SUMMARY

Decision making is necessary to provide a synthesis scheme to design activities and devise the most preferred design alternative. There exist several methods devised to model designer preferences in a graphical manner to aid the decision making process. For instance, the Conjoint Analysis has been proven effective for various multi-attribute design problems by utilizing a ranking- or rating-based approach along with the graphical representation of the designer preference. However, the ranking or rating of design alternatives can be inconsistent from different users and it is often difficult to get customer responses in a timely fashion. The high number of alternative comparisons required for complex engineering problems can be exhausting for the DM. In addition, many design objectives can have interdependencies that can increase complexity and uncertainty throughout the decision making process. The uncertainties in the attainment of subjective data along with increased model uncertainties can reduce the reliability of decision analysis results. To address these issues, this paper proposes the use of a new technique, the Improved Conjoint Analysis, to enable the modeling of designer preferences and trade-offs under the consideration of uncertainty. Specifically, a simulation-based ranking scheme is implemented and incorporated into the traditional process of the Conjoint Analysis. The proposed ranking scheme can reduce user fatigue and provide a better schematic decision support process. In addition, the incorporation of uncertainty in the design process provides the capability of producing robust or reliable products. The efficacy and applicability of the proposed framework are demonstrated with the design of a cantilever beam, a power-generating shock absorber, and a mesostructured hydrogen storage tank.