THE INFLUENCE OF DESIGN QUALITY ON PRODUCTIVITY IN CONSTRUCTION

THE INFLUENCE OF DESIGN QUALITY ON PRODUCTIVITY

IN CONSTRUCTION

ZHIGAO

FACULTY OF ENGINEERING MEMORIAL UNIVERSITY OF NEWFOUNDLAND

St. John'sCanada July1994

ABSTRACT

In many countries.the productivityin the designoffice has been a neglectedfield comparedwiththeoutpouringsof~searchand theoryon productivityinconseacdon.It may thusbeusefulto review the state of the an in construction research, as there arc parallels[0bedrawnwiththe design environment The author usedlenconstruction projectsto researchthe influenceof designquality on the productivityin consnuction. The ten projects werealldesigned by the author. Thedatawere collected after the constructionphaseand thedirectcostswereidentifiedwithrework(including redesign), repair. and replacement Finally the author gives fourequations for forecastingquality problemsrelatedtodesignin construction.The methodsof analysis usedarc statistical, optimization.simulationand fuzzy logic.Analysisof the data indicates ..natin construction thedeviations on projects accountedfor an average of 12.4%orlbe total project cost.Furthermore, design deviation average 78% of the total number of deviations, 79%of the total deviation C0515, and9.~%of the lotal projectcost.Since designcost is a small percentage of totalcosts, andan increaseindesignexpenditure can frequentlyreduce Iotallife cycle costs, it is important 10research jhe influenceof designqualityonproductiy.tyinconstruction.

ACKNOWLEDGMENTS

The author wishes to thank the Faculty of Engineering and Applied Science of Memorial University for affording him the opponunity of conducting this work.

Particularly, the supervision of Professor Dag A. FriisandlUs suggestion for an excitingresearch topic is greatly appreciated.The support and encouragementofDr.J.

J. Sharp.Associate Dean of the Faculty,Dr. Awad Hanna,the author'sformer supervisorare also acknowledged.

The author wishesfathankMr.Michael J. Hogan.Mr.Lain G.Smith. They gave me author manybeneficialideasand practicalsupportin the investigationanddata coUection.

Theauthorisgrateful10the FacultyofEngineeringfor financia1supponintheform of teaching assistantship and graduate support during his years of residenceat MemoriaJUniversity.

Finally,lhcpatience and underslBJlding of the author's wife and perentsduring the:

researchandstudyperiodis greatly appreciated.

ii

ABSTRACf ii

ACKNOWLEUG~NTS iii

CONTENTS iv

LIST OF TABLES vi

LISTOF FIGURES ix

CHAPTER 1.INTRODUCfION _ 1

1.1The Challenge ForTheCanadianConstruetionIndustty I t.2 ImprovingDesignQualityIsTheKeyTo Increasing

Construction Prodtlctivity 5

1.3Research Objecdve 9

1.4Data CollectionAnd ResearchMethodology 10

CHAPTER2.CONSTRUCTION MANAGEMENT ANDITSHISTORY._ 14

2.1Management 14

2.2.:HistoryOfConstnlctionManagement 14

CHAPTER3. PRODUCTIVITYMEASUREMENTS _ .14

3.1Introduction 24

3.2TheNeedToMeasure ConstructionProductivity 24 3.3ThreeMajorDifficultiesInMeasuringProductivity _ " 26

3.4 DifferentMeasurementOfProductivity 27

3.5WorkStudyModels 30

3.6 StatisticalMeasun::rnenl .31

3.7 UsingFuzzyLogicTo MeasureThe InfhlenceOfDesignInConstruction 33 CHAFfER 4.DESIGNPRODUCTIVITY.._.._.._ _ __ _ :•.38

4.1Introduction 38

4.2 DesignProcess 38

4.3DesignInputAndConstructionCOSr.: .40

4.4OptimumDesign ,42

4.5 DesignDecisions 47

CHAFfER5.CAUSEOFLOW PRODUCTIVITY

INCONSTRUCTION..._... .._....__...._...._.. .._._.. 49

5.1Introduction 49

5.2Deviation DataCOllocted 51

'.3Classifying DevjllrionData. .53

5.4Design Deviations ""., , ,53

5,,5ConstructionDcviation , ,5,5

5.6.AdministrationAndOwnersDeviation ,56

5.7FabricationDeviations , .56

S.8.TranspcrtanonDeviations, 56

S.9.OperationDevfations , , ,57

S.10AnalysisOfData. , , ~7

,5.11DesignDeviations IsTheMajorPartOfTotal Dcviations ,58

5.12 CostOfOcviations 61

5.13DeviationOfDistribution 63

CHAFfE R 6.FORECASTINGDESI GNAND CONSTRUCTION

DEVIATIONS 65

6.1Introduction 6,5

6.2Methodology 67

6.3LimitationsFoundInTheProjects 67

6.4Problem Categories. , 67

6.5Results 71

6.6 UsingFROB To ForecastThe DesignAndConstructionDeviation 77

6.7ForecastingConstructionDeviations And ConstructionMan-HoursUsing

Degree of Complexil)'of Construction .//7

CHAPTER 7.INCREASINGDESIGN PRODUCTIVITY I04

7.1Introduction .

7.2ManagementResponsibilities .

...104

. 105

7.3Optimization Design 106

7.4Communication AndFeedBackInDesignProcess I07

7.5 ComputerAidedDesign 111

7.6ControlOfChangesAndNonconformances l13

7.7Designer'sResponsibilityDuring Construction 113 7.8ResponsibilityOfConstructionManagementDuringDesign 114

CHAPTER 8 • CONCLUSIONSANDRECOMMENDATIONFOR

FURTHER RESEARCH _ _ 116

REFERENCES _ _ 120

List ofTables:

Table 1-1 RankingOfSelectedIndustriesAverageAnnualGrowth Rates 2 Tab le1·2 Construction Labor Productivity ComparisonOfFive COUntrics. ) Table1-3 CanadianLaborProductivityAndUnitLaborCoSlIndicatce.. . ,.4 Table 1-4 RangesOfPerceived Corporate Liabilities.... . 1

Table1-SRanges OfPerccivcdEmployee Liabilities , 7

Table 1-6ContractAdjustments By ClaimType 8

Table 1-7DescriptionsOfProjectsStudies... . 11

Table4-1 EffectOfExtta DesignInput, .41

Table5-1I>c:viation Distribution 50

Table5-2 DescriptionsOfProjectsStudies .52

Table5-3 NumberOfDeviationsAsPercentageOfTotalNumberOf~tions

OfEachProject 58

Table5-4 NumberOfDcsignDeviationsAsPercentageOfTotaJNumberOf

DeviationsOfEach Project. 00

Table5-5I>eviation CostsAs PercentageOfTotal ProjoctDeviations COsts....•...62 Table5-6AverageNumberOfDeviationsAs PercentageOfTotalNumberOf

DeviationsOf EachProject 63

Table6-1 EltampleOfCompositeProcedure. 80

Table6-2TheDefiniticnOf DegreeOfCcnplex 81

Table6-3 DescriptionsOfProjectStudics. 84

Table6-4TIleDegreeOfComplexityLiSI 88

Table6-5 The NumberOfDesignMan-HollnFor EachDiscipUne 90 Table6-6 Tbe Degreeof ComplexityofProjectsVC'l'$USDesignProblems

AndMan·HoUl'S 91

viii

ListofFigures

Fig I-I ProductivityImprovementOfDifferentIndustries(1961-1988) 3 Fig 1-2ComparisonOf Constr\lCtionProductivityand COSt 5

Fig.2-1 TraditionalService 16

Fig2-2Owner-BuilderScrvicc 18

Fig2-3Tum-KeyOrganization... . 20

Fig 2-4Professional Construction Managernent 22

Fig2-5Project Participants 23

Fig3-1FuzzyEstimateOfith BasicCriterion 35

Fig 3-2 Transferring ActualValueZUI(;(JInto IndexValue51.h(xJ 36

Fig 4-tDesign InProjectProcess 40

Fig 4-2EffectOfDesign eft'onOnProjectCost, 42

Fig4-3 A BuildingCostAnalysis 43

Fig 44Design DecisionAnd ConstructionCOSl 48

Fig5-1 DeviationDistributionByCostOfProject. .50 Fig5-2 NumberOfDeviationsAsPercentage OfTotalNumber

or

Deviations 59 Fig 5-3Design DeviationsAsPercentage

or

Total NumberOfDeviations 61

Fig5-4TheAverageOfTotalProjectDeviationCOSts 62

Fig5-6 AverageNumber

or

DeviationAsPercentageOfTotal Number

OfDeviadons. 64

Fig 6-1Example

or

Profile Curves 12

Fig 6-2SteelStructureDiscipline 74

Fig 6-3ConcreteStructureDiscipline 75

ix

Fig6-4Aal-Platc Sttueture Disciplinc...•.•.••._ M 76

Fig6-S Masonry Structure Disciplinc _ _ M M77

Fig6-6DesignDegree Of'ComplexityVen usDesignDeviation

OrMan-Hours..•.M M _•• •MMM••M MM M . .M _ _ _ 96 Fig6-7Construe:tionDegreeOfComplexity VersusConstructionDeviation

