Download PDF Winter Bridge on The Grainger Foundation Frontiers of Engineering December 13, 2024 Volume 54 Issue 4 This issue features articles by The Grainger Foundation US Frontiers of Engineering 2024 symposium participants. The articles examine cutting-edge developments in microbiology and health, artificial intelligence, the gut-brain connection, and digital twins. Beyond Digital Twins: Realizing Value through Digital Continuity Thursday, December 12, 2024 Author: Olivia J. Pinon Fischer and Dimitri N. Mavris The significant benefits of digital twins are clear, but more work needs to be done for the full potential of digital twins to be realized. Digital twins have taken center stage, revolutionizing how industries interact with physical systems by offering dynamic and virtual representations of complex systems and processes that can sense, diagnose, and predict the state of physical entities. A digital twin consists of three essential elements, as represented in figure 1: a virtual representation (model), a physical realization (article), and the transfer of data/information between the physical and virtual intended to increase the value of the physical article and its use (AIAA Digital Engineering Integration Committee 2020). As such, a digital twin can be succinctly defined as “a virtual representation of a connected physical asset” (AIAA Digital Engineering Integration Committee 2020). This connectivity, or bidirectional interaction (from the physical to the virtual and from the virtual to the physical), is central to the digital twin construct (NASEM 2023a), as it helps differentiate a digital twin, “an accurate description of an article that changes over time,” from a model, “a snapshot of the behavior of an object at a specific moment” (Wright and Davidson 2020). Digital twins have demonstrated their potential to revolutionize decision-making across science, engineering, and society as a whole (AIAA Digital Engineering Integration Committee 2023b; Ferrari and Willcox 2024; Kobryn 2019; NASEM 2023a; Pinon Fischer et al. 2022; Rasheed et al. 2020). Their development, applications, and contributions span multiple scales, from materials to individual components to complex systems of systems and missions outcomes, and extend across the boundaries of disciplinary teams and organizations across all phases of the product lifecycle (figure 2). In this context, the emphasis has recently been on the development and deployment of digital twins to support trades and decision-making across system, system-of-systems, and mission levels (Cantrell et al. forthcoming; Reddy et al. 2023; Schlichting et al. 2024) as well as across multiple phases of the lifecycle (Cantrell et al. forthcoming). Key Considerations The literature is abundant with publications that explore the benefits of digital twins, with a steady stream of studies emerging across diverse domains, sectors, and industries. However, as mentioned in Ferrari and Willcox 2024, those benefits can only be realized if the digital twin is purposefully “conceptualized, architected, designed, built, deployed and sustained.” Digital Twin Definition and Development The successful definition and development of a digital twin hinge critically on the ability to properly frame the underlying problem and articulate the ultimate purpose of the digital twin. Despite this, many past and ongoing efforts in digital twin development have adopted ad-hoc methodologies, where the focus prematurely shifts to building models without a thorough understanding or articulation of the problem space (Lu et al. 2020). These approaches often result in suboptimal outcomes, including poor model calibration due to inadequate or irrelevant data, limited interoperability between systems, and reduced model reusability. More critically, such methods frequently lead to the creation of models that provide a solution to the wrong problem and consequently fail to address the core questions and needs of the stakeholders. The consequences of inadequate problem framing are well documented in the literature, with several studies and panels emphasizing its centrality to successful digital twin implementation (Pinon Fischer et al. 2022; Rasheed et al. 2020; NASEM 2023b). The importance of problem framing is discussed extensively in Martin 2019 and involves a number of steps, as discussed below. Define Use Cases and User Groups: This step centers on specifying the key stakeholders and end-users of the digital twin, along with their unique perspectives and objectives. It includes identifying the primary issues to be investigated, the questions the digital twin is intended to address, and the decisions it will support. These elements collectively establish the digital twin’s purpose and guide the analytical approaches needed, subsequently informing data and modeling requirements (Boschert and Rosen 2016). Determine Scope and Context: The next step involves gathering and documenting the scope, context, perspectives, operating environment, key scenarios, primary constraints, existing investments in tools and methodologies, and other foundational assumptions (Martin 2019). Identify Data and Information Needs: Data plays a central role in digital twins, yet managing data presents numerous challenges, particularly in areas such as data integration, synchronization, and quality assurance. Additionally, challenges arise