetics,Beijing,China,14‐17October1996,pp.2276‐2281.

Reichstein,T.andSalter,A.(2006),"Investigatingthesourcesofprocessinnovationamong UKmanufacturing firms", Industrial andCorporateChange, Vol. 15,No.4,pp.653‐682.

Richtnér,A.(2004),BalancingKnowledgeCreation:ExaminingOrganizationalSlackand Knowledge Creation in Product Development, PhD Thesis, StockholmSchoolofEconomics,Stockholm,Sweden.

Robson, C. (2002), Real World Research: A Resource for Social Scientists andPractitioner‐Researchers,2nded.,Oxford,UK:BlackwellPublishers.

Roozenburg, N.F.M. and Eekels, J. (1995), Product Design: Fundamentals andMethods,Chichester,UK:JohnWiley&Sons.

Ruekert, R.W. and Walker, O.C. (1987), "Marketing's Interaction with OtherFunctional Units: A Conceptual Framework and Empirical Evidence", TheJournalofMarketing,Vol.51,No.1,pp.1‐19.

Ruffini,F.A.J.(1999),Productionsystemdesign.Frompracticetotheory,PhDThesis,UniversityofTwente,Enschede,Netherlands.

Ruffini, F.A.J., Boer, H. and van Riemsdijk, M.J. (2000), "Organisation design inoperations management", International Journal ofOperations& ProductionManagement,Vol.20,No.7,pp.860‐879.

Rönnberg Sjödin, D. and Eriksson, P.E. (2010), "Procurement Procedures forSupplier Integration and Open Innovation in Mature Industries",InternationalJournalofInnovationManagement,Vol.14,No.4,pp.655‐682.

Säfsten,K.(2002),Evaluationofassemblysystems:Anexploratorystudyofevaluationsituations,PhDThesis,LinköpingUniversity,Linköping,Sweden.

Säfsten, K., Harlin, U., Fjällström, S. and Berglund, M., (2008), "Proactive andReactive Ways of Managing Product Introduction", Proceedings of the 2ndSwedishProductionSymposium,Stockholm,18‐20November2008.

Sandkull, B. and Johansson, J. (1996), FrånTaylor tillToyota [in Swedish], Lund:Studentlitteratur.

Sawyerr, O.O., Ebrahimi, B.P. and Thibodeaux, M.S. (2000), "ExecutiveEnvironmental Scanning, Information Source Utilisation, and FirmPerformance: The case of Nigeria", Journal ofAppliedManagement Studies,Vol.9,No.1,pp.95‐115.

Schuh, G., Lenders, M., Nussbaum, C. and Kupke, D. (2009), "Design forChangeability", in ElMaraghy, H.A. (Ed.), Changeable and ReconfigurableManufacturingSystems,London,UK:Springer‐Verlag,pp.251‐266.

Shrivastava,P.(1985),"KnowledgeSystemsforStrategicDecisionMaking",JournalofAppliedBehavioralScience,Vol.21,No.1,pp.95‐107.

Sivadas, E. and Dwyer, F.R. (2000), "An Examination of Organizational FactorsInfluencingNewProductSuccessinInternalandAlliance‐BasedProcesses",TheJournalofMarketing,Vol.64,No.1,pp.31‐49.

Skinner,W.(1969),"Manufacturing–MissingLinkinCorporateStrategy",HarvardBusinessReview,Vol.47,No.3,pp.136‐145.

REFERENCES

104

Skinner,W.(1992),"Theshareholder'sdelight:companiesthatachievecompetitiveadvantage from process innovation", International Journal of TechnologyManagement,Vol.7,No.1/2/3,pp.41‐48.

Slack,N.(1987),"TheFlexibilityofManufacturingSystems",InternationalJournalofOperations&ProductionManagement,Vol.7,No.4,pp.35‐45.

Slack, N. (2000), "Flexibility, trade‐offs and learning in manufacturing systemdesign",InternationalJournalofManufacturingTechnologyandManagement,Vol.1,No.4/5,pp.331‐348.

Slack,N.andLewis,M.(2008),Operationsstrategy,2nded.,Harlow,UK:FTPrenticeHall.

Slack, N., Chambers, S., Harland, C.M., Harrison, A. and Johnston, R. (1998),OperationsManagement,2nded.,London,UK:Pitman.

Smith,P.G.andReinertsen,D.G. (1998),DevelopingProducts inHalftheTime:NewRules,NewTools,NewYork,NY:VanNostrandReinhold.

Souder, W.E. (1988), "Managing Relations between R&D and Marketing in NewProductDevelopmentProjects", JournalofProduct InnovationManagement,Vol.5,No.1,pp.6‐19.

Stalk Jr., G. and Hout, T.M. (2003), Competing Against Time: How Time‐basedCompetitionisReshapingGlobalMarkets,NewYork,NY:FreePress.

Strong, D.M., Lee, Y.W. and Wang, R.Y. (1997), "10 Potholes in the Road toInformationQuality",Computer,Vol.30,No.8,pp.38‐46.

Svensson, L. andNielsen, K.A. (2006), "Introduction and Background", inNielsen,K.A.andSvensson,L.(Eds.),ActionandInteractiveResearch:BeyondPracticeandTheory,Maastricht,Netherlands:ShakerPublishing,pp.1‐12.

Tanner,K.(2002),"Surveyresearch", inWilliamson,K.(Ed.),Researchmethodsforstudents,academicsandprofessionals:Informationmanagementandsystems,2nd ed,WaggaWagga,NSW, Australia:Centre for Information Studies, pp.89‐109.

