Implicit and Explicit Modeling of Uncertainty in Early Design Stages of Product Design: A Comparative Study

During product development, many decisions have to be made which affect the entire product life cycle. Especially in early phases in product development these decisions are subject to uncertainties when involving customer specific requirements. To support the designer in developing robust products that are insensitive to uncertainty, this paper compares two opposing approaches for modeling and assessing uncertainties: an explicit method in form of a Bayesian decision network with utility functions and an implicit method in form of a robust numerical optimization with sampling of a model-based product representation. Hence, a quantitative and qualitative comparison is provided regarding the modeling expense and the usability in early design stages. Regarding the integration into the product development process we propose a combination of the two approaches as complementary tools.