.png){kind=icon}
.png){kind=icon}
Return
A digital twin-driven hybrid intelligence framework for resilient and scalable cloud–edge orchestration in industrial cyber–physical systems
.png){kind=icon}
PRE
AI
M
DOI：10.1016/j.future.2026.108744.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
The rapid growth of the industrial internet of things and cyber–physical systems has intensified the demand for intelligent, scalable, and resilient cloud–edge orchestration under highly dynamic and uncertain operating conditions. While digital twins offer unprecedented opportunities for real-time system awareness and predictive control, existing studies often treat digital twins as passive monitoring or visualization tools, limiting their impact on autonomous decision making and decentralized intelligence. This paper proposes a digital twin-driven hybrid intelligence framework for adaptive and resilient task offloading in industrial cloud–edge cyber–physical systems. A functional digital twin continuously synchronizes the states of tasks, network conditions, and computational resources, enabling predictive system awareness and closed-loop control across the internet of things–edge–cloud continuum. Building upon the digital twin, a hybrid intelligence engine integrates machine-learning-based performance prediction with multi-objective optimization to jointly minimize end-to-end latency, energy consumption, and operational cost, while maximizing reliability, scalability, and system resilience. The proposed framework supports proactive offloading, failure-aware task migration, and selective re-optimization, enabling partially decentralized and scalable decision making suitable for large-scale industrial environments. Extensive simulation experiments under dynamic workloads, heterogeneous resources, and stochastic failures demonstrate consistent performance gains over cloud-centric, edge-only, and non-digital-twin baselines across multiple metrics. The results confirm that tightly coupling digital twins with hybrid intelligence mechanisms provides a practical pathway toward autonomous, resilient, and scalable orchestration in next-generation industrial cyber–physical systems and decentralized internet of things infrastructures.
Journal
F
IF:
6.1
Papers:
6.8K
Citations:
2.3W
Organization
No organization information available
Cited Papers