Teece, D.J., Pisano, G. and Shuen, A. (1997), "Dynamic capabilities and strategicmanagement",StrategicManagementJournal,Vol.18,No.7,pp.509‐533.

Terwiesch, C., Loch, C.H. and De Meyer, A. (2002), "Exchanging PreliminaryInformationinConcurrentEngineering:AlternativeCoordinationStrategies",OrganizationScience,Vol.13,No.4,pp.402‐419.

Thisse, L.C. (1998), "Advanced Quality Planning: A Guide for Any Organization",QualityProgress,Vol.31,No.2,pp.73‐77.

Thomke,S.andFujimoto,T.(2000),"TheEffectof'Front‐Loading'Problem‐Solvingon Product Development Performance", Journal of Product InnovationManagement,Vol.17,No.2,pp.128‐142.

Thompson, J.D. (1967), Organizations in Action: Social Science Bases ofAdministrativeTheory,NewYork,NY:McGraw‐Hill.

Trott, P. (2008), InnovationManagement andNew Product Development, 4th ed.,Harlow,UK:FTPrenticeHall.

Turkulainen, V. (2008), Managing Cross‐Functional Interdependencies – TheContingent Value of Integration, PhD Thesis, Helsinki University ofTechnology,Helsinki,Finnland.

Tushman,M.L.andScanlan,T.J.(1981),"BoundarySpanningIndividuals:TheirRolein Information Transfer and Their Antecedents", Academy ofManagementJournal,Vol.24,No.2,pp.289‐305.


105

Twigg, D. (2002), "Managing the design/manufacturing interface across firms",IntegratedManufacturingSystems,Vol.13,No.4,pp.212‐221.

Ulrich,K.T.andEppinger,S.D.(2007),ProductDesignandDevelopment,4thed.,NewYork,NY:McGraw‐Hill.

Utterback, J.M.(1994),MasteringtheDynamicsofInnovation,Boston,MA:HarvardBusinessSchoolPress.

VandeVen,A.H.andDrazin,R.(1985),"Theconceptof fit incontingencytheory",ResearchinOrganizationalBehavior,Vol.7,No.3,pp.333‐365.

Vandevelde,A.andVanDierdonck,R.(2003),"Managingthedesign‐manufacturinginterface",InternationalJournalofOperations&ProductionManagement,Vol.23,No.11,pp.1326‐1348.

Vincent, A. and Ross, D. (2001), "Personalize training: determine learning styles,personality types and multiple intelligences online", The LearningOrganization,Vol.8,No.1,pp.36‐43.

VonHaartman,R. andBengtsson,L. (2009), "Manufacturing competence: akey tosuccessful supplier integration", International Journal of ManufacturingTechnologyandManagement,Vol.16,No.3,pp.283‐299.

VonWeizsäcker,E.U.(1974),"ErstmaligkeitundBestätigungalsKomponentenderpragmatischenInformation",invonWeizsäcker,E.U.(Ed.),OffeneSystemeI:Beiträge zur Zeitstruktur von Information, Entropie und Evolution [inGerman],Stuttgart,Germany:ErnstKlettVerlag,pp.82‐113.

Voss, C., Tsikriktsis, N. and Frohlich, M. (2002), "Case research in operationsmanagement",InternationalJournalofOperations&ProductionManagement,Vol.22,No.2,pp.195‐219.

Weick, K.E. (1969), The Social Psychology of Organizing, Reading, MA: Addison‐Wesley.

Weick,K.E.(1995),SensemakinginOrganizations,ThousandOaks,CA:SAGE.Wheelwright, S.C. and Clark, K.B. (1992), Revolutionizing Product Development:

QuantumLeapsinSpeed,Efficiency,andQuality,NewYork,NY:FreePress.Wiendahl, H.P., ElMaraghy, H.A., Nyhuis, P., Zäh, M.F., Duffie, N. and Brieke, M.

(2007), "ChangeableManufacturing–Classification,DesignandOperation",CIRPAnnals–ManufacturingTechnology,Vol.56,No.2,pp.783‐809.

Wigren, C. and Brundin, E. (2008), "Interaktiv fältforskning – möjligheter ochomöjligheter", in Johannisson, B., Gunnarsson, E. and Stjernberg, T. (Eds.),Gemensamtkunskapande–deninteraktivaforskningenspraktik[inSwedish],Växjö:VäxjöUniversityPress,pp.95‐114.

Wu,B.(1994),ManufacturingSystemsDesignandAnalysis:Contextandtechniques,2nded.,London,UK:Chapman&Hall.

Yin,R.K.(2009),CaseStudyResearch:DesignandMethods,4thed.,ThousandOaks,CA:SAGE.

Zahay,D.,Griffin,A.andFredericks,E.(2004),"Sources,uses,andformsofdatainthe new product development process", IndustrialMarketingManagement,Vol.33,No.7,pp.657‐666.

Zahay, D., Griffin, A. and Fredericks, E. (2011), "Information Use in New ProductDevelopment:AnInitialExploratoryEmpiricalInvestigationintheChemicalIndustry",JournalofProductInnovationManagement,Vol.28,No.4,pp.485‐502.

MANAGEMENT OF DESIGN INFORMATION IN THE PRODUCTION …488464/FULLTEXT01.pdf · MANAGEMENT OF DESIGN...

Documents

Transcript of MANAGEMENT OF DESIGN INFORMATION IN THE PRODUCTION …488464/FULLTEXT01.pdf · MANAGEMENT OF DESIGN...