Answers are grounded in this paper and cite their sources.
This special issue reports on advancements in data-driven design methods, focusing on uncertainty quantification in engineering systems.
Advances in data acquisition, high-performance computing, and machine learning (ML) are reshaping the way engineers conceive, analyze, and optimize complex mechanical systems. Across applications as diverse as materials discovery, manufacturing process optimization, digital twins, robotics, and resilient network design, the engineering community increasingly relies on data-driven methods to model systems characterized by high dimensionality, evolving states, and heterogeneous sources of uncertainty. As these systems grow in complexity, uncertainty quantification (UQ) and design under uncertainty have become foundational to ensuring that engineered solutions are robust, reliable, and trustworthy. This special issue of the ASME Journal of Mechanical Design was conceived to bring together fundamental contributions that expand the capabilities of UQ and uncertainty-aware design optimization in the age of modern ML. The call emphasized interpretability, scalability, multi-fidelity reasoning, and the integration of new paradigms—including generative AI, probabilistic deep learning, graph-based representations, and large language models—into engineering workflows. It also highlighted the growing need to address uncertainty in multiscale phenomena, complex manufacturing processes, dynamically evolving cyber-physical systems, and high-risk decision environments.
No takes yet. Share an insight, caveat, or question.
Bostanabad et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: