Abstract Policymakers face persistent uncertainty, and they often find static approaches insufficient for long-term effectiveness. In this article, we address the following question: how can policymakers design adaptive policies as evolving cyber–physical–social systems that remain effective amid dynamic conditions? As our contribution, we develop a goal-oriented inverse design method (GOIDM) that enables policymakers to work backward from well-defined long-term objectives to identify feasible, robust policy interventions. Unlike traditional forward optimization or scenario-based approaches, in GOIDM, we integrate quantitative robustness metrics directly into the formulation through design capability indices (DCIs), enabling systematic evaluation of policy performance under uncertainty. We extend inverse design principles from engineering into policy contexts, and we define transition states, adaptation rates, and feedback-driven goal updates. We demonstrate GOIDM through a case study on sustainable development in the United Arab Emirates, where policymakers balance economic growth, environmental sustainability, and social equity. Using historical data and adaptive feedback, we achieve quantifiable results: an average adaptation rate of 0.042 across policy goals, DCI values exceeding 1.5 for all robustness constraints, and dynamic target adjustments, including a 23.8% increase in electricity consumption targets and a 5.3% increase in employment goals based on the observed 2023 transition state performance. We illustrate how GOIDM enables systematic adaptation, satisficing solution space identification via an interpretable self-organizing map, and iterative refinement toward 2030 targets. This framework is generalizable to domains such as resource allocation, supply chain optimization, and public health planning, where adaptive, data-driven decision-making under uncertainty is critical.
Mandegari et al. (Thu,) studied this question.