OrMan·Hours _._ _..M _ _ MMM_ _•• _ 102 Fig7-1BasicCommunicationSysrenInConstruction lndustry..M._ I09 Fig 7-2 Decision Impacts AsSeen By ARC. _•.•._ _ 112

CHAPTER 1. INTRODUCTION

1.1The Challenge forthe CanadianConstructionIndl1S1ry:

InCeneda,the COfIS!IUCDon industryisahigh proportion of totalrevenueincome,

Itisroughly8.3%of GNP.$50bnllon-pfusshareof theCanadian gross national product,(StatisticsCanada.1988).Thiswasalsotrueinthe Atlanticprovinces andin Ontario.In Quebec.plant processdesignwasas imponant as building projectSand,in British Columbia themostimportantsourceof revenuecame fromprojects relatedtc agriculture.fisheriesand forestry.Inthe prairieprovinces. thedominantsourceof revenue was fromoil,petroleumandnaneal gasprojectsIII36%.

The constructio nindustry is the largest industry butthevast majority ofits hundreds of participantsaresmallbustaesses. Acam:lingtoStatisticsCanadafigures for 1988.over90%ofthe110,000construction companiesinCanada have twentyor fewer employees.

Indesign,of the 2513 architecturaldesignfirmsinCanada.,only 176(8.17%) earned more than$1million.The averagefumin1988earneda fee income of S263,600 and had5 employees.Forty-sevenpercent of therevenuesware earned bythe largest sixtyfirms(J~ofthetotalnumberoffirms).(StatisticsCanada.1988).

Inrecentyears,Canadianconstructionindustryha~faced manychallenges.There are now,andwillcontinueto be, shortages ofresources,includingmaterials,equipment.

skilledworkers,and technicaland supervisorystaff.At the projectlevel, managementbas justbeguntointegratedesign, procurement.and COOS1l'UCtion intoone toralprocess.Then:

willbemore and more governmentalregulationon the safetyof design and on field

constructionmethods,environmentalconsequences ofprojccts,and personnelpoliciesat alllevels. Managementmustalso copewith neweconomicandculturalJUlities resulting from inflation,energy shortages,changingworlddevelopmentpatterns,andnew societal standards, Butthegreatestproblemisthat consuuctionproductivity improvementisvery slow.

Canadianconstruction productivity hasgrownrelativelyslowly.StatisticsCanada reponed(l99I)that from196 1101988,accordingtogrossoutput multifactored productivity,averageannualgrowthratesofconstruc tion productivityis0.5% only.See Table1·1andFig·t.l.

TableI·!

RankinlllohelectedlDdllStl'lcs anralllC aDDLI,1~WC'"rates(1961.1983)

Industries

"""" ....

Telocommunica1ioos asc

EVctrical&E1ec:lrOnieProd 1.10

TJ'IMPOfWionInduslries

I."

PlasticProducts 1.40

Wbolesaie& RetailTrade 1.3'

TIaIISJ)OI1aIionEquip 1.30

COllStructioa O.so

Food .40

SOUTce: StatisticsCanodtJ 1990·1991,CataloglU!12·204E

Fig1~1ProdUdivity Improvement or DitrerentIndustries

(Rankingor Seleded Indwtries According To Gross Output Multlractor Productivity. AverageAnd GrowthRates)

(1961.1988)

0.00 0.50 1.00 l.so 200 150 3.00 3.50 4.00 Growth PeruDtage

Comparedwith other developed countries.Canadianconstruction productivity growsslowly.See Table1·2:

Tlble 1·2 COILStrvl:doaLlIMwProductlvllJ Comparlso.ofF1ve Coulltries

COUNTJUES PROIlUCITVTlY

I" I'

r.... 19

V.K. 2.9

SinQ!IIYW\: 4.0

""""

^1.1

Source:J.K.YatesIfllunalional Labor PToducriviry«CostEngineering»Vol.3S/No.1 Jan.•1993

From 1986 to1990, Canadian construc tionproductivitydeclined3.7%.but consttuetionlaborunitcostincreased34.1%,secTable 1·3,andFie1·2

Table1·3

Canadianlaborprodudivil1 andunitlabor C05tIndicaton I .

orconstrudonmdustries l.Ibor ProdlJttiri^ullleal GDD'l

"""--

^Unit

"'- ^"' ... ^"'_ ... """''''''

'980 85-'" 83.30 "JO 83JO

""

^93.70

^92.,. . ... ,

^91.40

'982 100.10 ,,,.00 9'.30 83JO

'983 ^{101.90 104.-'0 91.'lQ} 83JO

'984 97-'" 98.30 93." ^9l.oo

'98'

^{97.., 96.10 92.10 9lJO}

'936

^{'00.00 100.00 100.00 100.00}

1987 ...90 96,. 107.40 1ll.2D

'988

....

^{94.10 113.60 '>lJO}

'989 97.10 94.00 111.40

'29'"

1990 96.30 95.40 111.90 134.10

(Source:StaJisdcsCanada1991 )

4

Flg1·1Compariso n orConstructionProductivityand Cost

:::f:+J:::[:F:fmj

..JC. ' .~""r : : , • ,

::!::I -l:Flrr:I::I::1

·· ··i· · ·· i··-·l· --·~-_ ·_~-_··-:--···}--··f··-·l···· i

: : : : : : ! : : .!

140.00

~12 0 . 00

j

_100.00

e

.. . _"

⁸⁰_60.00^.00

E

~

^40.00_20.00

0.00

0

..

.. ^.. .. ^'" ....

Yelr

o

..

..

1-..-

Labour ProduCllvl1y.- - . . 6 - -Unll labourcost

1.2 Improvi ngDesigl1Qualify IsThe Key To Increasing Construdlon Productivity

Itis beyondthe scopeoftlJisthesistodescribe Cunhcrthemanyproblemsfacing theCanadianconstructionindustry.From thewriter's investigationof some construCtion sitesandsomedesignfinns,we usuallyseclhcsituationthat labor blamethe conttaefOn;

contractorsblamelaborand designers;andtheownersblamethemallToooften.laboris madethe scapegoatforpoordesignqualityandpoorconstruction DWIagcmcnt

The reason for lowconstructionand design productivity isthe comparatively largenumbersand smallsizesof its businesses. its fragmentationanddivisiveness,andits service characceristics. During economic recession,many construction companies, especially design offices have verylimitedtime and resources available toallowthem to study thereamsof productivity data available,and develop a productivity program for their company.Itis surprising that productivity improvements have not occurredinthe Canadian construction industry.The vast majority of people involved in the industry do not mow about the information whichis available.This can also explain the lack of funding for productivity research in the construction industry compared with thatofthe manufacturing indusny,(Pric:e and Harris, 1985).

ASCE's (American Society of Civil Engineers) Hazardous Waste Liability Committee recently completed asurvey of civil engineering fmns practicing inthe hazardous waste arena. The intent oCthe survey was to provide infonnation on perceived liabilitiesinthis field,andmanagement tools to minimize those potential liabilities. The survey shows that about 30% of peccle think that Perceived Employee Liabilities depends highly on design.and70% on construction management (Table 1-4and Table:I~ 5). Furthermore, Diekmann (1985) investigated 427 construction projects, hepointed out: that the overall additive claim rate was 6%(i.e.six cents on the dollar) and , moreover, 72% of these increases were due to design error or owner initilled changes.

From Yates'sinvestigation, the design deviation cost is three timeshigher than construction costinmany countries. (J.K.Yatese.al,1993). Therefore.to improve the quality of the designisthe significant problem in the construction industry.

Table1-4

Range!orPerceived CorporaleLlabilltiet Pm:eivod Uabilityas Pettentage or TWLnicaI caleglXy ResllOl1dents

v low lo_ A Hih

RegulaotwylnlCrJKllllioo 60 30 0 10

FieldAsse:ssment 30

'"

^{so 0}

Monitoring 30 30 30 IO

LabonltoryAnal)'sis

.,

'" "

⁰

Penniuing 2S 2S so ⁰

PR:UminaryDesign 10 10 70 10

FInalDe.~ign '0 30 zo 40

ConstnlctionM"

'''''" "

⁰

"

⁷⁰

(SourtC: WayneTusa.1985)

Table I·S Rangn orPerceived EmployeeLiabililiet

Pm:eivedUabiliryas Pen:ent1gc or

Typeofpersonnel R

"' ""

V_low lo_ A _ Hi b

Field so zo

'"

¹⁰

...

