Towards a scalable implementation of digital twins - A generic method to acquire shopfloor data

Abstract New strategies in factory management focus on shopfloor data as a mean to achieve more flexible mass productions. In reality, the value adding usage of smart manufacturing approaches depends on the effort it takes to provide sensor and actuator data for further analysis. The concept of digital twins, that are based on a physical, a virtual and a communication component, promises a scalable solution to read and standardize shopfloor data. So far, there exists no generic method describing the implementation of the data acquisition as part of the communication component of digital twins. This contribution therefore focusses on necessary steps to consider when realizing a scalable data acquisition and examines how existing standards and open source solutions can support the implementation. The first part of the paper derives the necessary steps for acquiring shopfloor data. It starts by defining the architecture and aim of digital twins. Afterwards, a method to implement the generic data acquisition of digital twins is introduced. The approach is inspired by the Plug-and-Produce paradigm of the control engineering field and is adapted to the concepts of data acquisition and management. The second part examines the technical implementation of the proposed method. Identified approaches from a literature review are structured within the generic method. This describes the realization of the individual steps but also systematically differentiates existing approaches to build Digital Twins. With the aim of creating scalable architectures, a special focus is set on available open source solutions and standards when presenting the implementation part of the generic method.