in acquiring data and identifying suitable solutions for data storage, access, and security (Defense Business Board 2024; Margaria and Ryan 2023). The quantity and nature of available data are also critical, as they shape the appropriate modeling approaches to pursue. Identify Modeling Needs and Capabilities: This represents a critical step as the choice of modeling tools, languages, and platforms has significant implications on the future usability, scalability, extensibility, or maintainability of not only the models themselves, but that of the digital twin as well. Other elements also need to be taken into consideration and those include (Pinon Fischer et al. forthcoming): Model Scope: The aspects to be modeled are determined by the purpose of the digital twin and the specific decisions it is intended to support. For example, if the goal is to inform maintenance needs for aircraft brakes, there is no need to include a full aerodynamic model of the aircraft. Digital twins have demonstrated their potential to revolutionize decision-making across science, engineering, and society as a whole. Nature of Models: The nature of the models to be developed and/or used should be directly linked to the purpose and desired analytical/reasoning capabilities (descriptive, diagnostic, predictive, and prescriptive) of the digital twin. Models can be of various types (descriptive, mathematical, physics-based, data-driven, discrete, physics-informed, etc.), all of which have advantages and limitations that need to be clearly understood and balanced against the nature/type of the system of interest, the amount of knowledge about the specific process to be modeled, the amount of data available to train, calibrate, verify, and validate models, the needed or required generalizability and explainability of the models as well as the desired/required level of accuracy to be achieved. Model Fidelity: The fidelity of the models developed should be sufficient to address the relevant use cases and tailored to the available data for model calibration. Supporting the Sustainable Operation of Digital Twins The long-term operation of digital twins, which are expected to remain functional over the entire lifecycle of the system they represent—often spanning several decades—presents significant challenges in software maintainability (NASEM 2023a). A major issue arises from the inevitable obsolescence of current operating systems and platforms, as the tools and codes used today will likely be incompatible with future technological environments in 30 to 50 years. Furthermore, the individuals tasked with maintaining these systems in the future may not have been involved in their original development, leading to knowledge gaps that complicate upkeep and troubleshooting (West and Blackburn 2017). Additionally, the cost of operating and maintaining digital twins over such extended periods is expected to surpass their initial development costs (West and Blackburn 2017), bringing into question the responsibility of this cost to organizations. Digital Twins as Part of a Larger Digital Ecosystem As discussed, data and models represent the essence of digital twins. Consequently, the development and implementation of digital twins lay on the existence of an authoritative source of truth (ASoT) that integrates data and models across the product lifecycle. In other words, a digital twin represents only one element within a larger digital ecosystem (OUSD[R&E] 2023). A digital ecosystem encompasses the interconnected infrastructure, environment, and methodologies (processes, methods, and tools) that enable the storage, access, analysis, and visualization of data and models throughout a system’s life. An interoperable ecosystem facilitates the seamless flow of information, connecting various stakeholders and enabling collaborative decision-making from design and manufacturing to operations and sustainment. The successful definition and development of a digital twin hinge critically on the ability to properly frame the underlying problem and articulate the ultimate purpose of the digital twin. The foundation of a digital ecosystem lies in a consistent digital thread. A digital thread is a “linked set of digital artifacts whose consistency is actively managed over the life cycle of a product, process, or system” (AIAA Digital Engineering Integration Committee 2023a). This thread acts as the backbone of the ecosystem, enabling multiple digital twins to share authoritative data to better model and simulate the interactions and behaviors within the complex system-of-systems of interest. This resulting digital continuity extends the value proposition of digital twins to facilitate insights and optimizations not only at the level of discrete parts but also within integrated, multi-layered networks of systems and organizations, facilitating collaborative decision-making on a larger scale. Summary While much attention is rightly given to the development and benefits of digital twins, there is insufficient acknowledgment that their full potential cannot be realized without shared and interoperable reference architectures (Lockhart 2021), standards (Shao 2021), unified and consistent data sources, integrated tools, validated models, workflows, and robust data-sharing protocols. These elements are also foundational to the desired ability of multiple organizations and stakeholders to interact and collaborate openly within shared digital ecosystems. Such interoperability will be especially critical in addressing challenges involving complex systems of systems, whether in delivering timely, location-specific capabilities to warfighters or supporting sustained human presence on the Moon or Mars. In most cases, these ecosystems—and the integrated digital environments they rely on—will need for digital twins of various levels of fidelity and representing different laws of physics to be able to interoperate seamlessly. These environments have yet to be fully architected and developed, and they are urgently needed. References AIAA Digital Engineering Integration Committee. 