Robust Digital Twin Compositions for Industry 4.0 Smart Manufacturing Systems

Abstract

Industry 4.0 is an emerging business paradigm that is reaping the benefits of enabling technologies driving intelligent systems and environments. By acquiring, processing and acting upon various kinds of relevant context information, smart automated manufacturing systems can make wellinformed decisions to adapt and optimize their production processes at runtime. To manage this complexity, the manufacturing world is proposing the 'Digital Twin' model to represent physical products in the real space and their virtual counterparts in the virtual space, with data connections to tie the virtual and real products together for an augmented view of the manufacturing workflow. The benefits of such representations are simplified process simulations and efficiency optimizations, predictions, early warnings, etc. However, the robustness and fidelity of digital twins are a critical concern, especially when independently developed production systems and corresponding digital twins interfere with one another in a manufacturing workflow and jeopardize the proper behavior of production systems. We therefore evaluate the addition of safeguards to digital twins for smart cyber-physical production systems (CPPS) in an Industry 4.0 manufacturing workflow in the form of feature toggles that are managed at runtime by software circuit breakers. Our evaluation shows how these improvements can increase the robustness of interacting digital twins by avoiding local errors from cascading through the distributed production or manufacturing workflow.

Data Link for the Creation of Digital Twins

“…Employing microservices in the building of a digital twin has been proposed by several papers [6], [8], [9], [18]. Alasaam et al [8] used Apache Kafka to enable stream data processing with microservices.…”

“…However, the communication model is not described in detail, and sufficient information for the implementation of the concept is not provided. Preuveneers et al [9] introduced a digital twin architecture based on microservices, which allows easy addition of new features to the digital twin. To prevent local failures from propagating, the Digital Twins can be toggled by other DTs via RESTful interface.…”

“…For example, at simplest, a digital twin may only contain the metadata description, DT document, Link connects these blocks using an ordinary API gateway and enables the communication between the blocks by forwarding requests from one service to another. It lowers the number of connections compared to architectures in which systems communicate directly with each other presented by, for example, Preuveneers et al [9]. Lowering the number of connections simplifies the linking of services because each service does not need to contain information on how to connect to all other services.…”

“…Building a digital twin using microservices architecture allows better scalability and maintainability, and, robustness is improved since the failure of a single service does not likely cause the failure of the whole system [7] -in this case, the digital twin. Implementation of digital twin based on microservices is demonstrated, for example, in [8] and [9]. The major benefit of FDTF and building DT using (micro-)services is that existing systems can be used to implement features of a digital twin [5].…”

Digital Twins as Software and Service Development Ecosystems in Industry 4.0: Towards a Research Agenda

“…The traditional top-down enterprise management systems operate in a closed and hierarchical manner with higher-level controllers to plan, coordinate, and optimize production processes start to shift to changes that enable greater flexibility and adaptability to internal and external changes. In the world of networked production with multiple actors in complex manufacturing and production processes and ecosystems, the control process will inherently be a hybrid combination of hierarchical and heterarchical collaboration processes [15]. The equipment and technology vendors and software and system developers are all needed to become part of the ever-increasing complexity of larger and larger automated production networks [37], such as a digital twin network or ecosystem to support the production systems in industry 4.0 [15].…”