^{30 30 40 0}

""'''"

^{10 0 611 30}

C'onstruelionMana

m'"

^{0 0 0 100}

(SOURCE: WayneTusa.(985)

Ontheother hand.when ownersdiscovered that design-eonstruction and negotiatedcontractsinvarious fcrrns couldsignificantly reduce project duration,they intensifiedpressureson contraCtorstoget(acilitic.sintoproductionoroccupancy atlhc earliest possibleDlOmcntto maximize returns on investedcapital.Construction was increasingly programmedtoproceed simultaneouslywithdesign in lheinduslriaJand buildingfields.FromTable1-6wecanseethaiin pureadditive claims.dc5isnerrors accountfor39% oftotalclaims.Inpure deductive claims.ValueEngineering(mainly depends on design)accounts for 63%o(the deductions.(ValueEngineeringis • systematicapproachtoobtainingoptimum valuefor everydollars7'.nt).

TABLEI~Contract Adj ustme nts by ClaimType

PlIIlEADDmvECLAIMS PURE DEDUCTIVE

CLAIMS

T,,","""'" CIoimo ClmIpenll.lioolT_~lle"'i_ Cl....

...

ewm.".

N ._

- ' ^N_. - ' Do'~ - ' Do,. - ' ^{N_ .}.~ Don~.~

.... ... '000". .. ^'000"•

[),..;-- _{'" " '"} ..

^2An

..

²⁹⁰

_{" "}

¹⁹

_{" "}

Ch...~~tiotI

" " .. _{" "" " '"} • ., ,.

^Il

^,

"' ... -

^5l

_" .. _{" '"}

^1\

_" ^, .. " " "

Dift"erinI·iu.

.. _" ..

^{15 712}

_{" '"} . • · ^{, ,}

c:andiuon.

w... 19

,

" • ^, • '" "

^e

^• , ,

Vl1ueen

... _" . • • ^{, , , ,}

¹⁵

"

^2"

"

.trih

,

1

, ,

e

,

.00

" ^, · ^{, ,}

Olllc:r 19

• ,

1 12.02 19

, , , _{• • •}

T...

", " .

^{m reo6130 '00 un}

". ..

^'00

_'"

^'00

SOIInCfrom:

J _ ,

E.D~kNwt198$

Historicallyspeaking,productivityof employeesis a majorconce rn in all constructionorganizations.The optimization of the production of thetotalorganization canbeachievedlhroughthecoordinatedimprovementof the performance oftheindividual employees.For many years the productivityof blue collar workers in the construction industryhasbeenstudied and evaluated.However,severalstudies haverevealed thatpoor management was thecausefor poor worker productivity. Thisindicates that attentionis also nceded to improvetheproductivityof "white cellar" employees in construction.

Improvingdesignqualityis theprocessof increasingdesignproductivity.

Sometimes it canbethought of as a designmanagement problem. It shouldbewidely includedintheconstructionmanagem entfieldbothin Canadaandadler countries . Productivityinthedesignofficehas been a neglectedfieldcomparedwith theoutpourings of researchon productivityinconstruction or manufacturing activities.Soitmaybe useful10reviewthestateoftheartinconstructionresearch.Thereareparallelsthatcanbe drawnwith the designenvironment. What needs tobesaidat the outset, however,is that productivitygainsinthe design officewillnot come as caslly as those on the ccnseacdce site. Thisis because production of engineeringdesignsconsist of a large number of complex,vagueideas, interdependent tasks,which areDot easil y measured and flowclwtcd.

1.3 ResearchObjective

The research presentedinthisthesiswas conducted toidentify the causes and magnitude of quality problemsindesign and constructionandto determinethecosts associ atedwiththe quality problems.The author will usestatisticsand a fuzzy logical method to analyzetheinfluence of designduring construction.. Thedegreeof complexity is used toconsider some problemsof design and ccnseuctlce, Ftna1ly,the authorwill give four regression equations for forecasting designandconstruction deviatioos and

m.an-houn inpracdce,Theresearchwas conducledunder the guidanceoftheCanadian ConstrUCtion Codeand0LineseConstrUCtion Code..

1.'1nataColl«tlonand Rest armMet hodology:

Inordertoquanruysomeof thefacKln rhat affectdesign andconseuceon productivity,theproductivitydatawascollectedbythisauther.IIincludes[Weparts.The firstpan are tenrealconstructionprojecrs. TI1eauthor designed theseinthe paS(ten yearsin ChinLMost of thepracticalproblemsweredirecrly faced bytheauthor.The second partis.inSt.John's Newfoundland.Theauthorvisitedsomedesignrums,and visited twoconstructionsiteson a dailybasis andcollectedboth quanritativeand qualitative data.Thedatawere collected over two periods. oneis60days (C·Core Laboratory). anotherIs 90days (OenerailfospitaICancer Center).Differentcountrics' constructioncodesarc diffemlL However.theprinciplesCIawhichthey8Rbasedare very similar.Because ofthedL'fercncesinconstructioncode and availabilityof completion date,justthe followinglenprojects(fable1-7)are.usedinsubsequentanalysis.

Thcnameof the projcctsof the Table1-7isasfoIlowing:

A.BeijingPeople'sBrcU:astingStation.

B.HuabeiHospital

C.l.paneseCANONF"""'YWor1cshop D BeijingNormalUnivcnil)' classroomBuilding E.ChineseInternationalInvestmentCompanyBuilding FJapnneseTOSHIBAFICI01')'

O.BeijingFJcventhAsial'tGamesSportsman House H.Oaojaiayun Disaict

I.ZhongShanBuilding I.linanSupermarket

TABLE1-7,Descrlntions ofPreiects Stu dies PROJEcr TYPE OF TYPEOF TOTALINSTAUB>

STRUCTURE US. PROJEcrCOSTlVuB

A ConcrelcFI'8III$ lOCHNJCAL 9O,0Cl0,ooo

•

"""",F_ HOSPITAL 20,00:>,000

C

""""''-

^WORKSHOP

^sorooroo

0 SteelSltUCt1ue RBSIDBNIlAL 12.soo,ooo

•

SteelSlrUCture OffiCE 234.000,00o

F Stee1Suuctun: FACTORY 19,COO,OOO

0 MasonrySll'UCture RBSIDBNI1AL 1,soo,l)'JO

H

"""",,,'''''''''''

RBSIDBNI1AL 10,900,000

I

"""",,,'''''''''''

^Horn. 6J;,coo,OOO

1

""".-

^COMMERCIAL 16 QfY'1000

{Note :100 CanadianDollars

=

635RMB, in 1994eJ:change nie}

Some infonnalionwas also obtained from interviewswithcontractorsDot associatedwiththestudy sites.Themainfocus ofthedata coUectionwasaDthe measurementofwork complctedby thedesignandconstructionincludingconcrete frames, steelsrrucrures, masonrystructures,plusmasonrycrew~thefactorswhich affectedtheirproductivity.Theobserv er completed asimpledata sheeteverydaywhich servedas aguideforthe data collectionprocess. Some ofthe environme ntal data

II

included temperature, windspeed. humidity,and precipitationreadings.The weather informationrecordedon sitewas supplemented withweatherdata publishedfromtbe St.

John's weatherofficcfor theregion.(The infonnationof thetenChineseprojectsdidnot includeweatherconditions).Therewere measurementsof the amount of block and brick work completedduring the day as well as the crew sizesandwodcinghours.The size and type of materials used were considered importanttothestudyandwerealso collected.

Anyabsentees,overtime,oraccidentswere recorded.The qualitativedatacamefromthe observersown observationsof the work,frominformaldiscussions withthecrews,and from interviewswiththe site foremen.Thepurpose of obtaining qualitative data was to identify any significantproblems,delays,intel'TU ptions.or disruptionswhichoccurred duringthe shift.Adisruption is defined as"aneventoccurringon-sitethat adversely affect\:the crewsproductivityfor most oftileworkday"(SandCtSandTomas,J99I).

