Zusammenfassung: Vermehrt werden Daten in der Produktion erfasst und mit Hilfe maschineller Lernverfahren ausgewertet. Ziel dieser Datenanalysen ist es, Informationen über die Produktion zu erhalten. Typische Anwendungsbeispiele sind die Vorhersage der Produktqualität sowie die vorausschauende Wartung und Instandhaltung. Allerdings führen Datenanalysen aufgrund der domänenspezifischen Herausforderungen häufig zu keinen ausreichenden Analyseergebnissen. Die Kombination von Datenanalysen mit Expertenwissen stellt hierbei einen vielversprechenden Ansatz zur Lösung der domänenspezifischen Problemstellungen dar. Dieser Beitrag stellt hierzu eine neue Vorgehensweise für die Durchführung von Projekten des maschinellen Lernens mit wissensbasierten Ansätzen in der Produktion vor. In einem Ablaufplan werden notwendige Schritte und die Beteiligung der entsprechenden Experten vorgestellt. Für die Nutzung von vorhandenem Domänenwissen in der Produktion werden in diesem Artikel verschiedene Methoden zur Wissensmodellierung aufgezeigt und diskutiert. Der Ablaufplan und die Methoden zur Wissensmodellierung werden anschließend am Beispiel einer servopneumatischen Schweißzange validiert.

Sensitive robot systems are used in various assembly and manufacturing technologies. Assembly is a vital activity that requires high-precision robotic manipulation. One of the challenges faced in high precision assembly tasks is when the task precision exceeds the robot’s precision. In this research, Deep Q-Learning (DQN) is used to perform a very tight clearance Peg-in-Hole assembly task. Moreover, recurrence is introduced into the system via a Long-Short Term Memory (LSTM) layer to tackle DQN drawbacks. The LSTM layer has the ability to encode prior decisions, allowing the agent to make more informed decisions. The robot’s sensors are used to represent the state. Despite the tight hole clearance, this method was able to successfully achieve the task at hand, which has been validated by a 7-DOF Kuka LBR iiwa sensitive robot. This paper will focus on the search phase. Furthermore, our approach has the advantage of working in environments that vary from the learned environment. Keywords Assembly Peg-in-Hole Deep Learning Sensitive robotics

In the context of tolerance management, the consideration of manufacturing and assembly processes is becoming increasingly important. The main drivers for this are, above all, short development times and high-quality requirements, leading to tight tolerances. To overcome the resulting challenges, many publications address the process-oriented tolerance management. However, since multiple terms and definitions for describing activities that link tolerance management with the production process exist, it is hard to obtain a comprehensive overview on the topic. Therefore, this paper presents a review of existing approaches. The aim is to identify similarities and differences of existing approaches and present them with the help of a classification. For this purpose, among others, work from the areas of process-oriented tolerance management, stream of variation, state space modeling, and variation propagation in multistation manufacturing and assembly systems is considered. Based on the definition of the summarizing term “process-oriented tolerance and variation management”, a classification of this thematic area will be introduced.

Fuel cell systems will play an important role in the achievement of global climate goals, particularly in the building and transportation sectors. However, manufacturers face several challenges in the production of these systems, arising primarily from the complexity of products, processes, and equipment as well as the associated data processing requirements. This paper briefly outlines the challenges in the assembly of fuel cell stacks and introduces a specific design of a Digital Twin that addresses them with the help of tailored data management, data evaluation and information utilization. This is followed by the presentation of two potential application scenarios: the use of the Digital Twin as the basis for an expert system that accelerates scaling from prototype to series assembly and as a tool for end-to-end quality monitoring to continuously optimize the products and processes.

The special machinery sector is distinguished by a high level of complexity, driven by the consistently growing demand for customer-specific special solutions. In today's volatile market and highly competitive landscape, suppliers must be able to respond swiftly with tailored offers for specialized machinery. It is therefore essential to adopt a success-oriented approach to offer management, ensuring the detailed preparation of offers. One potential strategy for addressing these challenges in the design process is to prioritize the reuse of individual units (artifacts) from previous projects in new customer projects. Information to support the design process and the calculative preparation of an offer is relevant here. This article provides an overview of reuse strategies for the design of production systems. The approaches and mechanisms of reuse are described in more detail. We provide an overview of current research gaps in this field and a derivation of possible further developments for the development of a supplementary methodology for use in offer management in mechanical engineering.