Manufacturing Systems

Engineers are often uncertain about the application of machine learning (ML) due to the amount of different machine learning methods and the complexity of modeling. Thus, the use of ML applications in manufacturing companies remains behind the technical possibilities. This paper presents an intuitive ML guideline for engineers to reduce this uncertainty. The guideline comprises a process model with AI-based solutions to common problems of product development and production. An industrial example is used to demonstrate the functionality and the possibilities of the guide.

A fast flow of information throughout the entire supply chain is unavoidable for risk minimization and is not subject of a discussion in volatile times or crisis situations. The flow of information within the supply chain is characterized by various forms of transmission: EDI, cloud applications or other system interfaces are manifold in the areas of value-added networks for digital risk monitoring and process efficiency increase. If corporate processes are examined more closely, one area remains digitally underrepresented at the moment: The digital twin of a production tool. The handling of these production tools must now be taken to a new level.

The level of networking and complexity in intelligent manufacturing is constantly increasing. As a result, the demands on employees are growing, especially due to changing work tasks. In addition, the existing lack of skilled workers leads to bottlenecks. The use of remote technologies opens up new opportunities for collaboration, especially in maintenance. Considering the challenges of industrial maintenance, to what extent can remote technologies be used to efficiently meet these demands? This article provides an overview of the remote technologies currently used and discussed in practice and research. In addition, it shows which prerequisites must be created for an effective application of the technologies.

Volatility in market demand leads to temporary over- and under-utilization of production assets and stocks. Levelling (heijunka) as a lean method aims at de-coupling production from market volatility. The production program is spread as even as possible over time. This achieves high asset utilization, short lead times, and low inventories. There are validated heijunka methods for the manufacturing industry, but for the process industry this remains a research gap. This article describes the Product Wheel and its validation at a building material manufacturer in order to close that gap.

Due to the possibility of operator intervention, semi-autonomous systems allow for a better handling of complexity than fully autonomous systems. The use of a digital twin provides a novel interface for interaction with such systems. This paper describes the implementation of the control and user interface in a system with a digital twin. It is shown how the developed control architecture can be combined with different methods of human-machine interaction and virtual training. With this extended use of the control system by a digital twin the concept can be extended beyond the operation phase and can be used in other phases of the product life cycle.