Theobservercould not stayonthe construction site throughouttheentire shift which meantthatmuchof tile informationregardingdelays, anddailyprogress, dependtd upon informaldiscussionswithsite pet'SOMel.Theinformation obtainedoften depended onwhotheobserver interviewedand on theirevaluationofthe incident.Sometimesdetails ofthe incidentsmayhavebeenomitted.Thequalityofinformationimproved.aspersona1 rapport withsitepersonnelimproved.Cooperation from sitemanagementwas essential for thedata collectionprocess. Hence.creating agoodwortingrelation withthemWlU imperative.Tomaintainl;is relation.itwasimportantfortheobservertodirect inquires10 managerocntattimeswhentheywerenotbusywithjobmtatod activities. When questions were heldfortesabusytimestheinformationreceivedwaslDOItlin depthand weD thought out.Themanagementpersonnelfromthegeneralcontractorwerequitewillingtohelpin

12

the study, and eager despite some initialskepticism. The crews, subcontractors,and foremenalsocooperand.

The objectiveo(thedatacollectionwastoidentifyandquantifyas many (actorsas possible thataffecteddesignandconstruction productivity.The reason for this is that there are so manyfactorSdlatsimultanco uslyaffectproductivity,itisalmostimpossible10 unravelthem anddetermine their individual effectsduring a st-sdyof thissize.Toeven comeclosetocreating amodel whichcanaccuratelypredict theproductivity thatwillbe obtained duringa shift, wouldrequireconsiderableadditionalresearch.It wasdeemed more beneficialfor this study tottyand makeeveryoneintheindustrymore aware of the major causesofprodlOClivity lossesduring designandconstructionprojeccs,especialIyche relationshipbetweenthem.Thiswayone canttytotake steps 10 eliminate thenegative factors.

It wasnoC possible 10gelsuffici ent data on the design retered aspects ofthetwo Canadian projeclS10get a quantitativeindicationof the diff'Cttllces betweenCanadianand Chineseconstructionanddesignenvironments.TheCanadian data hastherefore been omittedsince itwillnotaddsufficient informationtogive any additionalinsightinlOtbe problem areastudied.Itismentionedhere,only becauseconsiderabletimeandeffort was spentintrying 10get enoughinformationto make8 meaningful compariSOll between CanadianandChinesepractice.Unfonunatelythis effort failed.

Inquantitative analysis,the author used fuzzy logic,optimization theoryand statisticalmethods.

13

CHAPTER 2. THE CONSTRUCTION MANAGEMENT AND HISTORY

Before studying the influence of design in construction, we should know something about construction management andits history,Management is critical in improving productivityinboththe design and constructionprocesses.

2.1Management

Usuallywedefinemanagement as the use of people and otherreso urces 10 accomplish objectives.Thisdefinition is applicable toallorganizationalstructures. both profit-oriented and not-fer-profit, (Boone and Kurtz 1992)

Management involves the creation of an environmentin which people can most effectively use other resources to reach Slated goals.It involves the implementationof four basic functions:planning,organizing,leading, and controlling.These functions play a roleinthe operadoncrall organizations.

Construction is the process whereby the designer's plans and specifications are convened into physical structures and facilities.Itinvolves the organization and coordination ofallthe resources for the project (labor, construction equipment, permanent andtemporarymaterials, supplies and utilities,money,technology and methods, andt:l.tne)tocomplete the project on schedule,withinthe budget, and acr..ording tothestandards ofquilityandperformancespecifiedbythedesigner.

2.2ConstructionManagemp.iltHistory

Historicallyspeaking,construction managementandproductivity ofthe construction site(or the factory floor) hasbeena focus of concern sinceat leastthetime

14

ofF.W.Taylor.whenhistheoriesof "Scientific Management"fIrstdrewattentiontothe fact that there were better ways of doing many things than thetraditional way.Indeed,it was Taylor'scontentionthat no fieldof activity,no matterhow simple. could notbe improvedby the applicationofscientificmanagement. andtosupport this viewheused his famousexampleofthepig-iron handler.So weUknowndidthis examplebeccce,that Taylorwas later to lamentthat "peopleseem to think that the wholeof scientific managementconsists inhandlingpig-iron"(Taylor1947).As isthe fate of many theories, Taylor iscurrentlyunfashionable.anditis nowadayscommonto pour scornon scientificmanagement.There are good reasonsfor this.Taylorappearsto have hadlittle respect for the inteUigence or humanity of his workers.and treated them as purely productionunits.But, it shouldnot be forgottenthat scientificmanagementcan lay claim 10somesignificant achievements.Its methodologylivesonunder thegeneral titleof

"methodsimprovement"studies. The modem version is more acceptablein!.halitalso takesaccountof the fact that tasksare carried out bypeople, whohave theirownneeds andare motivated (or good perfonnanceindifferentways.This canbeamibulCdtothe work of the"Human Relations"school ofmanagemen1lheories.Thework ofthisscbool (Mayo1949; McGregor1960j Herzberg1966) also owesadebt to scientific management,inthatitarosechieflytooppose the perceivedheartlessnessof scientific management

Thep~trealities ofconstructionmanagementwillnow beexaminedas wellas the realitiesof today. The following descriptionsfroma designerviewpoint, outline majordifferencesamongapproachesandexploresomeoftbeirvariations. advanlages.and disadvanlages. as wellastheirsimiJarities.

15

(i).TraditionalservicesInlhe19505

Here the owner employsa designer(architect.architect/engineer.or engineer)who firstpreparesthe plansandspecifications. then exercisessomedegreeof inspection.

moniloring•or controlduringconstruction.Constructionitself istheresponsibilityofa singlegeneralcontractorundercontract to the owner.Much of fhework mayactuallybe performedbyindividualtrade constructorsundersubcontract 10 the generalcontractor,

TRADITlONAL

1. Separate designer 2.Single general contractor 3. NumeroussubcontraelOfS

4.Fixedprice.unit price. guaranteedmaximum.

or cost plus a fixed fee constructioncontract 5.Negotiated professionalfeefor design services

Fig1.1 Trad itional Service

The traditionalapproach was a sequential,linear flow ofpreliminarydesign, consttuctiondocurcene,andsupervisionof consttuction.Codes and zoninglawswerenot too complexfor the moslpan.and the time for zoningandother public approvalsof a project was usually predictable.routine. andshort. Relations between architect/engineer

and clientandarchitect/engineerandcontractorwere stable,and the ideaof errors-and- omissionsinsurance bad not yet been born. Construction costs couldbereasonably projected using thecost-per-square-foot method.Cost consultants were fewand far between.Asthe pace ofconstructionquickened and the volume and complexit)-of buildingsgrew, theold traditional waysofscrvicedidnotfit the needsof thetimes.

(Z).ComprehensiveStrvice5

or

the 1960's(The Owner Builder)

Inthedecadeconstructionprojects becamelarger. Large projects became multibuilding projects. New towns were designed and built, city planning grewinto regionalplanning.More engineeringsystemsmade buildingsmorecomplexand difficult tosuperviseduring construction.The standatd form of architects/engineers agreements was changed; architects/engineersobservedconstruction, they no longer supervisedit.

Errors·and-omissionsinsurancewas invented,and with this protectioncameexposure, and the number of suits expanded again. Relations between architects/engineersand contractorswere notas wellcoonfutatedas they had been during the fifties.Withthe growth it followed thattimerequiredfor receipt of public approvals•....as longerinthe sixtiesthanin thefifties.

Therefore, many city,and county public works departments and private companies have performedboth theirown design work and some or allof the actual constructionwith theirown forcestomeetthischallenge.Thisapproachis often referred toas"forceeccoum,"Otherowners (orowners'representatives),whileretainingmanyof themanagementandconceptualdesiguresponsibilities, haveutiliz.cd consultantsfIXsome orallof the detailed design. and havedepended upon construction contractorsf«the actwol hiringandsupervisionof the laborforce.

17

Owner-Builder

1.Ownerresponsible fordesignand construction 2.~~~forceswcek coneactcrs and 3.FiXedprice,unit price,ornegotiated

constructioncontract

Fig 2-1 Owner - Builder Servi«

Owner-buijdershaveutilized manyof Ihe'contractualformsdiscussedabovefor Ihe traditional approach.and theyan:increasingly moving 10professionalconstruction managementmethods.Actually,theowner-buildercanbelikened to the design- constructor, except that the ultimateproductisutilizedin-houserather thandevelopedfor an outsideowner.Manyof theowner-buildershave developeddcsign-conslrUetdivisions that areof asizecomparablewiththose ofmany of lheIargcrturn-keybui1dcn.Howeva. itappearsthat thismethodof workis relativelylnrgeandrelativelyconstantover.long periodof time, and where project managementcanbeseparated fromoperational management

The owner-builder can employ all the techniques of the design-contractor,the professionalconstruction manager, and the traditionalapproach. Howeverat present the adventages of thistypeof approachare bestsuitedtoa~Iativelyfew,favorable situated companiesor agenciesnow.

