Virtual Commissioning to Digital Twins
Digital twins are computational representations of systems, processes, or living entities, utilized across domains such as manufacturing, automotive, aerospace, and robotics. They enable efficient process planning, improved reliability in cyber-physical production systems (CPPS), and cost reductions. Interpretations of digital twins vary, ranging from precise simulation models of CPPS under development—used for behavioral prediction without direct influence—to complex software solutions that dynamically represent and manipulate existing CPPS during operation. This conceptual spectrum reflects divergent applications in fields like mechanical engineering, automation, medicine, and computer science, highlighting a lack of standardized definition.
A significant conceptual gap exists between model-centric and system-centric interpretations of digital twins. The former involves abstract simulations of machine classes (“as-designed”), while the latter focuses on real-time representations of operational CPPS (“as-operated”). Bridging this gap is critical for seamless digitalization, as discrepancies arise from tolerances, environmental factors, and wear, causing “as-operated” systems to deviate from idealized models. Without continuous adaptation, these models lose predictive accuracy, leading to flawed analyses. To address this challenge, VC2DT aims to develop a methodology for the (semi-)automatic transformation of “as-designed” CPPS models into precise “as-operated” models during system commissioning, ensuring continuous alignment between simulations and real-world operational conditions.
Funding
Funded by the Deutsche Forschungsgemeinschaft (DFG).