Slide 1
Chapter 6 Value and Risk Management:Multi-Attribute Utility
TheoryComparison of strengths and weaknessesEnhance Value to
increase scoreManage RiskMost of the chapters figures are included
in the file.Instructor must decide how many and which examples to
use.Chelst & Canbolat Value-Added Decision Making1Figure 6.1:
MAUT process - Analysis
29/19/2011Chapter 6Chelst & Canbolat Value-Added Decision
MakingSteps in AnalysisCheck for data errorsInputsRanges and
preferencesAnalysis of uncertainty, strengths, and
weaknessesRobustness of optimal - weights changeValue Added and
Hybrid Improve values of highly weighted objectiveReduce
risk39/19/2011Chapter 6Chelst & Canbolat Value-Added Decision
MakingFigure 6.2: Objectives Hierarchy High Ceiling Kitchen
FixturesAmount of Light 0.274Dimmable 0.055Type of Light 0.205Max
Light Quality 0.534Operating 0.219Purchase 0.137Min Cost
0.356Change Bulbs 0.110Min Hassle 0.110Best Lighting System
1.00049/19/2011Chapter 6Chelst & Canbolat Value-Added Decision
MakingDiscuss weightsLight quality is more than 50% of weight
amount is most important with Type at .21 .Cost is a little more
than 1/3 rd of total with operating outweighing purchase cost.
Operating cost is also 0.22.4High Ceiling Data: 20 bulbs5Bulb
:Amount of Light (Lumens)Replace Bulbs
DimmableOperatingPurchaseType of Light65 Watt basic62010 or
moreNo27050Incandescent75 Watt basic9006 to 9No31576Incandescent65
Fluorescent7505 or fewerNo7090Fluorescent75 Fluorescent Dim9005 or
fewerYes80160Fluorescent75 Halogen Dim10206 to
9Yes315150HalogenReplace BulbsType of
LightCategoryUtilityCategoryUtility10 or more0Fluorescent06 to
90.5Incandescent.755 or fewer1Halogen19/19/2011Chapter 6Chelst
& Canbolat Value-Added Decision MakingNot linear with 3
categories with regard to type. Incandescent is 0.75 as good as
halogen true colors.Use lumens because watts has little meaning.75
watt fluorescent is based on standard packaging that lists
equivalence. The actual wattage is only one-sixth. The actual
wattage is used in calculating operating costs.5Errors in Input
Check Ranking Results GraphSymptom 1 One line goes above 1 or below
0Data input error in that alternative too large or too small added
a zero or left off a zeroSet range too narrow to include all
valuesSymptom 2 All lines on one measure outside rangeForgot to
reset range on that measure. LDW uses default range of 0 to 1.
Symptom 3 Ranking of all alternatives on a measure seems upside
downMost preferred and least preferred settings are reversed.
69/19/2011Chapter 6Chelst & Canbolat Value-Added Decision
MakingFigure 6.3: Ranking results graph for lighting system example
data input error (figure corrected from text)Utility 1.000
0.000Best Lighting SystemAmount of lightOperatingType of
lightPurchaseChange bulbs Dimmable65 watt basic75 watt basic75
Halogen Dim65 Fluorescent75 Fluorescent Dim9/19/2011Chapter 6Chelst
& Canbolat Value-Added Decision MakingFigure 6.4: Wrong
direction on operating cost measure reversed Fluorescents should
have best operating costs and not worst (near 0)
89/19/2011Chapter 6Chelst & Canbolat Value-Added Decision
MakingFigure 6.5: Results graph for lighting system example (figure
corrected from text)Utility 1.000 0.000Best lightingMax.
qualityMin. costMin. hassle75 Halogen Dim65 Fluorescent75
Fluorescent Dim65 watt basic75 watt basic9/19/2011Chapter 6Chelst
& Canbolat Value-Added Decision MakingFigure 6.5: Results graph
for lighting system example ERROR in GRAPH - ReplaceUtility 1.000
0.000Best LightingMax. QualityMin. CostMin. Hassle75 Watt Fluoro
Dim75 Watt Basic65 Watt Fluoro65 Watt Basic75 Halogen
Dim109/19/2011Chapter 6Chelst & Canbolat Value-Added Decision
MakingFigure 6.6: Objective (3) stacked bar results for lighting
system exampleRanking for Best Lighting System GoalAlternative75
Halogen Dim75 Fluorescent Dim75 watt basic65 Fluorescent65 watt
basicUtility 0.618 0.584 0.542 0.526 0.359Max Light QualityMin
CostMin Hassle119/19/2011Chapter 6Chelst & Canbolat Value-Added
Decision MakingHalogen earns most of its value from light quality75
fluorescent equal value from quality and cost11Figure 6.7: Measure
(6) stacked bar results for lighting systemRanking for Best
Lighting System GoalAlternative75 Halogen Dim75 Fluorescent Dim75
watt basic65 Fluorescent65 watt basicUtility 0.618 0.584 0.542
0.526 0.359Amount of LightPurchaseOperatingChange Bulbs Type of
LightDimmable129/19/2011Chapter 6Chelst & Canbolat Value-Added
Decision MakingMore details12Figure 6.8: Pairwise comparison
between the 75-watt halogen and equivalent fluorescent
139/19/2011Chapter 6Chelst & Canbolat Value-Added Decision
MakingAlmost all of positive difference involves type of light.