(3).FllSt-TrackService in 1970's.[Dcsign.col\SlruClorDesign-Managc(Tum·key») Thefast-trackmethodis organized toreduce the time neededfor construction, makingpossibleearlieroccupancy andreducing fmancingcosts.Infast-tracie,as soon as the schematic design is acceptedand public approvalsan:received, the construction manager(whosometimesis the m:hitcct)breaksthe projectdownintoaseriesofpbased bid packages.Workinthe field can be startediromediarely,and materialsand equipment requiringlong leadtimeare ordered. Fmal designand documentarion,for separatebilled packages,continue until the buildingiscompleted

During this decade a swarm of package builders and others invadedthedesign professional'sfield and offero..d owners a design-buildpackage.Architectsandenginocn countered by getting into the developmentbusiness themselves.Tbey organized joint venturesandconsoniums in any number of waysthatmightbeattractive 10 clients.

Design-buildusesthe fast track methodofconstruction,butitdoesso at a fixedprice and with.singleresponsibility.Altbiteel$ supplySttVicestothe point of single~nsibility and ROC totheowner,and~Icareisneeded10avoidconflictsof interest.

Someauthoritiesdifferentiatebetween"design-consmet" and"turnkey."General usage, however.treatsthem inlCt'Changeable.In thismethod,allphasesofaproject,from conceptthroughdesignandconstructionarchandledbythe same organizadon.(Fig2-3)

19

Design-Build

1. Slng'-linn responsiM lorboth designandconslruc:tton 2.Special\)'subcontraclOl"l 3.Fixedprice,gUaratlleedmaximumprICe,

orcoafplua afeede4ign-eonsUUCllon

.",."'"

1.Singlefirm respolllliblelor both designandcon8!NCllon 2.Fixedprioe,Ofnegoll.aledlndMduaJ

COl1.lnJeIIonoonlnl~orsubcon\l'lletl 3. Filledprice.guaranlMd maximumFJrio-,

orcoatplus IIIleede8!gn-eon.1nJcl1on

"."."'"

Fig 2-3 Tum-key Organization

Inthecase of design-consllUClor,theconstructoracts as a genen1 contractorwith single-firm.control ofallsubcontractors.Usually,butnotalways. there is some formof negotiated contractbetweendesign-construetor and owner,Inthecase ofdesign-manage, constructionispremedbya numberof independent contractorsina mannersimilartothe professiooalconstruction managementconcept.Undereitherdeslgn-conseactor design- manage,cons1lV::tion can readilybeperformedunder a phasedconsttu<:1ion programto minimize projcctduration.'Thisform of completing projects has teen usedfocthe majorityof process:-oriented heavy industrialprojectsconstructedinthe United Statesill the lastfewdecade. ReferencetoEngineeringNews-Record',(ENR) annual

wI

ofthe

20

.500largesldesigners show thai IJ..;design-constructorsare heavily representedinthe lOp20.

(4).Impart analJsis and dtsign servtcesin the 1980s.

(ProCessionalConstructionManagement)

Inthe eighties.allpanemsoCconstructionmanagementcontinue10beused.The traditionallinear Corm continue in useCorsmallerprojects.The trend is clear,however.

New methodshasbeen foundtoshonen constructiontime,andincreasetheproductivity oCdesignerandconstructor.ThatisProfessional constructionmanagement.

Professional construction managementinvolvesa three-party team of owner, designerandconstruction managerin anon-adversaryrelationship,and it provides the ownerwithanopponunitytoparticipate fullyinthe construction process. Its success dependsupon eliminationof adversarialrelationshipsamongteammembers.Shouldone ormoreof the team members introduce concepts or policies delrimentaltonaturally satisfactotyrelati~ships,the concept detericretesinto an adversarial sinratieu,with inevitablenegativeeffeclSuponboththeprojectand individualpanicipanlS.Barrie gaveI clear picturefor Professionalconstructionmanagement(Barrie 1992),see Fig2-4.

21

PROFESSIONAL CONSTRUCnONMANAGER

Genen.l ContrEtor ConstructionManager

1.1bfee-pa1)'lUrlItI(o_.~dai&na'.""

F'lCI'-IDllIIIIIaar Ittina: U I~IMMicr 2..filedprioeorllqOCilllo:lindo:pon&:rllf\Ib<:QI!l'aa<n 3.ConttN:tioftlllMAJft'''''''O'~• .,_bowna' 4.NtlJoti&ledpRIft:UiondfeefwccnlrCVCion rna\l& <maIl ...--widr. _llIimbon<montf«~

S.Ne,otWaI pMUliaftal lMadaip MrlP.

1.1lne-~IUmDfo_.doaipa, and COI'IIlNClionIIIonAp.

2.F'wdp;e..ClfIlt.OIiuo:I ~_Klian _a-K1IdncdylO'id!o_ . tConacnoo:tion - ...,...y.s 1llll0lII'neI"11lIt:nI1ll

nlftlllcldooalllfd.

4.NqocialcdplllfUlilllnalr. b __1.

_ .

lrCUlioIIlIUMC_

~pofaa.-lr.r.. dIltip-"-

Fig 14 ProfessionalConstruc tionMlnagt rntnl (DonaldS.Barr ie .1991)

Obviou sly, constructionprojectsarcincreasingincomplexity.Fillyyears ago, I personconceived his project.designedit,andbuiltithimself. We have progressed throughthe evolutionary stageofmasterbuilderstothepointthat we now have an ind ustry ofspecialists.A given projectis dependentuponnumerouspartics. including owners,designers. rananciers , consultants .accountants,anomey. ,cons D'Uctonand

govemmenlagencies(see Fig.2-5).Itis not surprising thai coordinationofallthe participants is a challengetothemoslcompetentmanaga.

Contractors Insurance Consultants Suppliers Vendors Archllec1s Engineers Union.

Anorneys

GeneralContractor

Fig2-05Project Participant (RichardL.Tucker, 1986)

23

CHAPTER 3. PRODUCTIVITY MEASUREMENT

3.1Introduction

Duetothe fact that misconceptions concerning productivity abound.itis necessarytorestate theimportance oftheconcept. Productivitygrowthisdirecdy linked withincreasingrealwealthofanation.ThUs.ifa genen.l.increase inpayisawarded,which isnot accompaniedby a commensurateincreaseinproductivity,theresultwillrnm:lybea corresponding adjustmentin prices of thosegoods and scrvices(i.c.• inflation).Itis not easy torigorousJy prove the relationship,ina particularindusay,orin the economyas a whole, but the concept accords with common sense, and few would thus dispute lIlat Increased outputwithina particularindustrywillbenefitat leastthe peopleconnectedwith that industry.

3.2 The Need For Measuring Deslgr. And Construction ProductivUy Themaininput resourcesindesignandconstructionan: labor (Including white collar andblue conar), materialsand equipmentThe cost of these resourcesform the direct costof every project However. materialcostunlikethatof labor and planaare usuallyoutsidetheconaoJ ofthecontractor.Moreover.laborand plant costs are unstable andvary withinthe limilSof their control.Labor,for example,is the only inputrcsourcc whose cost canbecompletely controlled on site. The greeter the control the lower the COSLControlcan achieve its aim only whenitis within its limits.These limits canbe known when the maximum or minimum utilization level and plant are known through productivitymeasurements.

Inrecent years, design and construction jobs have become vcryccmpeudve.

Clients want thebest service at the minimumcost, Contractorswho arc able to pucin thc

24

Ioweslbid,usuallywinthejob.Tobeabletobidlow,theestimatermustknowtheleve10f efficiency or such cose sensitiveresoerceslike whethertheavailable equipmentcan eccipleethejobwiminthisperiod.Thesecanbeknownby knowingtheirperformance leveltluooghproductivicymeasun:mentson previousjobs.

Laberispaidaccording to theconttibutioncoproductivity.This means

mat

every workermUSIbepaidfairlytoreeiproceea fairday'swork.The meaningofa fair day's work:isrelativeasfar astheworker or managemenlisconcerned. This controvmycan onlyberesolvedbya pre-determincdswwWd satisfactotyto bothmanagement andlabor throughknownfactsprovidedby measuringthepnxluctivhyotlabor,Managementneeds produc:tivilylevels as a basisof labor costcontrol

To motivlUCworkers,managcmenc institutesincentivcschemes.Management uses the resultsofproductivitymeasurements asabasisforthe paymentof wage incentive10 direc:tlabor.

