As supply chains grow increasingly complex and demand for autonomous delivery solutions rises, traditional logistics systems struggle with real-time coordination requirements across multiple transportation modes. The TRACE Reference Architecture (RA) addresses these challenges through a telecommunications-aware design that integrates multimodal transportation systems with advanced network capabilities. Following a layered structure, the proposed RA incorporates user interfaces, application services, data management modules, and physical infra-structure components, fostering real-time decision-making and coordination across logistics stakeholders. The architecture also leverages edge computing, distributed intelligence, and blockchain technologies to ensure secure data exchange while supporting autonomous vehicles (e.g., ground vehicles, drones) as part of multi-modal logistics operations. We demonstrate the architecture's versatility through three distinct implementation scenarios in Italy, Slovenia, and Greece, showcasing different network configurations including 5G Stand-Alone, Non-Stand-Alone, and Vehicle-to-Vehicle communications that sup-port autonomous delivery, platooning, and multimodal transport coordination. This paper illustrates how the TRACE RA provides a scalable foundation for next-generation logistics that balances operational efficiency with security and privacy requirements in the emerging B5G ecosystem.