Fast-Track Consensus Study on Foundational Research Gaps and Future Directions for Digital Twins

点击翻译按钮获取中文摘要

The National Academies of Sciences, Engineering, and Medicine is undertaking a study to identify needs and opportunities to advance the mathematical, statistical, and computational foundations of digital twins in applications across science, medicine, engineering, and society. A digital twin is a computer model that changes over time to represent the structure or behavior of a unique physical entity, such as a manufacturing process, piece of equipment, or even a person. Based on data inputs, the digital twin can be used to gain insight into present and future states of the physical twin. The exploration and use of digital twins is growing across domains, but many state-of-the-art digital twins are largely the result of custom implementations that require considerable deployment resources and a high level of expertise. Due to the individualized nature of many digital twin implementations, the relative maturity of digital twins varies significantly across problem spaces. Moving from one-off digital twins to digital twin implementations at scale will involve addressing foundational mathematical, statistical, and computational gaps. This study aims to highlight these critical research gaps and provide options to address them with the goal of advancing the use of digital twins across disciplinary communities.The proposed study by the National Academies of Sciences, Engineering, and Medicine, will highlight needs and opportunities to advance the mathematical, statistical, and computational foundations of digital twins in applications across science, medicine, engineering, and society. A digital twin assimilates observational data and uses this information to continually update its internal models so that they reflect the evolving physical system. The digital twin is therefore continuously improving and provides a dynamic digital history of the physical entity. These core functionalities can be augmented with feedback control and artificial intelligence, combined with ensembles of similar twins, or used in tandem with other predictive tools to analyze and diagnose operational states and to optimize performance under real-world conditions. Utilization of digital twins varies across disciplines. This study will address the following: (1) diverging definitions of digital twins and domain-inspired use cases; (2) foundational mathematical, statistical, and computational gaps for the continued development of digital twins; (3) best practices for digital twin development and use; and (4) opportunities to move the community and state of practice forward. Three domain-specific workshops will be held to explore the methods, practices, use cases, and challenges for the development and use of digital twins—focus areas include biomedical domains, Earth and environmental systems, and aerospace engineering. Four reports will be released by the National Academies during the course of this 18-month study: three short summaries of the domain-specific workshops and a consensus report focused on the cross-cutting foundational research gaps and future directions for digital twins.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.