Large Language Models (LLMs) have shown promising capabilities in supporting and automating a variety of clearly defined, structured tasks. The engineering of modular plants also follows standardized workflow structures and uses standardized artifacts such as P&IDs, flow diagrams or HAZOPs. Therefore, the engineering of modular plants has the potential to be significantly supported or automated by AI tools. The articles objective is to determine which aspects of the modular plant engineering workflow could be supported or automated by LLMs. This is done by reviewing the current state of the art and combining the modular plant engineering workflow which LLM-based capabilities for engineering support. This leads to two results. First, a mapping of current LLM capabilities to the steps within the modular plant engineering workflow. Second, a theoretical assessment of the general LLM capabilities in terms of their usefulness for the tasks within the modular plant engineering workflow. All in all, the proposed mapping of LLMs to the modular plant engineering workflow provides an overview of the ways in which LLMs can support engineers. In addition, the LLM capability assessment provides insights into the current state of the LLMs abilities and highlights where further research and development is needed.

In the era of Industry 4.0 (I4.0), Cyber Physical Production Systems (CPPS) and upcoming industrial transformations, there's a great impulse for smarter, more connected, and adaptable industries. To support this shift, our industrial devices need to be upgraded. They should not only do their usual tasks reliably but also support new technologies like Artificial Intelligence (AI), Machine Learning (ML), Digital Twins (DT), etc. seamlessly. This is where cognition in devices becomes significant. This paper showcases an innovative cognitive system design for edge devices for control, drawing inspiration from the well-established concept of cognitive control. This approach highlights the symphonious existence of conscious (controlled) and unconscious (automatic) processing. The system architecture demonstrates the concurrent processing of real-time tasks, like control loops, field devices, etc., and non-real-time tasks such as AI, ML, Neural Network (NN) models, DT models, etc. within the edge device. This corresponds to the unconscious and conscious cognitive functions in human beings. This paper outlines the characteristics of such cognitive devices. The potential technologies that can help in achieving these characteristics like virtualization, multi-core edge devices, concurrency, etc. have also been explored. Additionally, a proof-of-concept demonstration use case has been presented that exhibits the implementation of the Open Cognitive Control Systems (OCCS) architecture in edge devices. This work sets a stepping stone towards making industrial devices smarter.

The technology readiness levels of cloud infrastructure and edge devices have increased significantly in recent years. This means that companies now have a growing number of computing environments at their disposal that could be suitable for the computing and control tasks involved in their production processes. Consequently, the need arises to select the best computing environment based on objective criteria and metrics. In order to create the basis for such a decision-making process, first an overview of the definitions and a comparison between edge and cloud environments is provided. Then a decision-making framework is derived that aims to facilitate this selection process. This is accomplished by providing a set of criteria that can be detailed to analyze a given use case. This analysis is then represented in a custom radar chart, which offers decision-makers a compact but meaningful representation of the decision space. Finally, the limits and future potentials of such a decision-making framework are discussed.

BACKGROUND: Computational and data-intensive technologies such as model predictive control and artificial intelligence have the potential to increase process yields in the process industry. In modular plant designs, the flexible and difficult-to-predict use of a particular module presents the challenge of providing the right amount of data, computing power, and storage capacity to use these technologies in each module.OBJECTIVE: Simplifying the deployment and reconfiguration of compute-intensive applications for modular plants.METHODS: Reviewing the current state of the art and combining the modular plant concept with a cloud-based Control-as-a-Service (CaaS) approach.RESULTS: A cloud-based, containerized CaaS concept that supports the deployment compute- and data-intensive methods for modular plants. The complementing communication stack consisting of standards such as APL, TSN and OPC UA FX.CONCLUSION: The proposed architecture simplifies IT/OT integration, eases the deployment of compute-intensive technologies, and allows for GMP compliant pre-qualification of software modules. However, the limiting factors like plant safety, conformity to regulations and widespread architecture adoption are not covered in detail and will be subject of future research.