WP133: Probabilistic Consideration Method for Weight/Score-Based Decisions in Systems Engineering-Related Applications

Systems engineering (SE) uses a number of tools to structure project, product, and service development in an attempt to maximize workflow efficiencies and the satisfaction of project stakeholders and end users. Many of these tools—e.g., Quality Function Deployment (QFD) matrices—assign numerical weights and scores to SE considerations in order to organize and prioritize characteristics that are used to drive project decisions. Unfortunately, these weights and scores are susceptible to numerical variances due to human bias and other uncertainties. As a result, there is concern with regard to the robustness of these tools toward such variances and how this may impact major project decisions. To address this we introduce a probability-based method that uses computer simulations to evaluate the influence of numerical variance in weight/score-based SE tool implementations. This method can be used to develop confidence in weight/score-based decisions, determine if SE implementations should be re-evaluated, identify trends related to probabilistic variance, and set tolerances for such variances. The method can be expanded for use in other numerical based decision making frameworks that are susceptible to probabilistic variances.

Acknowledgements: This work is supported in part by the National Science Foundation via the Graduate Research Fellowship Program under grant 2008073003.