Efficiency and flexibility are continual demands on future production systems, and modularization has emerged as a strategy to address these challenges. In the process industry, the Module Type Package (MTP) standard was introduced to simplify the integration and configuration of modular production systems. Transferring the MTP technology to other indus-trial domains is an active research field. This paper describes how the MTP can be used to integrate robotic systems into mod-ular manufacturing plants without compromising the MTP standard. This fills a gap in existing literature and provides promising scientific and commercial opportunities. The study presents a layered concept encompassing all necessary compo-nents for the integration of robotic systems, from the robot's control to the orchestration of the modular plant. A prototypical implementation demonstrates the feasibility of the proposed ap-proach, successfully performing a typical robotic task, a Pick-and-Place operation, using MTP-based services. The imple-mented services, though representing basic robot functionali-ties, lay the groundwork for more complex services or nested service structures. In conclusion, this research contributes to the understanding of the potential integration of robotic systems in MTP-based modular production environments, paving the way for more sophisticated applications, and addressing open questions for future exploration. The prototypical implementation is publicly available on GitHub.

In the era of Industry 4.0, data is becoming increasingly vital for future manufacturing and digital twin applications. Currently, primarily asset-created data is utilized, while the potential benefits of incorporating external data from connected production systems remain untapped. This underutilization is largely due to the complex and time-consuming nature of integrating external data. Automating this process could significantly improve efficiency and unlock valuable insights for more effective decision-making. In this paper, we present the requirements for a system aimed at automating the integration of external data, considering factors such as data heterogeneity and the need to support diverse communication protocols. We further introduce an assistant system that facilitates this automated integration, enhancing manufacturing processes and digital twin applications. Furthermore, we provide a prototypical implementation of the assistant system in the context of a collaborative robot utilizing information from its surroundings obtained through external camera and LiDAR imaging. By addressing the challenges of external data integration, this research contributes to the advancement of data-driven manufacturing and the optimization of digital twin technologies.

Ideally, modern modular production plants are controlled via services with uniform interfaces and state machines. But in detail, services are of diverse nature. Due to this diversity, orchestration systems face a challenge of correctly "understanding" the available services and recognizing how and when these services can be used. This paper identifies fundamental service properties, enabling the categorization of services. Furthermore, the authors propose to assign an information model containing the service properties to each service. Consequently, an orchestration system can independently recognize, interpret, and accordingly coordinate the information about service properties and variants; they become explorable. Orchestration systems are thus put in a position to deal explicitly with different services of diverse control methods on their own.

The Digital Twin is one of the future key technologies of digitization and Industry 4.0. Through coupling the vast amount of available static and dynamic data about a physical asset with intelligent software functions, it aims for simplifying the increasingly complex functions and interconnections of automation systems. This makes it necessary to provide the Digital Twin with updated and high-quality data about the physical asset. However, little attention is paid to data not generated by the asset itself but from external data sources. This is partly due to a lack of methodologies for identifying and evaluating external data relevant for a specific Digital Twin. The contribution of this work is a categorization of data for the Digital Twin and a methodology to support the identification of data from outside of the physical asset. The work-in-progress refers to the context of industrial automation and is to be validated in the future in laboratories containing discrete manufacturing facilities at Pforzheim University and the University of Stuttgart.

This contribution pursues the approach to extend real and operative plant components by virtual elements to improve efficiency, safety, and reliability in the automation engineering process. Although some research has been done in the field of mixing real and virtual representations of different plant components, the presented Virtual-In-The-Loop approach promises a wide range of new applications. To enable a systematic investigation of its possibilities and limitations, we present a first-time categorisation of the opportunities for mixing real and virtual plant components and interfaces. Furthermore, we propose a terminology that enables the unambiguous naming of the different forms of mixtures. The results are intended to serve as an orientation and classification guide for future work in the field of Virtual-In-The-Loop Engineering.