Digital Twin-Enabled Vehicle Routing in Logistics and Transportation

Vehicle routing in modern logistics increasingly depends on live sensing: GPS and Internet-of-Things telemetry, roadside perception, and vehicle-to-everything (V2X) data streams change the operational state faster than static route plans can absorb. Digital twins have been proposed to link real-time sensing with routing decisions, yet the literature is fragmented across production, emergency, urban, supply-chain, and Internet of Vehicles (IoV) domains. This systematic review synthesizes 30 peer-reviewed studies of digital-twin-enabled vehicle routing retrieved by a two-stage search of Scopus, ScienceDirect, IEEE Xplore, and Google Scholar, and reported in accordance with PRISMA 2020. Each study was coded for domain, digital-twin role, coupling granularity, evidence maturity, evidence type, and evidence confidence. Four roles recur in the routing loop: monitoring and synchronization; simulation and training; adaptive re-optimization; and infrastructure, data governance, and model management. Twenty-five studies contribute primary evidence and five are conceptual or secondary review sources; simulation-only validation dominates (19 of 30 studies), and only three studies are validated on industrial or field data, so reported operational benefits remain largely unconfirmed. No meta-analysis was performed, because the corpus is too heterogeneous for pooled estimation. The review contributes a role taxonomy, an evidence-maturity mapping, and a routing-specific reference architecture in which synchronization quality, decision latency, and sensor-data governance are proposed as first-class design variables, together with three testable design hypotheses and a research agenda for sensing-driven, cyber-physical routing systems.