Digital Twins for Marine Robotic Platforms: Architectures, Applications, and Challenges Across AUVs, ROVs, USVs, Underwater Gliders, and Emerging Platforms

Digital twin (DT) research spans autonomous underwater vehicles (AUVs), remotely operated vehicles (ROVs), unmanned surface vessels (USVs), underwater gliders, and emerging marine robotic platforms; however, reported systems differ widely in terms of their physical–virtual coupling and validation maturity. A structured Web of Science search retrieved 116 records, of which 107 were retained for analysis. This review synthesizes architectures, enabling technologies, applications, coupling modes, validation settings, and deployment barriers across the perception–decision–control chain. The reported workflows address model construction, multiphysics simulation, virtual testing, monitoring, control validation, and decision support. Within the purposively selected 11-case full-text audit, two cases were classified as DT-enabling methods or components, five as digital models or virtual prototypes, three as digital shadows, and one as a bidirectional DT prototype; none documented sustained operational deployment. Four constraints recurred across platforms: the trade-off between model fidelity and computation; uncertain marine environments; communication and synchronization limitations; and fragmented interfaces and standards. Physics-informed learning, multi-fidelity modeling, edge computing, and interoperable reference architectures provide candidate development paths, but the evidence suggests that the central gap lies less in producing sophisticated virtual models than in establishing sustained, interoperable, and bidirectional physical–virtual operation.