Modemconsuuctionis·very complex,chCRare timeson lbesitewhen cenain.

activi tieswiDbecoinson withwhichthecontrae1orhasDOpreviousexperieeee, Meanwhilethe valueofthe produas refertothewbole produaiocprecess,notonlythe lastorfinalsection.Sodesign.transportation,procurementandmanagementetc.should be included.Hencetheindirectoost anddirecc:coscshouJdbccakulatedatthesamelime.

IIis,cherdore,veryimportant10measuretheproductivityof laborandplant involvedin suchan activitybothasa costccnecteeesure and asatllttoricaJrecord.

A foreman,or superinlendentonsiee,or design office staff,maynotice an operationthacdoes DOCprogress atanacceptablerate,this canbestbeinvestigated by productivilyanalysis.

3.3. Three Major Difficulties In MeasuringProductivity

Sometimes,formeasurementpurposes,productivityis looselydefrnedintennsof the ccrput of goodsor services for a given unitamount of resourceinput capital.labor, knowledge, and materials.Thereare three major difficulties which seriously limit the usefulnessof suchmeasurements.

First. a single productivity measurementis not very useful byitself, and canbe used onlyfor comparisonswith the past or with other producers,i.e.,measurementsare indicatorsof relativeefficiency(NationalResearchCouncil1979).A companyrecording a productivitymeasurementcomparable withthe rest of the industryonly leams that it is as efficient (or inefficient) aseveryoneelse.Inordertoobtainan efficiencymeasurement inabsolutetermsitis necessaryto have a theoreticalmaximum withwhichtocompSR.

ThisIs neteasy10come byinthe contextofthe design office. One would hesitateto hazarda guessatthe absolute: efficiencylevelsprevailinginthe averagedesignoffioc,but figuresavailable suggest that on the construction sites there is still a long waytogo.

Tucker(1986)quotesfiguresfroma P"...,ject indicating that only20%ofman·houn were used effectivelyinputtingtheprojecttogether.

The second difficultywithproductivity measurementsisthatitis difficult 10 measurewhat itisreallydesiredtoknow.Inparticular,itisdifficult10measurethequality of the finished design, although some useful progress has recently beenmadein this arca(ConsD'Uction IndustryInstitute1986}. Some of theideasinthis publicationwill be mentioned later, butitis worth noting at this point that productivity(oreffectiveness}

shouldbedefined in terms ofliltoutput of finished goods and services; in this case complete,cceseaeted andoperatingdesigns.

26

The third difficultyis. there is no general standard for measuring"Brain Work".

For instance.the IWO designers A and 8, can designa hote!intotally different ways.

MaybedesignerAspendsten days for that work, but designerBspendsthirtydaysfor thesamehotel.Itishardtosay whether A's efficiency ishigher than B's.because B's design maybe benet than A's(for Instance,save more money. better satisfy owner etc.).

Thatis whyjusta few productivity experts research whitecoUarproductivity.With the help offuzzylogic,wemaybeabletosolvesomeoCthe"soft" productivity problems.

3.4DifferentMeasurementsorProductivity

Different measuresof productivity serve different purposes.Itis imJX)rtant to choose a measure thatis appropriatctl"lthepurpose. Work·studymodelsserve different goalsthanproductivity models,Substantive discussionsrequirethe knowledge of the defmitionbeingused

There are a number of measure: of productivity that have applicationin economics, construction and design.Ineconomics.where the objective istodevelop measuresfor uscinpolicyplanning,lotal factorproductivity(IFP)isdefined as follows:

ECONOMICSMEASUREMENT:

TotalOulpUt Total Input

Dollarsof output Dollarsof input

(Equation3-1)

TotalOutput

Labour+Materials+Equipment+Energy+Capital

27

Inthis thesis.ifeconomic productivity is mentioned.all resultsare calculated using Equatioo3-1orEquation3-2.

CONSTRUCTION MEASUREMENT:

Inconstructionwe can use this equationtocalculatethe generalpnxfuetivity:

Output(Oollan)

Productivity_ (Equation3-3)

Design+Inspection+Construction+Right-of-way (OoUars) meanwhile •itis usual 10 measureproductivity withreference to project or task performance. Commonly productivity is defined as output per labor cost or outputper laborhour.Allt:mativclythe inversecanbeused.so that labor productivity canbedefined as:

Output

Labour Productivity

=---

Design Productivity

(Equation3-4)

The question of productivity in the design officeismore complex.Current confusion about design productivityappearstostem from atleasttwoproblems:

(1)Nonstandardterminology

(2)To use numericalvaluesto calculate or measure the thinkingand idea generation processes of engineersand lUtbitectsis not easy.

28

Therefore.how 10definethedesign producrivityisalsoa difficult problem, becauseitincludesnotonly product(drawings)•butalso services.In mostdesignfirms, the managerusuallyuses buildingarea divided by designworkhours to define design productivity.

DeslgnProd uctivity=.:B.:U'.:ldi.:"ng=,A.:rea=.::o:::fDe=sig",n:ofS"qc::u==f"'.:'I:c..) DesignTIme( Man-hour)

(Equation 3-5a)

Maybeamore accuratedefinitionIhat can be used by governmentalagenciesfor specificprogram planningand by the privatesectorforconceptualestimateson individual projeclsis:

DesignProductivity

=

=-:::-:-D"F",<",dO-.:II"".,..::.)

=.,..,-,,,---,

(DS+EC+MC+CS+PC)(dollars)

OF:designingfee (designincome)

DS:designer's salary

EC: equipmentcost

MC:materi.ilscosl

CS:consU1lCtionservice

PC: prime planningcost 29

(Equatlon3^{o S b})

3.5 Work Study Models

Aworkstud)' methodIssometimescalleda time-motionstudy.'Thestud)'isdone intwo phases. The preferredmethod of doing the work isflJ'Stdetermined(themotion stud)')and then a timestudy is done to determine me standard time10perfonnmetask.

Commondatacollectiontechniquesusedare timelapsephotograph)',videophotography, stopwatch timingand wod::sampling. Results are commonly presentedusing gang and crewbalancecharts, processchartsand materialflowcharts.

Work samplingis atechniqueinwhichalargenumberof observationsaremade over aperiodoftimeof a consttuctionactivity.The crafts people,machinesandprocesses arestudied andthe percentageoftimespentina nurnberofwori::Slatesis noted.

The selection of classificationsfor workrequires great care.Inconstructionfour

"workstates"arecommonlyusedand aredescribed below:

~This classification ofwork.deals with activities that directly contributeIOconstnlCtionof the project.Examplesincludecraftsmenusingtools,a welder weldinga workeroperatingacoecreevibratoree .

["diM 'Work'This classificationof work is necessarywortinsupportof,but notanintegral pan of.directwork. Examplesincludeacraftsmancleaningup,an employee transponing material, workers studyingdrawings ora craftsmangiving instnleti.oolOhishelpen.

IdW.Idle classificationcovers activity,orlack thereof, that is unrelated10the projectand unexplained.Examplesincludean employee standingidlewhile a secondone cleansup,a craftsmanwalkingemptyhanded,employeeschattingwhilegettinga glassof

30

I!du;.this classification refm to inactivity thalisretaed to unavailabilityoftools or queuing.Examples include craftsmen waitinginline at the tool shed, employees waitir.gfurmaterials to be pickedu" byacrane oremployecs waitingfordirtttion.

Scarfuto (1985)present san exempte ofco nstructio n worksampli ng and categorizestimespent as directwork,indirectwork,idletimeand delaytime.Louis and Borcherding(l986)studied the correlation between the resultsof work sampling measurementsand actual productivity. Resultsshowed a closerelationship between the two.Inaddition,the usefulness of worksam pling informationasapplicatorin the productivityprojection modelwas demonstrated.Thomas (1991)offered anopposing opinion.The hypothesi s that direct worlepercentagesfromworksamplingstudies canbe usedtopredictlabor productivity was earned.Data and observations of the investigation was that direct work cannotbeused to predict laborproductivity.

3.6StatisticalMeasurement:

To improve productivity.the impact of eacb of thevariablesmentionedby Koehn(1986)on labor productivity canbeassessed using statistic.\l methods, and specific attentioncanbethen givento those particular parameters that adversely impact prtduetivity.

