Mathematics of Digital Twins
Abstract
Design and deployment of Digital Twins (DTs) of a physical asset require seamless integration of numerical models and observational data, both of which often present themselves with various fidelities. Despite considerable advancement in hardware, computing and data analytics, complex systems of practical importance to the DoD missions defy their robust and reliable high-fidelity, DT-like representations. The main culprit in this failure is the chasm between the complexity of the relevant phenomena and the lack of targeting capabilities of the current modeling techniques. We propose to develop a mathematical and numerical paradigm of a DT that is synchronized with a physical asset, powered by model and data integration, evolved by data acquisition, and optimized by mission outcomes, throughout the life cycle of the physical asset. The DT includes models, data, connectivity, analytics, and actions. The DT can describe the past, the present, and the future of the physical asset, while being optimized and controlled by specific mission objectives. Our work will streamline the simulation of complex interconnected systems, which is crucial to DoD mission. These systems encompass software and hardware components, multiple levels of physics, varying scales, complexities, decision-making processes, and uncertain structures. Our fully integrated DT framework will accelerate and automate the design, engineering, analysis, reverse engineering, and control of such systems, while facilitating decision-making under uncertainty.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Feb 06, 2025
- Source ID
- FA95502410237
Entities
People
- Daniel M. Tartakovsky
Organizations
- Air Force Office of Scientific Research
- Stanford University
- United States Air Force