From purpose to technology: A requirements-driven approach to designing digital twin implementations

The digital twin (DT) paradigm promises to take permanent, bidirectional data exchange between the physical asset and its digital representation into account at its core. Nevertheless, as with any digital model, a DT is always an abstraction of reality. This chapter introduces a guide to creating purpose-driven, geometric-semantic DTs. It proposes a DT implementation framework with use-case-derived requirements and an integrity metric for evaluation. Potentially involved technologies for implementation are reviewed in three phases: raw data acquisition and processing, data enrichment, and geometry provision. The chapter elaborates requirement-dependent implementations for the use case of "Efficient HVAC maintenance." A compact variety of further DT use cases across several domains is provided, each along with specific requirements and implementation suggestions to fulfill them. The chapter presents a guide that allows a structured design of successful DT implementations with a sound selection of technologies and processing approaches.