Manystatisticalmethodsare available that measuretheimpact of one variable (the dependent variable) on another variable (theindependent··m able).In:....ditionInbeing able to predict the value of the dependent variablebasedon information abc>utan independent variable, a measureofsRngthofthe mmonship betweenthesevariables am

31

alsobedetennined.One measureofthc Strellgth oftherelationshipbetweentwounables

%and y Is called1Mco~cienl0/ linear corretauo« (r)•or simplythe cceeledcn coefficient Given" pair of observationsl;l"i.,Yi ),the samplecorrelationcoefficieeu-cen becomputed as:

(Eq 'lation 3-6)

where:

S,,-

LXY'(r,xllLY) n

(Ly)2

S,,_

^{Ly2 •- n -.}

Inordertofindtheproportion (,2 ) of the totalvariablesofthey-velues lhatIItC accounted for bytheindependentvariablex,lite foUowing equationcan beused:

(Equation 3·7)

32

Similarly,1·~ representsdlat proportionofthetotalvariabilityofthey -values thatarenotaccountedfor by!he.r variable.Theequationsdescribedabovecanbeusedif.

and onlyif.there is a linear relationshipbetween x andj, Other models become necessary when Ihe relationship between rand y isnotlinear;that is,Jincreases or decreaseswith x but notina linearfashion. The approach usedinthis studywas theIlU1k correlationcoefficient.whichmeasuresthe monatomicrelationshipbetween1 andr.that is,J increasesor decreaseswithx evenwhentherelationbetweenJand:ris nonlinear.

3.7Using Fuzzy Relational Deta-Base To Measure The Problemsor DesignandCcnstrueuon.

(I). Concept or Fuzz)' Set:

Fuzzinessrepresents situationswhere membershipinsets cannot be definedon a yeslnobasis becausethe boundariesof sets arevague(Zadeh 1965).The centralconcept offuzzy-settheoyis themembcrslUp function,which numericallyrepresentsthedegreeto which an element belongs to a set.Ina classical set, a Sharpor unambiguousdistinction exists betweenthemembersandnonmembersof the set.Inother words.the valueof the membership functionof eachelementin theclassicalsetiseither1 for memben(thosethat ccnainly belongintothe set ) or 0 for numbers (thosethatccrtainlydotlO1).However, manysets,such as the sets of complexsystem,nice houses. beautifulplace, and numbers muchgrc;tterthan1.0.do notex.hibit thischamcteristic,that is. their ooundaries arefuzzy,

The analysisofthebasiccrileriafor thedesignlconstruetior. problem is estimated.

asfuzzyvaJucstocharacterizetheiruncertainty.Thefuzzyvaluesare numbenthatbelong to a given set (interval )witha degree of membership. To evaluate the various design/constructionproblems under oncenainty. letz.{X)bea fuz.zyvalue for theith

33

basiccriterion.andlet Itsmembership function m[l.-(.l")] beacrapczoid(Fig3-1).where xisoneeteeem(problem name)ofthediscretesetofdesign/constrUCtion problems.U the:bapeZOidisreduced10a venaJ.1ine,ittq:lttSm1Saso-calkdcrisp(nonfurzy)number.

AIcvel-c:utconcept(DongendShah 1987)canbeusedIodcflnCtheinlel'ValrLeachbasic crilerionatvariousdegrttSofmembership.Themembership dep 'eCforanuncenainvalue can be determinedusing"expert judgment"basedonexperience and observation variability.AsshowninFig3-2.^Q(;c)istheinterValvalueoftheithbasicaiterionuthe membcshipdegreeh(l.e..a<ZuI.x)<b).

1. If BESZI >WCflZI ,then Si,h(x)•

r.

^1,

~Z/,h(X) .WORZI

Y

(BESZ1-WORZ/),

a,

2).11BESZic WC1AZi,then S/,h(x).

~

^{[Zi,h(x)-WOR Zi}

^Y

^(BES1

^{a ,}

^,

^{n -}

^WORZi),

34

Z_BESZII,h_cZi,h(X)(Xl < WORZi~m;SZi

J

Z I,h(x )<..WORZi (EquallonU)

Z i,h (Xl ",BESZi ]

BESZicZl,h(Xl<WORZi Z i,h(x )>.,woRZi

(Equation:H)

Il"(Zi(r»)

1.0

0.0

IDOStlikelyintervai

I I

Largcst likelyintcrval

Fig3-! Fwz)'E.stimateaf

it "

BasicCriterion Zi(r)

Since: the unitsofthebasiccriteriaarediffered!(suchIStechnicalfaaonDOl

bemaCJF;pressedinunits ...allwhilethecostisindoll&rs).thusmakingit difficuhto compare them directly.chc.ctualvalue of eacbbasiccriterionlZI,h{.zJJshouldbe tranSformedinto an index.Usingthebest value(BESZI)ofZlandthewontvalue (WORZ;)ofZ;fortheIlhbastecriterion,theactualvalueZi,h(x)canbetransfCll1DCd intoanindexvalueS;,Iz(x) asindicatedby(Fig3-2)

Toassign the best and wontvalucs(i.e. BESZjand WORZ, )of theI thbasic criterion,oneof twooptionscanbeused.The firstistoassign the bc.standworstvalues of theithbasiccriterionaccordingtotheoverallbestandworst valuesoftheIthbasic criterion amongthe design/constnl(..1ionproblemsconsidered.Thesecondoptiooisto assignthebestandwontvalues oftheithbasicailCriOllaccordingtotheopinionof an

3S

expert. Since the actual valuez..,A(.r)islUIintervalwith lowerbounda andupperboundb (Fig3·1 ).the indexvalueSi,h(x)resultingfromZi)l~lis also aninterval (Fig3-2),

a)BESZi>WORZt SiJ(,r)

Black Gray White

1.0

d

I---+----~./"

.,/(1

WQRlE

b) BESZi<WORZi

b BESZi

BESZi • b WQRli 'ZJJt(,rJ

FlgJ..2 TransferringActuaJValueZt,I(q intoIndoVa!Ui:S I,A(,rJ

36

Indesignarxtconstructionproductivity measurement,manyproblemsarc~fuuy~.

For instance.it is very hardtomeasureone engineer'sidea using "input"or "output".To measuredesign quality is also acomplex procedure.What is the meaningof "Excellent design?"~Verygoodconstruction?".From"poor"(0)to"excellent" ( I ), we have'.ogo througha"gray" area.( from0 to1).

(2).Fuzzy Relation Dala·Base

Inthis thesis the author wants to use the Fuzzy Relation Data-Base systemto analyze the relationship between design and construction. Because the Fuzzy Relation Data·Base(FRDB)( Candel,1986) model ...asdesigned to satisfy the requirementsfor sound formalfoundations,real-worldinformationmodels. individualization,and user's convenience.The RelationalData-Basestructurecombined with the theoryof fuzzysets providea solid theoretical foundation. The querylanguagepermits"natural-language-like"

expressionsthatare easily understood by users.and canbe furtherdeveloped to incorporate fuzzy inferences or productionrules.Anexperimental FROB systemwas developedto teS11hefeasibilityofanImprecise information system.. The FROBhasbeen usedasa useful tooltoevaluate thecomplexityof engineeringproblemswherethereare conflictingobjectives.the objectiveshavevarying degrees of importance,and valuesof input variablesm unoertain(A.KandeI1986,Lecet. aI.I991,z.x.Hee.aI.1m).This is a multiJc:vel, multiobjective methodusingfuzzysets10representtheuncc:naintyininput variables. The specific objectives of this studyarctwofields.Thefirst objective is to developanevaluationsuppon systembasedona FRDBmethod(Table3-1andTable3-2), thesecondis to applytheevaluationsuppon systemto forecastthedeviationof designand constn:.etion.

InChaptet 6,the authorwillintroduce the usc of the SingleFactorEvaluation methodofFRDBsystemtoevaluate thecomplexityofdesignandconsuuctionprojects.

37

Chapter 4. DESIGN PRODUCTIVITY 4.1.IntrodudiCMI

Thedesign productivityproblemiscloselylinkedtothedesign qualityproblem.

Alotof researchshowsthaihighquality designcantangiblyreducetheCOlI of construction. Therearetwoways10discuss the design quality problems. dlatartthe economicalandtechnicalway.Inthis chapter,theauthorwill • fromancconomica.Iview point.discusstheinflucnceofdesign onproductivityinconstruction;thecompressionof designcost andconstructioncost;optimum designanddesign decisions .

Theterm"design"isdefined as thecreatio nofplansandspecificatio ns that resultin theallocation ofresources.toaccomplisha project(Dickmonand Robenshaw 1975.) .Design consists ofthreesub processes:(I)de finingthe problem. orthe

"conce ptualphase"; (2)g~eratingandevaluatinc alternatives,or the"preli millAl)'

design phase";and (3)reducingthe best solutionto.descriptionforconsae uce,orthe

"detaileddesign phase."Insomeengineering disciplines.desi.."consistsof•tesCand

revisio nstage; this mgcisno«gcnenll y applkabl cinCOrl$l2'UCdou.Ontheocherhand.

