In automotive enterprises, digital twins (DTs) are often developed as target-specific systems for specific phases of the vehicle lifecycle, leading to fragmentation. This fragmentation hinders cross-phase analysis and decision-making. This paper presents a novel federated digital twin (FDT) system architecture that enables the integration of information from fragmented DT systems, with minimal custom development requirements. The FDT system accesses information from heterogeneous member DT platforms (where each platform hosts DTs of the same phase of the lifecycle of multiple vehicles) while preserving the platforms’ operational independence. An illustrative example, focused on vehicle component defect traceability, demonstrates the architecture’s efficacy. A preliminary evaluation indicates that the architecture reduces integration complexity while enabling responsive decision-making in critical scenarios. Although further evaluation at industry scale is necessary, the architecture is expected to enhance operational efficiency by supporting decisions that require information from across the product lifecycle.