A Digital Twin for Production Control Based on Remaining Cycle Time Prediction

The recent industrial context pushed manufacturers to invest heavily in digitization for a more efficient use of their equipment and scarce resources. The digitization of industrial environments allows the establishment of digital decision-support tools such as digital twins, to exploit the shop-floor data for making more accurate decisions considering the real system state. Existing literature focuses on the development of specific digital twin components as well as methods that are typically developed and tested without an integration within a digital twin architecture. This paper proposes a complete digital twin framework with the purpose of aiding production planning and control operations. The focus is on the design of a production control service that manages the material flow in the real system using simulation-based predictions of the remaining cycle time. Preliminary experiments are done by applying the digital twin architecture on a lab-scale model, demonstrating the applicability of the proposed approach.