«lILSuuctiondesignmust includethe revi siorasanduloIelJtttalionsthaioccurdurina consttuction. Consmx:tiondesig n im;:ludesplanninZ.cost. sched ule,and quality function s dlatleadto thespecificationofconstnlCdonconditiollJ.Insome countries.

detaileddesign includeconsaucti o n dnl.wing,even includematerialsspecification,for instance. steel barspecification.woodspecification.or materialslistandinsl.a1lldon spcci..fication.

4.2.DesignProcrss

ThepogressofIprojectfromtheinitialidea10engineeringrealityis depiCltd by

• processsucb as dlatshowninFiBA.1(McOmrgc" 1988).Thisdiagrunpves aJricturc 38

which is !OOrough and simple.There areoeerespecu which couldbe addedtoFig,4-1. suchas commissiOning. operating, maintaininS, and decommissioning.These processes arealsoveryimportantHowever, thefact is thatdifferenldesignoffices have diffm:nt customers,and different projects have differentcharacteristics.All of these have been omitted inthis as notessentialtothe argumentItisessentialthatthesef" tonbetaken intoaccount when assessingthelife cost of a projecL From this figure.wecan seethat notonly designisap~uc:tprocess bulalso a serviceprocess,fromproject investigation untilthe constructionhasbeencompleted. Serviceisa continuousprocess.

Fwthermore,designsare not producedjust by draftsmen(orbyengineen).but bya systemof resources workingas ateam.Toreturn 10the conseuctionsiteanalogy,it is no good speedinguptheree of concrete placementifthatonlymeansthatthe operator hasto wailforthefmnworltcrew.Ina similarway delaysaoddisruptionsarethebiggest impediment 10 pl'Oductivity(and morale) on the site. Hence design office output is affected by failures o( the systemto(unction properly.Failure of the system usually maniteslSitselfin sbonageofakeyresource,materiaJ.labor,equipment, or information (especiallyconcerning decisionson criticalissues). Themostimportantof these,illthe design officeconIcAl,is information.The enginewhich drives thedesign systemisthe brainpowerof its members,andthe fuelwhichpowersitisinformation.Thedesignoffice is the quintessentialinfonnation processing system. Informationdrawn froma hostof sources- themarketplace,designcodes,ICChnical. material.the client'stermsof reference,the knowledgeandexperience of thedcsigners--isprocessedintosomething whichis(hopefully)elegant,useful,andeconomicaL

39

F'lg 4-1 ProjectDesign Process (JohnF.McGeorge,1988)

Itwasstaled earlier that productivity mUSIbedefinedinlermsof the OIOrpUt of finishedgoods and services, Le., activity does not equal productivity.Inorder tobe productive, an activity mUSl contribute towardstheattainmentof the desftd gOlli. Itis necessary Ihatthedraftsmanisnotjust active, but productive,andtoachievethiJrequires thatattcnnonbegiveD to thenatureofthe design process.

4.3.DesIgnInputand Construction Cost

The second feature of the process is that the costof completing eachstage increasesrapidly,in moreor lessexponential fashion,asindi~bydie cost pynmid.

Thecostofdesignis generally consideredtoberoughly between2'lJand10%ofthetotal 40

costs (Institution ofCivilEngineers1985).Theimplicationsof !hisarcinteresting.Since designcostsrepresent only a smallpro portionofIOtalcosts,itbecomeswonhwhi le 10 increasethedesign effort significantlyinorderto achievecomparativelysmall reductions inconstructioncost.A simple example(McGeorge. 1988).using a design cost of 5%is shownin Table4-1.

EffectolExtr. Oeshra]alit

CMU

. ...

'.m Ori_

"''"'"

ees

""' ..

Iksignoost SO SO+(SOlSOtl>)oo7S

ConsuuctioaCOlit 'SO 9SO·(9S0xIMI)=8SS

T""'_

^{1000 830}

OveraUSlvin. 70l7tl>

Inthiscaseithasbeen assumedthatISO% increeseindesign inputyields I 10%savingin consb'Uctioocost.The netresultiSI790reductionin totalcost, orasaving almost3time.s thecost of the extradesignwort.11shouldbequestioned" of course • whethersuchfiguresare realistic,IlJId the evidenceIvailableindicateSthat they are.ID fact, ratherthanoverstate.~sean:hcarriedout on CODStnK:tabUity.Forexample.indicates thatconsO'UctabililyandvalueengineeriDg reviews typically yieldCODSttllCtiODcost

41

savingsof10to20 timesthecostofthe extra design inpulCBusintss RoundTable 1982).

Even these figuresuDderstatethe true potentialfor improvements,as consaucubility exercises.bydeflDition. aim atreducingonly construction costs,andthe pictureis improvedbyincluding operatingandmaintenancecosts,which frequently exceed constructioncost.To we one example, recently compiledfigureson hospiulsindicate thatoperatingcosts exceedthecapitalcostof buildinga hospitalwithinonly2to3 yean of operation(NauooalBuildingResearchlnsuttKe198$).

4.4Optimum Design

(I).Thebalanee between designinput andconslrudioncost Naturally,improvementscannot continuetobemadeindefinitely.Atsome point an increase in designinput winyieldan insignificant savingonconstruction,andtotal costwillbehigher.lbus,plotting design effort againstIOtalcostyieldsa curvelikethe onesho wninFig4-2. A rational policythenwoul dbe toaimfortheoptimumde3ign effort'0',butthe implkationsoftheforcgoiJlgdiscussion are that most engineering projectsfanwell 10the\eft,at somepoint'lI' •Thesoa::essof COOSttUCl8bility progrvnsis a resultofRCOgnUingandICIingupotlthissimpleancIobvious fet.

42

0= Optimum design Input a=TypicalproJectla.el

a

DESIGN INPUT

o

f1g4~2 Erred orDesignEffort on ProJed Cost

(2). TheBalancebetween Mullifactorcost

Note.usually the"optimumNor"cheapest"means fromalotal cost viewpoint, anddoes notreferto any individualcost Wehaveto understandthatthe mosteconomic designis notnecessarily thecheapest;itistheonewhich givesthebest valueformoney.

Inthe productive sector of theeconomy.theobjectiveis toproducegoods and services.

notfortheir own sake. bUIinorder10makea profi t.Thefinalcriterionforthe entrepreneuris!hediffm:nce betweenrevenueandexpenditureand the relationshipollhis difference10thecapitalemployed.The buildingisoneof these expenses.

43

For instance,the cost of a building to an owner canbe visualized in threepans.

The first pan is thecostofconseucecn,thisis basic becausethe owner'sneedseSlablish what functions and size a buildingshould have. The resulting design sets construction costs.Allother costs an: related to, orare a functionof, construction costs.The second partof the cost ofabuildingincludes all those expenses thatanowner must incur. other lhan construction, before a building can be occupied. Thethirdpart of the cost of a buildingtakesinall theexpensesincurredbythe ownerduringitsuscfullife.These IlII1er an:life-cyclecosts.The cost of operationsandmaintenance dependsinlargepartonthe initialcostandon the qualityof thearchitecturaland engineeringbuildingsystems.Low initialbuildingcosts achievedby thesacrificeof qualitycan resultinhighlife-cyclecosts.

Figure4-3illustratesthese threepans of the costofa building. TheIltst cclumn shows that,ontheaverage,architectural buildingsystemsaccountfor about39percentof the construction costs, structural engineering systems 25 percent, and mechanical engineering systems36%percent The second column illustrated other costs before occupancy.Landisa large variableand canbesignificantindowntownwbanlccadces.

(Sometimesthe land cost may accountfor upto50-100percentor moreof the building.) Thethirdcolumn showsthat totalbuildingcostsover many yearsdwarf the initialcost of construction.

44

''''''

THE COST OF ABUILD IN G

~CT\J""

\ ~""', .. ^,,'"

'1

AEH:lIIATDlS

IlUIDt<G FHN«:lNG

,...,..,

SYSTEMS

....

^TAXES

SmucTIJRAL .~ MAt/T£NAHC€

WlJlt<G

~

SYSTEM lAI<l _\ <>'ERATDlS

""

--=

MACHANtCAL ElEcm<:AL

§

WID"" <XlNST

SYSTEMf> 36"'- COOTll

Fl14..J ABulldlncCostAnalysts

Wecan use thistoexpresstherelationship0(thevariouscost foroptimiu~

purpose'

Minimumcost.Construction Cost + Other directand indirectCost ... Ufe·CycleCost

s.t, [Construction Cost>0 Otherdirect andindirectcost> 0

Ufe-Cycle Cost >0

45

]