Adaptive digital twins for predictive decision-making: Online Bayesian learning of transition dynamics
arXiv:2512.13919v3 Announce Type: replace Abstract: This work shows how adaptivity can enhance value realization of digital twins in civil engineering. We focus on adapting the state transition models within digital twins represented through probabilistic graphical models. The bi-directional interaction between the physical and virtual domains is modeled using dynamic Bayesian networks. By treating state transition probabilities as random variables endowed with conjugate priors, we enable hierarchical online learning of transition dynamics from a state to another through effortless Bayesian up
The increasing complexity of operational systems and the advancements in Bayesian inference and real-time data processing enable more sophisticated approaches to predictive modeling.
This development offers a pathway to more resilient and efficient infrastructure through self-adapting digital models, significantly impacting capital expenditure and operational reliability in critical sectors.
The ability for digital twins to autonomously and continuously update their internal models based on real-world data allows for proactive decision-making rather than reactive responses, moving towards truly adaptive systems.
- · Civil Engineering Firms
- · Infrastructure Operators
- · AI/ML Platform Providers
- · Predictive Maintenance Software Developers
- · Traditional Inspection Services
- · Reactive Maintenance Providers
- · Inefficient Asset Management Practices
Adaptive digital twins will reduce unexpected failures and extend the lifespan of physical assets by continuously optimizing maintenance schedules.
This enhanced predictability will lead to lower insurance premiums for infrastructure projects and increased investment in long-term, data-driven planning.
The widespread adoption of self-adapting digital twins could redefine liability frameworks in infrastructure, shifting responsibility towards AI-driven predictive capabilities.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.LG