Open Access

Mario Luber, Sarah Wagner, Marc Wegmann, Johannes Schilp

• Due to the progress of Digital Production and the Industrial Internet of Things, continuous shop floor data is available with high coverage, accuracy, and in high detail for Production Planning and Control software. Detailed Capacity Planning and Scheduling (DCPS) can benefit by applying the Digital Twin concept and Machine Learning for an accurate and automated virtual representation of the production resource. However, the effort and difficulty required for data connection, data preparation, and modelling are high. Connection standards enable interoperability and plug-and-play software, and constitute an opportunity to reduce the effort and difficulty. This article compiles requirements regarding the virtual representation of the production resource for DCPS. It then proposes the concept of a plug-and-play, Machine Learning Digital Twin to meet these requirements. The elements of the according Digital Twin software are described, and the need for future research is identified.