Digital twins are a concept often reduced to simulation. While simulation has been used in logistics for a long time, digital twins extend the common notion of simulation through its capacity to support near real-time planning and decision-making. Digital twins enable a dynamic, detailed, and functional real-time representation of physical assets to monitor their performance, anticipate their future state, or control how their resources are used. Its approaches and methods to designing, developing, implementing, and updating digital models that allow modularity, reuse and evolution of their components have not yet been exhaustively developed. It is not the first time the modeling & simulation community has attempted to create real-time simulation models and tools. However, it is the first time the technologies required to develop such advanced performance assessment and planning and control systems are available to enable the functions needed to support real-time adaptive planning and decision-making. Several conceptual frameworks have been proposed to build a coherent and common understanding of planning and control systems, but its interpretation and adaptation to specific systems still require the development of dedicated methodological and technological architectures. By exploring the body of knowledge before the popularization of the term digital twins, this paper proposes a critical analysis of the technological requirements and methodological steps to implement digital twins in logistics. Ultimately, the goal of this project is to propose a methodological framework dedicated to the design, development, and operation of digital supply chain twins.
Domingos et al. (Tue,) studied this question.