Let decision-makers direct the search for robust solutions: An interactive framework for multiobjective robust optimization under deep uncertainty

The robust decision-making framework (RDM) has been extended to consider multiple objective functions and scenarios. However, the practical applications of these extensions are mostly limited to academic case studies. The main reasons are: (i) substantial cognitive load in tracking all the trade-offs across scenarios and the interplay between uncertainties and trade-offs, (ii) lack of decision-makers’ involvement in solution generation and confidence. To address these problems, this study proposes a novel interactive framework involving decision-makers in searching for the most preferred robust solutions utilizing interactive multiobjective optimization methods. The proposed interactive framework provides a learning phase for decision-makers to discover the problem characteristics, the feasibility of their preferences, and how uncertainty may affect the outcomes of a decision. This involvement and learning allow them to control and direct the multiobjective search during the solution generation process, boosting their confidence and assurance in implementing the identified robust solutions in practice. • The proposed interactive framework involves decision-makers in the solution process. • Decision-makers learn the problem’s characteristics and limitations. • Decision-makers can explore how uncertainty may affect the outcomes of a decision. • This involvement/learning allows decision-makers to direct the multiobjective search. • A new hypothetical problem is proposed for benchmarking multiobjective DMDU methods.