MOReCriT: Understanding Multi-Objective Optimization Results in Resource Criticality Analysis through Visual Analytics

Energy transition technologies depend on an expanding set of raw materials that are geographically concentrated and often sourced from politically or economically unstable regions, introducing substantial supply risks. To account for these uncertainties, multi-objective optimization models incorporate material-related criticality alongside system costs, enabling the exploration of trade-offs between economic performance and resource risk. Although this trade-off is expressed in two objective dimensions, each Pareto-optimal solution represents a high-dimensional configuration of technological and capacity decisions shaped by numerous criticality assumptions. Existing tools typically rely on isolated visualizations of objective values or tabular inspection of these complex results. Such static plots and fragmented material flow visualizations make cross-cutting reasoning across the trade-off frontier unwieldy and time-consuming. Thus, we present MOReCriT, a visual analytics tool for the interactive exploration of solution spaces derived from multi-objective optimization. The system enables experts in resource and energy modeling to interactively analyze Pareto fronts, trace solutions back to their underlying assumptions, and compare alternative technology pathways through coordinated, interlinked views. We evaluate MOReCriT with four domain experts. The results indicate that integrating established visual representations into coordinated, interlinked views substantially improves the interpretability of high-dimensional Pareto spaces. Compared to existing analysis workflows, our approach enables faster exploration and supports informed refinement of optimization parameters by making the impact of modeling assumptions explicit. This also revealed previously overlooked patterns in which certain technologies did not follow the expected proportional shift along the cost–criticality trade-off, obscured by the complexity of the result space.