2“System Thinking” Design Heuristic Claim of SEBetter systems engineering leads toBetter system quality/valueLower costShorter scheduleSYSTEMDESIGNDETAILPRODUCTIONINTEGRATIONTESTTraditional DesignTimeRiskSavedTime/Cost“System Thinking” DesignTimeRisk
3NASA Tracking 1980s % Investment in System Engineering Effort (SEE) Total Program Overrun32 NASA ProgramsIn the late 1980s, Werner Gruhl of the NASA Comptroller’s office set out to improve cost estimation on NASA projects. As part of his effort, he mandated that NASA projects track costs to a common Work Breakdown Structure (WBS) that would allow gathering data across projects. This additional tracking was funded as part of each project. Over several years of live and historic projects, he developed the chart shown above that shows the impact of front-end investment (i.e. system definition and analysis) on the accuracy of cost estimation.% Investment inSystem Engineering Effort (SEE)
4Time-Phased Sensitivity of SE to Total System Life Cycle Cost %Commitment to Technology,Configuration, Performance, Cost, etc.10075Cost Incurred50System-Specific Knowledge25Ease of ChangeNEDConceptual-PreliminaryDesignDetailDesign andDevelopmentConstructionAnd/orProductionSystem Use,Phase out, and DisposalCommitment, System-Specific Knowledge, and CostSystems Engineering is important early in a program to influence the design when incurred costs are low and design changes are easy.
5Concept and Technology Development PreliminarySystem DesignPhaseConceptual SystemDesign PhaseDetail Design and DevelopmentPhaseSystems EngineeringHighIndividual Design DisciplinesDesign InfluenceAs system design and development progresses during system development, the flexibility to implement system design changes is reduced and the cost impact of introducing changes is increased. Systems engineering has its greatest opportunity to influence design decisions earlier in systems design and development when the system design is still in evolution.LowSystems Design and Development ProgressSystems Engineering is important early in a program to influence the design when incurred costs are low and design changes are easy.
6Conclusions SE effort improves development quality Cost, schedule, subjectiveHypothesis is supported by the dataOptimum SE effort is 10-15%Matches data from NASA projectsCost, schedule overruns are minimizedHowever, note wide dispersion of dataQuality of the SE effort mattersLower quality SE reduces effectiveness