Fluorescents are bad with true light color13Figure 6.9: Sensitivity
analysis for weight placed on Amount of Light => NOT
sensitive
149/19/2011Chapter 6Chelst & Canbolat Value-Added Decision
MakingVertical line is current weight 0.27Halogen is best at that
pointAs weight increases best solution becomes even better relative
to others since it provides the most lightDecrease it would have to
decrease by more than 50% to change ranking.Conclusion not
sensitive to this weight change14Figure 6.10: Sensitivity to weight
placed on Type of Light Sensitive to decrease
159/19/2011Chapter 6Chelst & Canbolat Value-Added Decision
MakingNot sensitive to increaseHowever if weight assigned to type
of light decreases below approx 0.18 solution changes fluor with
dimmer is bestWith zero weight fluorescent becomes much better.
Remember fluorescents are the worst at producing true
colors15Figure 6.11: Sensitivity to weight placed on Operating Cost
Sensitive to increase
169/19/2011Chapter 6Chelst & Canbolat Value-Added Decision
MakingEven small increases in weight on operating cost sends
halogen to 2nd and then 3rd place. Steep decline.16The Best
Alternative So Far May NOT Be The Final Answer: ValuesEnhance best
alternative(s) or create better Hybrids through: Value Analysis
& ManagementEnhance Best or 2nd beat Alternative Identify a
highly weighted but weak measure level in the best
alternative(s).Creatively identify a way to improve the
alternatives measure level and specify associated changes in other
measure levels such as added cost.Hybrid: combine two alternatives
in a way that builds on the strengths of the best alternatives.
Evaluate the newly formed alternatives.179/19/2011Chapter 6Chelst
& Canbolat Value-Added Decision Making17Create a Lighting
HybridIn best solution identify a weak measure with significant
weightOperating cost - weight 0.22Creative alternative: 50-50High
quality light where needed half of kitchenFluorescent with lower
cost where true color quality does not matter189/19/2011Chapter
6Chelst & Canbolat Value-Added Decision MakingFigure 6.12:
Ranking Hybrid Lighting System: 50-50 split in kitchen Reduce
operating costRankin for Best Lighting System GoalAlternative75
Halogen Hybrid75 Halogen Dim75 Fluorescent Dim75 Watt basic65
Fluorescent65 Watt basicUtility 0.640 0.618 0.584 0.542 0.526
0.359Max Light QualityMin CostMin Hassle19Bulb Amount of
LightChange Bulbs DimmableOperatingPurchaseType of Light75
Fluorescent Dim9005 or fewerYes80160Fluorescent75 Halogen Dim10206
to 9Yes315150HalogenHybrid 50-50 split9606 to 9Yes200160Halogen -
Fl9/19/2011Chapter 6Chelst & Canbolat Value-Added Decision
MakingHybrid was based on the weakness in operating cost for
halogen. - 19The Best Alternative So Far May NOT Be The Final
Answer: ValuesEnhance best alternative(s) or create better Hybrids
through: Value Analysis & ManagementEnhance Best or 2nd beat
Alternative Identify a highly weighted but weak measure level in
the best alternative(s).Creatively identify a way to improve the
alternatives measure level and specify associated changes in other
measure levels such as added cost.Hybrid: combine two alternatives
in a way that builds on the strengths of the best alternatives.
Evaluate the newly formed alternatives.209/19/2011Chapter 6Chelst
& Canbolat Value-Added Decision Making20Figure 6.13: Stacked
bar ranking for kitchen remodeling exampleAlternativeBuild
RiteQuality BuildCost ConsciousUtility 0.651 0.630 0.462Max.
