Edge Intelligence and Digital Twin Synergy for Low-Latency Autonomous Control in Dynamic Cyber-Physical Systems
Abstract
Cyber-Physical Systems (CPS) are evolving towards higher autonomy and real-time responsiveness, posing stringent demands on low-latency decision-making and dynamic environmental adaptation. Traditional cloud-centric control architectures suffer from inevitable network delays and bandwidth bottlenecks, limiting their applicability in time-critical scenarios. This paper proposes a novel synergistic framework integrating Edge Intelligence (EI) and Digital Twin (DT) for autonomous control systems. By deploying lightweight AI models at the network edge and establishing high-fidelity virtual replicas of physical assets, the framework enables real-time perception, local decision-making, and predictive control. This study elaborates on the architectural design, communication protocols, and security mechanisms of the proposed framework. Through experimental validation on smart grid load regulation and autonomous mobile robot navigation, the results demonstrate that the EI-DT synergy reduces end-to-end latency by over 60% and improves control stability by 30% compared to cloud-based approaches. This research provides a viable solution for latency-sensitive and dynamic autonomous control applications.
