Carmen Krahe, M.Sc.
Forschungs- und Aufgabengebiete:
- Maschinelles Lernen auf 3D-Modellen
- Industrie 4.0
- Produktionsplanung und -steuerung
- Steuerung flexibler Produktionssysteme
Vorlesungsbetreuer der Vorlesung Integrierte Produktionsplanung im Zeitalter von I4.0 (IPP)
Detailplanung: Produktionsplanung und -steuerun
- AIAx – Artificial Intelligence goes CAx
- MoSyS - Menschorientierte Gestaltung komplexer System of Systems
|seit 11/2018||Wissenschaftliche Mitarbeiterin am Institut für Produktionstechnik (wbk) des Karlsruher Instituts für Technologie (KIT)|
|10/2011 - 10/2018||Studium des Wirtschaftsingenieurwesens am Karlsruher Institut für Technologie (KIT)|
|05/04/1993||Geboren in Karlsruhe|
|[ 1 ]|| Hofmann, C.; Brakemeier, N.; Krahe, C.; Stricker, N. & Lanza, G. (2018), The Impact of Routing and Operation Flexibility on the Performance of Matrix Production Compared to a Production line. Advances in Production Research, Hrsg. Schmitt, R. & Schuh, G., S. 155-165.
An increasing number of product variants and a decrease in demand certainty challenge manufacturing companies. Lean, flow-oriented production lines are best-practice to assure efficient production in a predictable environment. However, with the increase in complexity and uncertainty, more flexible production systems such as matrix production currently receive much attention. Having neither a common takt time nor a rigid linkage, they offer new degrees of freedom regarding process order and machine choice. This paper contributes to answering the question under which conditions a matrix production is favourable compared to a production line. To answer this question, the effects of MTTF and MTTR as driving factor to choose a matrix production over a production line are analysed. Regarding the material flow in the matrix, the benefits of routing flexibility and operation flexibility concerning throughput time, tardiness and output of the matrix production are evaluated. The results show that a rule based approach has its limits especially regarding the exploitation of operation flexibility. For low levels of routing flexibility, the rule based approach tends to generate sup-optimal solutions due to a lack of coordination between the agents.
|[ 2 ]|| Lanza, G.; Klenk, F. & Krahe, C. (2019), Track & Trace als Basis für die Kreislaufwirtschaft in der Automobilindustrie, Werkstoffe in der Fertigung, S. 22-23.
Die Kreislaufwirtschaft zielt darauf ab, den aktuellen „take, make & dispose“-Ansatz der linearen Wirtschaft, der Auslöser massiver Ressourcenverschwendung ist, durch die mehrfache, wertschöpfende Zufuhr von Produkten in den Produktionsprozess zu verändern. Hierdurch kann der Ressourcen- und Energieverbrauch reduziert sowie die Wirtschaftlichkeit der Unternehmen erhöht werden. Eine Voraussetzung für die Implementierung der Kreislaufwirtschaft ist die lückenlose Nachverfolgbarkeit von Produkten, Komponenten und Materialien. So wird ermöglicht, dass Produkte, die am Ende ihres Lebenszyklus wieder in den Besitz eines Unternehmens gelangen, leichter identifiziert, aufgearbeitet und weiterverwendet werden können. Der vorliegende Artikel zielt darauf ab, eine Übersicht aktueller Track & Trace-Technologien zu geben und die Möglichkeiten aufzuzeigen, wie mittels einer zentralen Austauschplattform ein Mehrwert aus Daten generiert werden kann.
|[ 3 ]|| Krahe, C.; Iberl, M.; Jacob, A.; Lanza, G. & , . (2019), AI-based Computer Aided Engineering for automated product design - A first approach with a Multi-View based classification. Procedia CIRP, Volume 86, 2019, S. 104-109. 10.1016/j.procir.2020.01.038
Today’s success of industrial companies is largely determined by engineering competence and the digitization of all corporate processes. The design process and know-how of engineers is strongly individual and a rule-based description of their approach can often not be done at all or only with high effort. Existing knowledge can therefore only be passed on to other engineers with difficulty, which in particular increases the effort required for familiarisation. A further problem is the lack of an overview of existing components within a company, which very often leads to multiple designs and unnecessary waste of time for the engineer. The aim of this approach is to extract the implicit knowledge from existing CAD models with the aid of machine learning methods and thus to make it formalizable. In addition, a suitable classification and similarity analysis should quickly point out existing components. For this purpose, an AI-based assistance system is to be created. Based on the existing database, the assistant first points out to the engineer already existing, but very similar components. For that, the component type currently in construction firstly is identified and then very similar components are searched within the detected scale that are finally suggested to the engineer. The engineer now only has to parameterize the proposed components according to his application. In a further step, the assistant should also be able to suggest useful next design steps, which it has learned on the basis of the CAD data already available and their design history. The implicit experience knowledge that is contained in the existing CAD models thus ensures a design suitable for production and the avoidance of errors in the design.
