EAGER: Adaptive Digital Twinning: An Immersive Visualization Framework for Structural Cyber-Physical Systems

美国的基础设施系统包括一系列对国家经济、安全和完整性至关重要的资产、系统和网络。每个社区的成员，从个人家庭到全球公司，都依赖这些基础设施系统来蓬勃发展并保持高质量的生活。这个基础设施是复杂的、相互依赖的、相互联系的和多样化的，包括我们饮用的水、我们使用的电力、移动我们的交通服务以及连接我们的通信系统。今天为社会服务的这些基础设施系统中的许多都是在第二次工业时代建造的，在许多情况下，由于用于保护它们的资源减少，它们处于失修状态。虽然我们继续改进设计方法并实施更可持续的保护战略，但现代基础设施系统仍沿用其早期发展中使用的许多历史方法，尚未实现现代化。随着社会对技术的依赖持续增长，基础物理基础设施系统必须得到保留，但也必须进行现代化，以确保这些系统能够作为未来所需的智能和灵活的网络物理系统(CP)。该项目将探索一种高风险/高回报的方法，使用人工智能知情的数字双胞胎来实现基础设施系统的现代化。基础设施系统的数字孪生将在物理世界的可测量数据和虚拟世界中的模拟过程之间形成协作反馈环路，提供对更广泛的CPS框架的特定领域的适应，为决策提供信息。该项目应用于大型结构系统的领域，将测试这样的假设，即使用这些结构系统的数字孪生表示的沉浸式参与将使参与者能够观察、交互并将与这些系统相关联的复杂行为机制与其操作环境相关联。为了验证这一假设，研究设计将探索一系列技术创新，包括建立人工智能模型，以模拟基于模拟的结果和基于实验的测量。利用这些技术创新，我们将能够1)了解人工智能公式模型在多大程度上能够有效地模拟大型结构系统的模拟和实验的复杂力学行为；2)评估人工智能公式模型的发展在多大程度上实现了数字孪生兄弟所需的模拟和实验之间的实时、双向交互；以及3)表征在沉浸式环境中部署人工智能公式模型如何允许最终用户观察和表征在役结构系统的运行状态。这项工作的成功将通过融合对这些复杂的网络物理系统的实验和数字描述，并创造必要的新过程来克服行为的理论描述和现实生活中的结构响应之间存在的知识差距，为结构系统在其运行环境中的实时决策奠定基础。这个项目的结果将通过被推荐的期刊、会议记录和学生论文传播给更广泛的研究社区。这个奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

Infrastructure systems in the United States include a diverse series of assets, systems, and networks that are vital to the nation’s economy, security, and integrity. Members of every community, ranging from individual families to global corporations, rely on these infrastructure systems to thrive and maintain a high quality of life. This infrastructure is complex, interdependent, interconnected, and diverse, encompassing the water that we drink, the power that we use, the transportation services that move us, and the communication systems that connect us. Many of these infrastructure systems that serve society today were built during the second industrial age, and in many cases are in a state of disrepair with decreasing resources to preserve them. While we have continued to improve design approaches and implement more sustainable preservation strategies, modern infrastructure systems still follow many of the historical approaches used in their early development and have not been modernized. As societal dependence on technology continues to grow, the underlying physical infrastructure systems must be preserved, but also modernized to ensure that these systems are equipped to serve as the smart and agile cyber-physical systems (CPS) the future demands. This project will explore a high-risk/high reward approach to modernizing infrastructure systems using artificial intelligence-informed digital twins. The digital twinning of an infrastructure system will form a collaborative feedback loop between the measurable data of the physical world and simulated processes in the virtual world, providing a domain-specific adaptation of the broader CPS framework necessary to inform decision-making.Applied to the domain of large-scale structural systems, this project will test the hypothesis that immersive engagement using a digital twin representation of these structural systems will enable participants to observe, interact, and contextualize the complex behavior mechanisms associated with these systems in their operational environment. To test the hypothesis, the research design will explore a series of technology innovations including the formulation of artificial intelligence models to emulate both simulation-based results and experiment-based measurements. Leveraging these technology innovations, we will be able to 1) understand to what extent can artificial intelligence formulated models effectively emulate the complex mechanical behaviors of simulation and experimentation of large-scale structural system; 2) evaluate to what extent does the development of artificial intelligence formulated models enable the real-time, bi-directional interaction between simulation and experimentation required of a digital twin; and 3) characterize how the deployment of artificial intelligence formulated models within an immersive environment allow end users to observe and characterize operational states of an in-service structural system. Success of this work will be realized through the fusion of experimental and numerical descriptions of these complex cyber physical systems and the creation of novel processes necessary to overcome the knowledge gap that exists between the theoretical descriptions of behavior and real-life structural response, forming a foundation for real-time decision-making for structural systems in their operational environments. Results from this project will be disseminated to the broader research community through refereed journals, conference proceedings, and student dissertations.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.