QualityMin. CostMin. Hassle219/19/2011Chapter 6Chelst &
Canbolat Value-Added Decision MakingStrength in quality and not as
good on cost21Kitchen remodeler - weights22
9/19/2011Chapter 6Chelst & Canbolat Value-Added Decision
MakingKitchen remodeler - Data23Chapter 5MeasureBuild RiteQuality
BuildCost ConsciousTotal labor cost$34,000$26,000$25,000Total
material cost$20,000$12,000$10,000Cost overrun history0% (p=0.33)
2% (p=0.34)7% (p=0.33)2% (p=0.33) 5% (p=0.34)9% (p=0.33)6% (p=0.33)
9% (p=0.34)15% (p=0.33)Duration kitchen unavailable13 weeks10
weeks9 weeksWeeks of delayOn time (p=0.33),1 week late (p=0.34)2
weeks late (p=0.33)1 week late(p=0.33)2 weeks late(p=0.34)3 weeks
late (p=0.33)2 weeks late (p=0.33)3 weeks late (p=0.34)4 weeks late
(p=0.33)Cleanliness scaleCleanMessyDirtyFollow-up and resolution
scaleAdequateHighly responsiveAdequateCreativity scaleHighly
creativeCreativeMundaneBrand & store reputation scaleTop of
line2nd Best Brand2nd Best BrandPercent use of
subcontractors25%40%65%Fit and finish scaleExcellentGoodGoodYears
in business 12 (Good)8 (OK)22 (Excellent)Quality of references
scaleExcellentGoodOKChelst & Canbolat Value-Added Decision
MakingIn these data the uncertainty associated with contractor is
independent of the other contractors experience.In some instances
the uncertainties are linked. For example in the bulb example the
uncertainty regarding hours of oepration would be the same for all
3 bulbs and would have a commensurate impact on operating cost.
23Figure 6.14: Comparison of top two kitchen remodelers
249/19/2011Chapter 6Chelst & Canbolat Value-Added Decision
MakingQuality build Weak but not correctable (creativity and
references) Weak and correctable (fit and finish)24Enhance 2nd best
Best on CostSpend a little more money and improveIn 2nd best
solution identify a weak measure with significant weight that can
be improvedFit and finish (0.12 weight)Cleanliness and percent use
of contractors (minor)Creativity and References can NOT be
improvedCreative alternative: Spend money to improve fit and finish
(and other benefits)Less subcontractingCleaner259/19/2011Chapter
6Chelst & Canbolat Value-Added Decision MakingImprove Quality
Build on Fit-Finish & add cost26MeasureQuality BuildQuality
Build + Value EnhancementTotal Labor Cost$26,000$29,000Total
Material Cost$12,000$12,000Cost Overrun History2% (p=0.33), 5%
(p=0.34),9% (p=0.33)2% (p=0.33), 5% (p=0.34),9% (p=0.33)Kitchen
Unavailable10 weeks10 weeksWeeks of Delay1 week late (p=0.33),2
weeks late (p=0.34),3 weeks late (p=0.33)1 week late (p=0.33),2
weeks late (p=0.34),3 weeks late (p=0.33)Cleanliness Created
ScaleMessyCleanFollow-up and Resolution ScaleHighly
responsiveHighly responsiveCreativity ScaleCreativeCreativeBrand
& Store Reputation scaleModerate priceModerate pricePercent Use
of Subcontractors40%20%Fit and Finish ScaleGoodExcellentYears in
Business but GroupedGoodGoodQuality of References
ScaleGoodGood9/19/2011Chapter 6Chelst & Canbolat Value-Added
Decision MakingFigure 6.15: Stacked bar results after value
enhancement: kitchen remodelerAlternativeQuality Build + Value
EnhancementBuild RiteQuality BuildUtility0.6780.6510.630Max.
QualityMin. CostMin. Hassle279/19/2011Chapter 6Chelst &
Canbolat Value-Added Decision MakingThe Best Alternative So Far May
NOT Be The Final Answer: RisksEnhance best alternative(s) through:
Risk Analysis & ManagementIdentify a highly weighted measure
with significant uncertainty in the best alternative(s).Assess the
impact on the MUF of reducing downside risk on that measure.Develop
a strategy for reducing the downside risk even if it changes other
measure levels. Evaluate any newly formed
alternatives.289/19/2011Chapter 6Chelst & Canbolat Value-Added
Decision Making28Value Analysis & Risk Management Conformal
Coating Process Printed BoardsGlobal Electronic will install a new
conformal coating process because the upcoming Powertrain Control
Module (PCM) design requirements are incompatible with the existing
coating process at the plant. These coatings are applied to the
printed wiring boards to protect circuitry from environmental
exposure after the installation of all surface mount devices, but
before final assembly of the module. The process should ideally be
capable of selectively applying the coating to various areas of the
circuit board, coating some areas while avoiding others.