|[ 4 ]|| Greinacher, S.; Overbeck, L.; Kuhnle, A.; Krahe, C. & Lanza, G. (2020), Multi-objective optimization of lean and resource efficient manufacturing systems, Production Engineering, Band 14, S. 165-176. 10.1007/s11740-019-00945-9
In the manufacturing industry, target-oriented and efficient use of resources is gaining importance, alongside economic optimization. The economic and organizational optimization of manufacturing systems according to the lean principles is only partly compatible with the goals of resource-efficient manufacturing. Therefore, an approach is sought to improve individual analyses of manufacturing systems. This paper proposes an approach for the multi-objective optimization of lean and resource-efficient manufacturing systems. To predict the dynamic effects of several configurations of manufacturing systems, material, energy, and information flows of a discrete event simulation are coupled with an assessment model, based on objectives of lean and resource-efficient manufacturing. Using design of experiments, Gaussian process meta-models are computed for the behavior of the simulation model. These meta-models allow the approximation of the system behavior to be computed in a short period of time and enable extensive multi-objective optimization and more adequate decision-making support systems. The proposed approach is tested in the metalworking industry.
|[ 5 ]|| Krahe, C.; Bräunche, A.; Jacob, A.; Stricker, N. & Lanza, G. (2020), Deep Learning for Automated Product Design.. Procedia CIRP, Volume 91, 2020, S. 03. Aug. 10.1016/j.procir.2020.01.135
In today’s business environment, the trend towards more product variety and customization is unbroken. Due to this development, the need of agile and reconfigurable production systems emerged to cope with various products and product families. To design and optimize production systems as well as to choose the optimal product matches, product analysis methods are needed. Indeed, most of the known methods aim to analyze a product or one product family on the physical level. Different product families, however, may differ largely in terms of the number and nature of components. This fact impedes an efficient comparison and choice of appropriate product family combinations for the production system. A new methodology is proposed to analyze existing products in view of their functional and physical architecture. The aim is to cluster these products in new assembly oriented product families for the optimization of existing assembly lines and the creation of future reconfigurable assembly systems. Based on Datum Flow Chain, the physical structure of the products is analyzed. Functional subassemblies are identified, and a functional analysis is performed. Moreover, a hybrid functional and physical architecture graph (HyFPAG) is the output which depicts the similarity between product families by providing design support to both, production system planners and product designers. An illustrative example of a nail-clipper is used to explain the proposed methodology. An industrial case study on two product families of steering columns of thyssenkrupp Presta France is then carried out to give a first industrial evaluati
|[ 6 ]|| Kandler, M.; Lanza, G. & Krahe, C. (2020), Development of a Human-centered Industry 4.0 Philosophy. New Developments in Sheet Metal Forming, S. 123-135.
Industry 4.0 is increasingly finding its way into production. New technologies and instruments are increasingly changing production processes and, in particular, production work. New instruments to support employees make it possible to increase labour productivity, which is a key factor contributing to economic growth. While at the beginning the benefits were unclear and the investment strategy was missing, today the lack of acceptance and unsuitable organizational structures in the companies are the biggest obstacles. This is reminiscent of the first failure of Lean Management. Only after several years with the anchoring of Lean principles as a philosophy and a sustainable improvement culture, Lean Management was successful. What is needed is an approach that involves employees in incremental process improvements using Industry 4.0 in order to support the acceptance. To develop this approach, the respective success factors of Lean Management and Industry 4.0 are first studied and then combined to form an integrated process optimization approach. The result is a human-centred optimization approach that promotes an increased acceptance of Industry 4.0 and supports employees in coping with future flexible work content.
|[ 7 ]|| Krahe, C.; Kalaidov, M.; Doellken, M.; Gwosch, T.; Kuhnle, A.; Lanza, G. & Matthiesen, S. (2021), AI-Based Knowledge Extraction for Automatic Design Proposals Using Design-Related Patterns. Procedia CIRP, S. 397-402. 10.1016/j.procir.2021.05.093
Engineering competence and the digitization of all processes along the product development process are highly decisive for today's success of industrial companies. The design process is very individual and strongly based on design engineers' experience. Part of this knowledge and the result of the design approach are fixated in the existing variations of the product generations, but are difficult to extract and to formalize. Conclusions about design-related patterns between products of different generations or variants can be drawn from the model tree representing the design engineer?s thinking process for each individual CAD model. However, the model tree has hardly been used so far. The aim of this paper is to examine whether there exist any common design patterns between CAD models of certain component classes by the exemplary use case in the area of mechanical engineering. To identify patterns and to extract knowledge out of complex data sets, Machine Learning (ML), especially Deep Learning, has proven an immense capability. Finally, based on the learned patterns, meaningful next design steps are to be proposed in the form of an assistance system. The results show that there exist common design patterns for various classes of components. It is illustrated on an exemplary component class that those patterns can be used to train an assistance system based on Recurrent Neural Networks (RNNs). The corresponding design patterns were extracted from data of an industrial application partner. By transferring these design patterns to the development of new product generations or variants, on the one hand the design process itself and thus the time to market can be shortened. On the other hand, the knowledge from previous product generations contained in those patterns can be preserved. For further research the design patterns of CAD models extracted by ML algorithms is a contribution to faster knowledge extrapolation.