Refine
Document Type
- Journal article (14)
- Conference proceeding (4)
Is part of the Bibliography
- yes (18)
Institute
- ESB Business School (17)
- Technik (1)
Publisher
- Elsevier (11)
- De Gruyter (1)
- Gesellschaft für Informatik e.V (1)
- IEEE (1)
- LIT Verlag (1)
- Lange (1)
- MDPI (1)
- Springer (1)
Gesellschaftliche und industrielle Trends im Zuge der Digitaliserung induzieren Veränderungsprozesse in der Industrie. Eine hohe Flexibilität und schnelle Entscheidungsfindungsprozesse stellen entscheidende Wettbewerbsvorteile für Unternehmen dar, um zukünftig erfolgreich am Markt agieren zu können. Um dies zu ermöglichen, müssen aggregierte Echtzeitdaten und Prognosen unmittelbar sowohl am Ort der Wertschöpfung als auch dezentral zur Verfügung stehen. Die Entscheidungsunterstützung mit Hilfe geeigneter Visualisierungen ist ein maßgeblicher Bestandteil von Shopfloor Management Systemen. Aufgrund der steigenden Anforderungen wurde das konventionelle und analoge Shopfloor Management in den letzten Jahren verstärkt durch digitale Lösungen ersetzt. Ein ganzheitlicher Shopfloor Management Ansatz, der die Trends und die daraus resultierenden Herausforderungen für die Industrie abdeckt, ist aktuell nicht vorhanden. Zukünftige Shopfloor Management Lösungen sollen diese Lücke schließen. Hierfür wurde ein ganzheitliches System entwickelt, welches Produktionsinformationen in Echtzeit unmittelbar am Shopfloor visualisiert, eine integrierte flexible Planung und Steuerung der Produktion beinhaltet sowie die Mitarbeiterbedürfnisse berücksichtigt. Eine flexible und individuelle Schichtplanung durch die Mitarbeiter und eine umfassende automatische Beanspruchungsbeurteilung sind dazu integriert worden. Zudem ermöglicht das System die Prognose und Visualisierung von Produktionsinformationen und unterstützt die Anwender bei der Durchführung strukturierter Shopfloor-Meetings. Dadurch werden Entscheidungen direkt auf den Ort der Wertschöpfung verlagert.
Learning factories on demand
(2021)
Learning Factories are research and learning environments that demonstrate new concepts and technologies for the industry in a practical environment. The interaction between physical and virtual components is a central aspect. The mediation and presentation usually occur directly in the learning factory and are thus limited in time and concerning the user group. A learning factory- on-demand- can be provided by dividing and virtualizing the individual components via containers and microservices. This enables both local operation and operation hybrid cloud or cloud systems. Physical components can be mapped either through standardized interfaces or suitable emulators. Using the example of the Learning Factory at Reutlingen University (Werk150), it will be shown how different use cases can be made available utilizing software-based orchestration, thus promoting broader and more independent teaching.
Conventional production systems are evolving through cyber-physical systems and application-oriented approaches of AI, more and more into "smart" production systems, which are characterized among other things by a high level of communication and integration of the individual components. The exchange of information between the systems is usually only oriented towards the data content, where semantics is usually only implicitly considered. The adaptability required by external and internal influences requires the integration of new or the redesign of existing components. Through an open application-oriented ontology the information and communication exchange are extended by explicit semantic information. This enables a better integration of new and an easier reconfiguration of existing components. The developed ontology, the derived application and use of the semantic information will be evaluated by means of a practical use case.
Artificial intelligence is considered to be a significant technology for driving the future evolution of smart manufacturing environments. At the same time, automated guided vehicles (AGVs) play an essential role in manufacturing systems due to their potential to improve internal logistics by increasing production flexibility. Thereby, the productivity of the entire system relies on the quality of the schedule, which can achieve production cost savings by minimizing delays and the total makespan. However, traditional scheduling algorithms often have difficulties in adapting to changing environment conditions, and the performance of a selected algorithm depends on the individual scheduling problem. Therefore, this paper aimed to analyze the scheduling problem classes of AGVs by applying design science research to develop an algorithm selection approach. The designed artifact addressed a catalogue of characteristics that used several machine learning algorithms to find the optimal solution strategy for the intended scheduling problem. The contribution of this paper is the creation of an algorithm selection method that automatically selects a scheduling algorithm, depending on the problem class and the algorithm space. In this way, production efficiency can be increased by dynamically adapting the AGV schedules. A computational study with benchmark literature instances unveiled the successful implementation of constraint programming solvers for solving JSSP and FJSSP scheduling problems and machine learning algorithms for predicting the most promising solver. The performance of the solvers strongly depended on the given problem class and the problem instance. Consequently, the overall production performance increased by selecting the algorithms per instance. A field experiment in the learning factory at Reutlingen University enabled the validation of the approach within a running production scenario.
