These slides were presented at the NAFEMS World Congress 2025, held in Salzburg, Austria from May 19–22, 2025.

Abstract

Virtualization of engineering has the potential to address strategic goals like reduction of time-to-market, managing an increasing product complexity and variety, or raising the flexibility to react on the market. Companies can realize it at various levels. The lowest level provides selective support for product development without substituting any physical test '“ hardware-based development is accompanied by a small number of simulation methods. This level primarily helps to raise the quality of a product but has minor influence on time to market. The highest level enables virtual release of a product including relevant certification without any physical prototype. This is not only the level that enables a significant shortening of development time. It is also the best base for any digital twin operation during product lifecycle. Any virtualization level between the described extrema must fit to a company individually. But there are factors that have an influence on the effectiveness regardless of which aimed level. Simulation methods enable the identification of the behavior or characteristic of products, materials, or process virtually '“ in the best case without physical prototypes or physical testing. Mostly, companies start with investigating and implementing simulation methods '“ either bought or licensed from expert tool providers or developed by themselves. These methods are crucial requisites - but they are just the visible peak of the "virtualization iceberg". It is evident that even powerful high-quality tools need validated input data and usage ability to provide high-end prognosis quality. To set up a simulation model with realistic behavior, simulation engineers need sufficient system understanding. Finally, the provided simulation results must be verified and confirmed by adequate testing, analysis, and comparison. To guarantee continuous quality, the whole simulation procedure must be confirmed and documented. This also must include the relevant usage limitations. The required level of detail is dependent on the level of experience of the users, who will apply the validated procedure within a product development. But there are several further factors that unleash the potential of virtualization and lead to an efficient virtualized product development: process design, dedicated organization, efficient data handling, numerical optimization and robustness analysis, procedure automation, tangibility, and trust, as well as data-driven design methods. There are still many companies who cannot imagine or do not dare to follow the path towards virtualization consequently. Once the virtualization level is rising and numerical support becomes significant, it is not sufficient any more to stick to the former hardware-related development process. Time periods, number of gateways and milestones as well as synchronization points for functions and data must be defined adequately. The digital prototype must become the leading element for development '“ together with the implementation of relevant responsibilities in the organization. Synchronization and integration of the high number of numerical disciplines and methods '“ e.g., by data provision or by system simulation - is becoming crucial for cross-disciplinary optimization. The real benefit from virtualization is appearing, when parameter variation helps to explore the solution space in an early design phase or the robustness of a solution, when MBSE models enable an highly automated validation of new functions or product variants, when tangibility of results allow direct exploration of data for decision makers, or when AI-based algorithms assist in the product design or smart product functions. The presentation will give a comprehensive experienced-based picture of potentials and dependencies as well as of requisites and recommendations for a successful way towards engineering virtualization.