1

Introduction

Digital twins, which are virtual representations of natural, engineered, or social systems, hold immense promise in accelerating scientific discovery and revolutionizing industries. This report aims to shed light on the key research needs to advance digital twins in several domains, and the opportunities that can be realized by bridging the gaps that currently hinder the effective implementation of digital twins in scientific research and industrial processes. This report provides practical recommendations to bring the promise of digital twins to fruition, both today and in the future.

THE SIGNIFICANCE OF DIGITAL TWINS

Digital twins are being explored and implemented in various domains as tools to allow for deeper insights into the performance, behavior, and characteristics of natural, engineered, or social systems. A digital twin can be a critical tool for decision-making that uses a synergistic combination of models and data. The bidirectional interplay between models and data endows the digital twin with a dynamic nature that goes beyond what has been traditionally possible with modeling and simulation, creating a virtual representation that evolves with the system over time. The use cases for digital twins are diverse and proliferating—including applications in biomedical research, engineering, atmospheric science, and many more—and their potential is wide-reaching.

Digital twins are emerging as enablers for significant, sustainable progress across industries. With the potential to transform traditional scientific and industrial practices and enhance operational efficiency, digital twins have captured the attention and imagination of professionals across various disciplines and

sectors. By simulating real-time behavior, monitoring performance to detect anomalies and exceptional conditions, and enabling predictive insights and effective optimizations, digital twins have the capacity to revolutionize scientific research, enhance operational efficiency, optimize production strategies, reduce time-to-market, and unlock new avenues for scientific and industrial growth and innovation.

Digital twins not only offer a means to capture the knowledge and expertise of experienced professionals but also provide a platform for knowledge transfer and continuity. By creating a digital representation of assets and systems, organizations can bridge the gap between generations, ensuring that critical knowledge is preserved and accessible to future workforces and economies.

In the present landscape, “digital twin” has become a buzzword, often associated with innovation and transformation. While there is significant enthusiasm around industry developments and applications of digital twins, the focus of this report is on identifying research gaps and opportunities. The report’s recommendations are particularly targeted toward what agencies and researchers can do to advance mathematical, statistical, and computational foundations of digital twins. Scientific and industrial organizations are eager to explore the possibilities offered by digital twins, but gaps and challenges often arise that impede their implementation and hinder their ability to fully deliver the promised value. Organizations eager to use digital twins do not always understand how well the digital twins match reality and whether they can be relied on for critical decisions—much of this report is aimed at elucidating the foundational mathematical, statistical, and computational research needed to bridge those gaps. Other technological complexities pose challenges as well, such as network connectivity and edge computing capabilities, data integration issues and the lack of standardized frameworks or data structures, and interoperability among various systems. Additional challenges include organizational aspects, including workforce readiness, cultural shifts, and change management required to facilitate the successful adoption and integration of digital twins. Furthermore, ensuring data security, cybersecurity, privacy, and ethical practices remains a pressing concern as organizations delve into the realm of digital twins.

COMMITTEE TASK AND SCOPE OF WORK

This study was supported by the Department of Energy (Office of Advanced Scientific Computing Research and Office of Biological and Environmental Research), the Department of Defense (Air Force Office of Scientific Research and Defense Advanced Research Projects Agency), the National Institutes of Health (National Cancer Institute, National Institute of Biomedical Imaging and Bioengineering, National Library of Medicine, and Office of Data Science Strategy), and the National Science Foundation (Directorate for Engineering and Directorate for Mathematical and Physical Sciences). In collaboration with the National

Academies of Sciences, Engineering, and Medicine, these agencies developed the study’s statement of task (see Appendix A), which highlights important questions relating to the following:

• Definitions of and use cases for digital twins;
• Foundational mathematical, statistical, and computational gaps for digital twins;
• Best practices for digital twin development and use; and
• Opportunities to advance the use and practice of digital twins.

The National Academies appointed a committee of 16 members with expertise in mathematics, statistics, computer science, computational science, data science, uncertainty quantification, biomedicine, computational biology, other life sciences, engineering, atmospheric science and climate, privacy and ethics, industry, urban planning/smart cities, and defense. Committee biographies are provided in Appendix F.

The committee held several data-gathering meetings in support of this study, including three public workshops on the use of digital twins in atmospheric and climate sciences (NASEM 2023a), biomedical sciences (NASEM 2023b), and engineering (NASEM 2023c).

REPORT STRUCTURE

This report was written with the intention of informing the scientific and research community, academia, pertinent government agencies, digital twin practitioners, and those in relevant industries about open needs and foundational gaps to overcome to advance digital twins. While the range of challenges and open questions around digital twins is broad, it should be noted that the focus of this report is on foundational gaps. The report begins by defining a digital twin, outlining its elements and overarching themes, and articulating the need for an integrated research agenda in Chapter 2. The next four chapters expound on the four major elements of a digital twin as defined by the committee: the virtual representation, the physical counterpart, the feedback flow from the physical to the virtual, and the feedback flow from the virtual to the physical. In Chapter 3, fitness for purpose, modeling challenges, and integration of digital twin components for the virtual representation are discussed. Chapter 4 explores the needs and opportunities around data acquisition and data integration in preparation for inverse problem and data assimilation tasks, which are discussed in Chapter 5. Automated decision-making and human–digital twin interactions, as well as the ethical implications of making decisions using a digital twin or its outputs, are addressed in Chapter 6. Chapter 7 looks at some of the broader gaps and needs to be addressed in order to scale and sustain digital twins, including cross-community efforts and workforce challenges. Finally, Chapter 8 aggregates all of the findings, conclusions, gaps, and recommendations placed throughout the report.

REFERENCES