Parallel grippers offer multiple applications thanks to their flexibility. Their application field ranges from aerospace and automotive to medicine and communication technologies. However, the application of grippers has the problem of exhibition wear and errors during the execution of their operation. This affects the performance of the gripper. In this context, the remaining useful life (RUL) defines the remaining lifespan until failure for an asset at a particular time of operation occurs. The exact lifespan of an asset is uncertain, thus the RUL model and estimation must be derived from available sources of information. This paper presents a method for the estimation of the RUL for a two-jaw parallel gripper. After the introduction to the topic, an overview of existing literature and RUL methods are presented. Subsequently, the method for estimating the RUL of grippers is explained. Finally, the results are summarized and discussed before the outlook and further challenges are presented.
Modern production systems are characterized by the increasingly use of CPS and IoT networks. However, processing the available information for adaptation and reconfiguration often occurs in relatively large time cycles. It thus does not take advantage of the optimization potential available in the short term. In this paper, a concept is presented that, considering the process information of the individual heterogeneous system elements, detects optimization potentials and performs or proposes adaptation or reconfiguration. The concept is evaluated utilizing a case study in a learning factory. The resulting system thus enables better exploitation of the potentials of the CPPS.
The functionality of existing cyber-physical production systems generally focuses on mapping technologic specifications derived from production requirements. Consequently, such systems base their conception on a structurally mechanistic paradigm. Insofar as these approaches have considered humans, their conception likewise is based on the structurally identical paradigm. Due to the fundamental reorientation towards explicitly human-centered approaches, the fact that essential aspects of the dimension "human" remain unconsidered by the previous paradigm becomes more and more apparent. To overcome such limitations, mapping the "social" dimension requires a structurally different approach. In this paper, an anthropocentric approach is developed based on possible conceptions of the human being, enabling a structural integration of the human being in an extended dimension. Through the model, extending concepts for better integration of the human being in the sense of human-centered approaches, as envisioned in the Industrie 5.0 conception, is possible.
Artificial intelligence is a field of research that is seen as a means of realization regarding digitalization and industry 4.0. It is considered as the critical technology needed to drive the future evolution of manufacturing systems. At the same time, autonomous guided vehicles (AGV) developed as an essential part due to the flexibility they contribute to the whole manufacturing process within manufacturing systems. However, there are still open challenges in the intelligent control of these vehicles on the factory floor. Especially when considering dynamic environments where resources should be controlled in such a way, that they can be adjusted to turbulences efficiently. Therefore, this paper aimed to develop a conceptual framework for addressing a catalog of criteria that considers several machine learning algorithms to find the optimal algorithm for the intelligent control of AGVs. By applying the developed framework, an algorithm is automatically selected that is most suitable for the current operation of the AGV in order to enable efficient control within the factory environment. In future work, this decision-making framework can be transferred to even more scenarios with multiple AGV systems, including internal communication along with AGV fleets. With this study, the automatic selection of the optimal machine learning algorithm for the AGV improves the performance in such a way, that computational power is distributed within a hybrid system linking the AGV and cloud storage in an efficient manner.
The paradigmatic shift of production systems towards Cyber-Physical Production Systems (CPPSs) requires the development of flexible and decentralized approaches. In this way, such systems enable manufacturers to respond quickly and accurately to changing requirements. However, domain-specific applications require the use of suitable conceptualizations. The issue at hand, when using various conceptualizations is the interoperability of different ontologies. To achieve flexibility and adaptability in CPPSs though requires overcoming interoperability issues within CPPSs. This paper presents an approach to increase flexibility and adaptability in CPPSs while addressing the interoperability issue. In this work, OWL ontologies conceptualize domain knowledge. The Intelligent Manufacturing Knowledge Ontology Repository (IMKOR) connects the domain knowledge in different ontologies. Testing if adaptions in one ontology within the IMKOR provide knowledge to the whole IMKOR. The tests showed, positive results and the repository makes the knowledge available to the whole CPPS. Furthermore, an increase in flexibility and adaptability was noticed.
Cyber-Physical Production Systems increasingly use semantic information to meet the grown flexibility requirements. Ontologies are often used to represent and use this semantic information. Existing systems focus on mapping knowledge and less on the exchange with other relevant IT systems (e.g., ERP systems) in which crucial semantic information, often implicit, is contained. This article presents an approach that enables the exchange of semantic information via adapters. The approach is demonstrated by a use case utilizing an MES system and an ERP system.
