A perspective on component-level digital twins for surface marine propulsion system health monitoring: need and feasibility

Failures in marine propulsion systems pose significant safety and economic risks through loss-of-propulsion events. Although statistically infrequent, emerging maritime trends such as energy transition, alternative propulsion technologies, and Arctic navigation are expected to increase vulnerability to component failures, underscoring the need for advanced condition monitoring. This perspective article establishes component-level digital twins as an important but underexplored research direction. It introduces a scale-based characterisation of DTs from nano-level phenomena to process-level systems. It identifies the key barriers to practical implementation, including computational cost, dynamic modelling limitations, data scarcity and ownership constraints, communication challenges, and limited industry–academia collaboration. To address these gaps, the paper proposes the Digital Twin Research Hub (DTRH), a conceptual framework integrating risk-based prioritisation, physics-based modelling, surrogate modelling, and data-driven approaches. An experimentally validated bond graph-based simulation for torsional responses with placeholders for surrogate models is presented as an initiative. Finally, through a scholarly discussion, the study establishes the need for and feasibility of component-level DTs as an underexplored research direction for improving the reliability of marine propulsion systems through enhanced condition monitoring.