A TOGAF-Driven Proactive Digital Twin for Production Planning and Resource Optimization

Abstract

Modern production systems operate under high variability and frequent disruptions, reducing the effectiveness of classical static planning approaches. Although Digital Twin technologies enable real-time monitoring and simulation, many implementations lack embedded optimization mechanisms for prescriptive decision support. This paper proposes a TOGAF (The Open Group Architecture Framework)-driven Digital Twin framework for adaptive production planning, in which a Mixed-Integer Linear Programming (MILP) model is continuously re-optimized using real-time shop-floor data. The Digital Twin provides bidirectional synchronization between the physical system and its virtual decision model, enabling closed-loop adaptive control, dynamic resource reallocation, and proactive disturbance handling. The framework is structured within a TOGAF-aligned enterprise architecture that integrates business objectives, information flows, and optimization logic. The approach is validated through computational experiments based on an industrial case study, including static–dynamic comparison, scalability assessment, and robustness analysis under capacity disruptions. Results demonstrate a reduction in total completion time, an increase in average resource utilization, and preserved feasibility under severe disturbances, confirming the effectiveness and practical applicability of Digital Twin–based dynamic decision support.