Pre-screening of a wide variety of available processes has reduced
the number of viable candidates to three. The team realized that
these processes vary widely in ability to accommodate design
changes (flexibility), weight, initial investment costs, material
costs, etc. 299/19/2011Chapter 6Chelst & Canbolat Value-Added
Decision Making29Goals Hierarchy & Weights: Coating Process
SelectionSelect Best Coating Process (1.000)Performance (0.
213)Cost(0.324)Time(0.148)Reliability(0.315)Flexibility
(0.157)Weight(0.056)Coating Control(0.130)Material (0.167)Labor
(0.037)Facilities & Tooling (0.093) Development Time
(0.148)Scrap (0.028)Foreign Material(0.185)309/19/2011Chapter
6Chelst & Canbolat Value-Added Decision Making30Coating
Process: Development Time Uncertainty (Risk)Selective Spray Sil-Gel
Potting Coat and Extract
DTPrDTPrDTPr280.15160.40280.10320.45200.50300.20360.35240.10340.60480.05400.1031Chapter
6Chelst & Canbolat Value-Added Decision
Making31MeasureSelective SpraySil-Gel PottingCoat and
ExtractFlexibilityHighMediumLowWeight (Gr)623020Coating
Control011Foreign MaterialSuperiorExcellentGoodFacilities &
Tooling C. 30000025000110000Labor Cost ($)400001000020000Material
C. ($)1700061500063000Scrap C. ($)95000011000Development Time
(W)34.1918.9933.48Alternatives & Data: Coating ProcessRed cells
illustrate expected value of probabilistic data32Chapter 6Chelst
& Canbolat Value-Added Decision Making32Results: Coating
Process SelectionSelective Spray: most preferred alternativeBest
alternative affected by uncertainty. Sil-Gel a close second: 7%
lessAt the extreme it may be better than the best.Development time
involves significantly more uncertainty in the best alternative
(Selective Spray) than for the 2nd best
33Chapter 6Chelst & Canbolat Value-Added Decision
Making33FOCUSHybrids: Both Risk and Value ManagementRisk
ManagementSignificant uncertainty in development time of best
alternativeInvest $40,000 to eliminate risk of unusually long PD
timeValue ManagementCoating Control is a weakness of the highest
ranked alternativeInvest $60,000 to improve coating control reduce
scrapRisk and Value ManagementCreate hybrid that combines both risk
and value managementSlightly improved overall scoreDramatically
reduce chance of lower performance.349/19/2011Chapter 6Chelst &
Canbolat Value-Added Decision Making34Risk Management: Coating
Process SelectionGlobal Electronic contacted the Selective Spray
supplier to reduce uncertainty in development time. The supplier
asked $40000 more for tooling premium to work overtime and reduce
development time to a range of 28 weeks to 32 weeks.Development
Time (Weeks)Probability280.15320.45360.35480.05Development Time
(Weeks)Probability280.40300.40320.20OLD NEW35Chapter 6Chelst &
Canbolat Value-Added Decision Making35Risk Management Coating
Process Selection (Cont.)$40000 more investment reduces uncertainty
significantlyAmount added to the Facilities and Tooling Cost of new
alternative.Overall average score only slightly improved but less
downside riskSelective Spray + Risk Mng. is the best
alternative
36Chapter 6Chelst & Canbolat Value-Added Decision
Making36Hybrid Alternative Through Value & Risk ManagementThe
Selective Spray supplier can not reduce facilities and tooling
cost, or labor cost The supplier asks $60000 to upgrade coating
application nozzles that Improve coating control from 0 to 1Coating
control provides an assessment of the process ability to apply
coating where it is needed, as well as preventing coating bleed
into undesirable areas of the printed wiring boards. 0: Problem
areas may affect function1: Problem areas dont affect
functionReduce scrap cost from $95,000 to $10,000 per yearAsks
$40000 to reduce development time to a range of 28 weeks to 32
weeks 379/19/2011Chapter 6Chelst & Canbolat Value-Added
Decision Making37Create Hybrid AlternativeThrough Value Analysis
& Risk ManagementGenerate a new alternative that is a hybrid of
the Selective Spray and the suppliers new offerImprove coating
control, reduce scrap cost and development time Pays $100000
(=$40,000+$60,000) that increases facilities and tooling cost to
$400,000MeasureSelective SpraySelective Spray + Value
Mng.FlexibilityHighHighWeight5.65.6Coating Control01Foreign
MaterialSuperiorSuperiorFacilities & Tool300000400000Labor
Cost4050040500Material Cost1778017780Scrap
Cost9500010000Development T.34.1929.6638Chapter 6Chelst &
Canbolat Value-Added Decision Making38Enhanced Alternative: Higher
Value and Less RiskSelective Spray+Value&Risk Mng. is the most
preferred alternative
39Chapter 6Chelst & Canbolat Value-Added Decision
Making39Significant improvement in value and less riskFigure 6.28:
MAUT SME/Decision makers meeting agenda
409/19/2011Chapter 6Chelst & Canbolat Value-Added Decision
MakingMore Examples from textWarehouse location selectionCoating
Value management figuresDisposition of weapons grade
plutonium419/19/2011Chapter 6Chelst & Canbolat Value-Added
Decision MakingFigure 6.16: Objectives hierarchy for warehouse
selection example
429/19/2011Chapter 6Chelst & Canbolat Value-Added Decision
MakingFigure 6.17: Non-linear utility functions for parking
spaces
439/19/2011Chapter 6Chelst & Canbolat Value-Added Decision
MakingFigure 6.18: Stacked bar ranking for warehouse site selection
exampleAlternativeFedCo PropertiesCenter DriveProspect
ParkNorthbrook Business CenterUtility 0.600 0.579 0.421 0.377Max.
