Häufig gestellte Fragen
What distinguishes a digital twin from a 3D model or a simulation?
A 3D model or a classic simulation is static and represents a fixed design or calculation state, whereas a digital twin is continuously synchronised with real-time and historical data from the real object. This ongoing feedback loop to reality is the defining characteristic: if the physical asset changes, for instance through wear or altered operating parameters, this is reflected in the twin. A 3D model or simulation model can be a building block of the twin, but without the continuous data connection it does not replace a digital twin. A data warehouse is not a twin either, since it historicises data but does not represent the current state and behaviour of a specific object.
What role does the ERP play in a digital twin?
The digital twin is not a standalone ERP module but a cross-system concept in which the ERP supplies the commercial and master-data facts. The ERP typically contributes material numbers, bills of materials, order, serial number, maintenance and cost information that anchor the technical twin in business terms. The real-time and simulation logic, by contrast, usually resides on an IoT or engineering platform, while PLM systems contribute design and lifecycle states and MES systems the manufacturing execution. Clean master data in the ERP is the prerequisite, since without reliable material numbers and bills of materials no consistent representation can be built.
What prerequisites and interfaces does a digital twin require?
A digital twin requires IoT sensors connected to a data or asset platform, integrated CAD and engineering data, and sufficient computing capacity for simulations, complemented by subject-matter maintenance of the model. Data exchange frequently takes place via industrial protocols such as OPC UA or via REST APIs, and large volumes of time-series data should be stored in a way that does not overload the ERP. In practice, it is advisable to start with a narrowly defined use case rather than a complete plant model. A clean data and interface architecture matters more than the choice of any single product, and the specific implementation should be agreed with the respective vendor.
Are there norms and standards for digital twins?
For manufacturing digital twins, the ISO 23247 series of standards provides a conceptual framework describing so-called Observable Manufacturing Elements such as personnel, equipment, material, processes and products, and their synchronisation with the digital representation. In addition, the Asset Administration Shell (Verwaltungsschale) is regarded as the technical foundation for the machine-readable description of industrial assets in the context of Industrie 4.0; it is standardised in IEC 63278-1:2023. These standards reduce integration risks and lessen dependence on individual vendors, since data is described interoperably. Adoption in practice is uneven, however, so actual standards compliance needs to be checked on a project-by-project basis.
Is a digital twin worthwhile for mid-sized companies, and what does it cost?
For capital-intensive machinery or complex production, a digital twin can pay off for mid-sized companies too, with the first quick wins often arising in predictive maintenance for individual larger machines. According to industry figures, asset-level twins for a few critical machines start at roughly several tens of thousands of euros per machine, while twins of entire factories (system twins) can reach six- to seven-figure budgets. Licence fees typically account for only a smaller share of total costs; the majority goes to implementation, customising, training and data migration. Industry studies report double-digit percentage reductions in unplanned downtime and maintenance costs, but the actual economics depend heavily on the use case and the existing system landscape.
What are typical use cases for a digital twin in production?
A frequent use case is predictive maintenance, where sensor data on vibration, temperature, running time or energy consumption indicates wear early and the system generates a maintenance proposal before a failure occurs. Beyond that, the twin serves process optimisation, the simulation of what-if scenarios and production planning in combination with ERP and MES systems. In practice, three maturity levels are distinguished: the Digital Twin Prototype as a product model before manufacturing, the Digital Twin Instance as the representation of a specific unit with its history, and the Digital Twin Aggregate, which combines many instances for fleet or plant analyses. The real benefit only emerges from linking technical early warning with commercial planning in the ERP, for example checking spare-parts availability and cost-effective technician deployment.