2020. Digital Twin: Definition & Value. The American Institute of Aeronautics and Astronautics and Aerospace Industries Association. AIAA Digital Engineering Integration Committee. 2023a. Digital Thread: Definition, Value, and Reference Model. 2023. The American Institute of Aeronautics, the Astronautics and Aerospace Industries Association, and the International Association for Engineering Modelling, Analysis, and Simulation. AIAA Digital Engineering Integration Committee. 2023b. Digital Twin: Reference Model, Realizations & Recommendations. The American Institute of Aeronautics, the Astronautics and Aerospace Industries Association, and the International Association for Engineering Modelling, Analysis, and Simulation. Boschert S, Rosen R. 2016. Digital twin—the simulation aspect. In: Mechatronic Futures: Challenges and Solutions for Mechatronic Systems and their Designers, 59–74. Hehenberger P, Bradley D, eds. Springer. Cantrell SA, Margolis CH, Krauss MR, Pinon Fischer OJ, Mavris DN. Forthcoming. Digital Twins for Sustainment-Oriented Wargaming. Proceedings, AIAA SciTech Forum (2025), Orlando, Florida. Defense Business Board. 2024. Assessment of the Department of Defense: Creating a Digital Ecosystem. Department of Defense. Defense Science Board. 2024. Digital Engineering Capability to Automate Testing and Evaluation. Department of Defense, Office of the Under Secretary of Defense for Research and Engineering. Ferrari A, Willcox K. 2024. Digital twins in mechanical and aerospace engineering. Nature Computational Science 4(3):178–83. Kobryn PA. 2019. The Digital Twin Concept. The Bridge 49(4):16–19. Lockhart T. 2021. The Hitchhiker’s Guide to the Digital Engineering “Galaxy.” Utah Engineers Council Journal. Online at: https://uec-journal.thenewslinkgroup.org/the- hitchhiker s-guide-to-the-digital-engineering-galaxy/. Lu Q, Parlikad AK, Woodall P, Ranasinghe GD, Xie X, Liang Z, Konstantinou E, Heaton J, Schooling J. 2020. Developing a digital twin at building and city levels: Case study of West Cambridge campus. Journal of Management in Engineering 36(3):05020004. Margaria T, Ryan S. 2023. Data and data management in the context of digital twins. In: The Digital Twin, 253-78. Crespi N, Drobot AT, Minerva R. Springer. Martin JN. 2019. Problem framing: Identifying the right models for the job. INCOSE International Symposium 29(1):1–21. NASEM [National Academies of Sciences, Engineering, and Medicine]. 2023a. Foundational Research Gaps and Future Directions for Digital Twins. The National Academies Press. NASEM. 2023b. Opportunities and Challenges for Digital Twins in Engineering: Proceedings of a Workshop—In Brief. National Academies Press. OUSD(R&E)[Office of the Under Secretary of Defense for Research and Engineering]. 2023. DOD Instruction 5000.97. Pinon Fischer OJ, Matlik JF, Schindel WD, French MO, Kabir MH, Ganguli JS, Hardwick M, Arnold SM, Byar AD, Lewe J-H, and 5 others. 2022. Digital twin: Reference model, realizations, and recommendations. Insight 25(1):50–55. Pinon Fischer OJ, Sabri S, Chen Y. Forthcoming. Fundamentals of digital twins, modeling appraoches, and governance. In: Digital Twin: Fundamentals and Applications. Sabri S, Alexandridis K, Lee N. Springer Cham. Rasheed A, San O, Kvamsdal T. 2020. Digital twin: Values, challenges and enablers from a modeling perspective. IEEE Access 8:21980–22012. Reddy AV, Schlichting GS, Murphy DM, Oroz JC, Pinon Fischer OJ, Sudo AM, Mavris DN, French MO, Beattie C. 2023. JADC2 in a contested logistics environment: The role of digital twins. Proceedings, NATO AVT-369 Research Symposium on Digital Twin Technology Development and Application for Tri-Service Platforms and Systems, Oct 10–12, Bastad, Sweden. Schlichting GS, Reddy A, Murphy DM, Oroz JC, Pinon Fischer OJ, Mavris DN. 2024. Contested logistics operating under digital support. Proceedings, AIAA SCITECH 2024 Forum, Jan 8–12, Orlando, Florida. Shao G. 2021. Use case scenarios for digital twin implementation based on ISO 23247. National Institute of Standards: Gaithersburg, Maryland. West TD, Blackburn M. 2017. Is digital thread/digital twin affordable? A systemic assessment of the cost of DoD’s latest manhattan project. Procedia computer science 114:47–56. Wright L, Davidson S. 2020. How to tell the difference between a model and a digital twin. Advanced Modeling and Simulation in Engineering Sciences 7:13. About the Author:Olivia J. Pinon Fischer is principal research engineer and chief, Digital Engineering Division, and Dimitri N. Mavris is distinguished regents’ professor and director, both at Aerospace Systems Design Laboratory, the Daniel Guggenheim School of Aerospace Engineering, Georgia Institute of Technology.