Operational DesignMin. Loss of Current EmployeesMin. Total CostMin.
Distance to Key FacilitiesMax. SpaceMax. Building
Appearance449/19/2011Chapter 6Chelst & Canbolat Value-Added
Decision MakingFigure 6.19: Comparison of Center Drive and FedCo
Properties facilitiesOverall Utility forFedCo PropertiesCenter
DriveDifference 0.600 0.579 0.020Total DifferenceNumber of Truck
DocksLease and Maintenance CostOffice and Lab SpaceWarehouse Floor
SpaceAppearanceMaterial Handling CostNumber of Parking
SpacesDistance to Niles FacilityCenter DriveFedCo
Properties459/19/2011Chapter 6Chelst & Canbolat Value-Added
Decision MakingFigure 6.20: Stacked bar ranking after value
management for warehouse site selection
469/19/2011Chapter 6Chelst & Canbolat Value-Added Decision
MakingFigure 6.21: Stacked bar ranking for the coating
processes
479/19/2011Chapter 6Chelst & Canbolat Value-Added Decision
Making47Figure 6.22: Comparison of selective spray and sil-gel
pottingOverall Utility forSelective SpraySil-Gel PottingDifference
0.702 0.650 0.051Total DifferenceMaterial CostCoating
ControlFacilities&Tooling CostFlexibilityForeign
MaterialWeightDevelopment TimeLabor CostScrap CostSil-Gel
PottingSelective Spray489/19/2011Chapter 6Chelst & Canbolat
Value-Added Decision MakingFigure 6.23: Stacked bar results after
value management
499/19/2011Chapter 6Chelst & Canbolat Value-Added Decision
MakingFigure 6.25: Ranking alternatives after value and risk
management
509/19/2011Chapter 6Chelst & Canbolat Value-Added Decision
MakingFigure 6.26: Ranking results including cutoff value in
Logical Decisions Alternative75 Halogen Dim75 Fluorescent Dim75
watt basic65 Fluorescent65 watt basicUtility 0.618 0.584 0.542
0.526 0.359Alternative failed at least one cutoff519/19/2011Chapter
6Chelst & Canbolat Value-Added Decision MakingFigure 6.29:
High-level objectives hierarchy for disposition of surplus
weapons-grade plutonium
529/19/2011Chapter 6Chelst & Canbolat Value-Added Decision
MakingFigure 6.29: Overall ranking for disposition of surplus
weapons-grade plutonium
539/19/2011Chapter 6Chelst & Canbolat Value-Added Decision
MakingFigure 6.31: Sensitivity analysis for Reactor alternatives
for disposition of surplus weapons-grade
plutonium549/19/2011Chapter 6Chelst & Canbolat Value-Added
Decision MakingFigure 6.32: Sensitivity analysis for Reactor
alternatives for disposition of surplus weapons-grade
plutonium559/19/2011Chapter 6Chelst & Canbolat Value-Added
Decision Making
Office and Lab Space
Truck Traffic Handling
Number of Truck Docks
Number of Parking Spaces
Lease and Maintenance Cost
% Loss of Employees
Distance to Niles Facility
Distance to Headquarters
Max. Operational Design
Min. Total Cost
Max. Building Appearance
Min. Loss of Employees
Material Handling Cost
Min. Distance to Key Facilities
Appearance
Max. Space
Select Warehouse
Warehouse Fl. Space
Soci-economic
Natural Environment
Cost
Timelines
Intl Cooperation
Diversion
ES&H
Human H&S
Theft
Non-proliferation
Operational Effect
Pu Disposition
Irreversibility