Abstract
A digital twin (DT) is a virtualization of a physical asset built upon a set of computational models that dynamically update to persistently mirror a unique asset of interest throughout its operational lifespan, enabling informed decisions that realize value. This talk presents a progressive view of recent advancements in predictive DTs, emphasizing the integration of probabilistic graphical models, Bayesian inference, and risk-aware optimal control spanning applications from civil infrastructure to aerospace systems.
The first part introduces a DT framework for civil engineering structures, where the asset-twin system is formalized via a dynamic Bayesian network. Reduced order models and deep learning are leveraged to assimilate structural health data from sensors and support condition-based maintenance and lifecycle optimization [1].
The second part extends this formulation to incorporate robustness with respect to low-probability, high-consequence events using parametric Markov decision processes. Here, state transition dynamics are treated as random variables, and risk functionals are used to define robust control objectives. By embedding probabilistic model checking within the DT framework, we enable adaptive policy refinement that balances operational cost against probabilistic risk exposure [2].
Finally, we propose a modular DT architecture for multi-spacecraft on-orbit servicing missions, supporting decentralized state estimation, subsystem-level health monitoring, and coordinated control. The framework handles uncertainties in the kinematic states as well as the health of the propulsion subsystem and demonstrates the generation of mission-robust policies through simulation of rendezvous scenarios involving partially degraded assets [3].
This progression highlights the scalability and adaptability of probabilistic DTs across domains with stringent performance and reliability requirements.
[1] M. Torzoni, M. Tezzele, S. Mariani, A. Manzoni, and K. E. Willcox. “A digital twin framework for civil engineering structures”, Computer Methods in Applied Mechanics and Engineering, vol. 418, p. 116584, 2024. doi: 10.1016/j.cma.2023.116584.
[2] M. Tezzele, S. Carr, U. Topcu, and K. Willcox. “Adaptive planning for risk-aware predictive digital twins”, in Physics-based and Data-driven Modeling for Digital Twins, Eds. K. Cherifi and I.V. Gosea, SEMA SIMAI Springer Series, 2025. arXiv:2407.20490
[3] S. Henao-Garcia, M. Kapteyn, K. E. Willcox, M. Tezzele et al. “Digital-Twin-Enabled Multi-Spacecraft On-Orbit Operations”, in AIAA SCITECH 2025 Forum, American Institute of Aeronautics and Astronautics, Inc., 2025. doi: 10.2514/6.2025-1432.
Contatto:
