Digital Twin Entanglement: Enhanced measurement & reference benchmarking

Digital Twins (DTs) are a fundamental paradigm for monitoring Cyber–Physical Systems (CPS). However, accurately quantifying the entanglement between a DT and its Physical Twin (PT) remains a critical challenge. In heterogeneous CPS, maintaining this alignment is highly complex: telemetry streams are generated at varying frequencies, reflect differing criticality levels, and face constant disruption from network latency and computational interference. Existing monolithic entanglement metrics, such as ODTE, fail to jointly capture these dynamics; by treating all data loss uniformly and ignoring ordering violations, they mask subtle state desynchronization. To resolve these limitations, we propose the Enhanced Overall Digital Twin Entanglement (E-ODTE) metric. E-ODTE explicitly models the fine-grained structure of CPS telemetry by decomposing the evaluation into a granular, per-property representation. Furthermore, it introduces a data continuity component to penalize out-of-sequence updates and incorporates semantic weighting to encode operational criticality. For rigorous validation, we introduce the Digital Twin Entanglement Benchmark (DTEB), a deterministic reference baseline founded on strict update alignment. E-ODTE is evaluated within a real-time stress-testing framework, mapping native Real-Time Operating System (RTOS) scheduling priorities directly onto its weighting structure. Experimental validation utilizing a Franka Emika Panda robotic setup controlled by a PREEMPT_RT kernel demonstrates that E-ODTE consistently and faithfully tracks the DTEB baseline, even under severe workload preemption. Beyond its effectiveness as a reliable surrogate metric, E-ODTE acts as a powerful diagnostic instrument, successfully exposing latent telemetry buffer anomalies that remain invisible to conventional evaluation approaches.