11 Interfakultäre Einrichtungen
Permanent URI for this collectionhttps://elib.uni-stuttgart.de/handle/11682/12
Browse
81 results
Search Results
Item Open Access Concepts and methods for the design, configuration and selection of machine learning solutions in manufacturing(2021) Villanueva Zacarias, Alejandro Gabriel; Mitschang, Bernhard (Prof. Dr.-Ing. habil.)The application of Machine Learning (ML) techniques and methods is common practice in manufacturing companies. They assign teams to the development of ML solutions to support individual use cases. This dissertation refers as ML solution to the set of software components and learning algorithms to deliver a predictive capability based on available use case data, their (hyper) paremeters and technical settings. Currently, development teams face four challenges that complicate the development of ML solutions. First, they lack a formal approach to specify ML solutions that can trace the impact of individual solution components on domain-specific requirements. Second, they lack an approach to document the configurations chosen to build an ML solution, therefore ensuring the reproducibility of the performance obtained. Third, they lack an approach to recommend and select ML solutions that is intuitive for non ML experts. Fourth, they lack a comprehensive sequence of steps that ensures both best practices and the consideration of technical and domain-specific aspects during the development process. Overall, the inability to address these challenges leads to longer development times and higher development costs, as well as less suitable ML solutions that are more difficult to understand and to reuse. This dissertation presents concepts to address these challenges. They are Axiomatic Design for Machine Learning (AD4ML), the ML solution profiling framework and AssistML. AD4ML is a concept for the structured and agile specification of ML solutions. AD4ML establishes clear relationships between domain-specific requirements and concrete software components. AD4ML specifications can thus be validated regarding domain expert requirements before implementation. The ML solution profiling framework employs metadata to document important characteristics of data, technical configurations, and parameter values of software components as well as multiple performance metrics. These metadata constitute the foundations for the reproducibility of ML solutions. AssistML recommends ML solutions for new use cases. AssistML searches among documented ML solutions those that better fulfill the performance preferences of the new use case. The selected solutions are then presented to decision-makers in an intuitive way. Each of these concepts was evaluated and implemented. Combined, these concepts offer development teams a technology-agnostic approach to build ML solutions. The use of these concepts brings multiple benefits, i. e., shorter development times, more efficient development projects, and betterinformed decisions about the development and selection of ML solutions.Item Open Access Data processing, analysis, and evaluation methods for co-design of coreless filament-wound building systems(2023) Gil Pérez, Marta; Mindermann, Pascal; Zechmeister, Christoph; Forster, David; Guo, Yanan; Hügle, Sebastian; Kannenberg, Fabian; Balangé, Laura; Schwieger, Volker; Middendorf, Peter; Bischoff, Manfred; Menges, Achim; Gresser, Götz T.; Knippers, JanItem Open Access SyKonaS - Projektbericht. Nr. 1, Konflikte in der Energiewende: Definitionen und Typologien(Stuttgart : Verbundvorhaben SyKonaS, Zentrum für interdisziplinäre Risiko- und Innovationsforschung der Universität Stuttgart (ZIRIUS), 2022) Minn, Fabienne; Wassermann, Sandra; León, Christian D.; Püttner, Andreas (Mitwirkender); Liebhart, Laura (Mitwirkende); Wolf, Patrick (Mitwirkender)Das Forschungsprojekt "SyKonaS: Systemische Konfliktanalyse mittels Szenariotechnik" hat zum Ziel, gesellschaftliche Konflikte und deren Wechselwirkungen in der Energiewende zu verstehen, zu antizipieren und Lösungsvorschläge zu entwickeln. Im Rahmen dieser Zielsetzung wurden im Arbeitspaket 1 des Projektes die Konflikte der Energiewende empirisch aufgearbeitet und eine systematische Typologie von Energiewendekonflikten entwickelt (Task 1). Im vorliegenden Bericht werden das Vorgehen und die erzielten Ergebnisse beschrieben.Item Open Access Autonome Entscheidungsfindung in der Produktionssteuerung komplexer Werkstattfertigungen(Stuttgart : Fraunhofer Verlag, 2020) Waschneck, Bernd; Bauernhansl, Thomas (Prof. Dr.-Ing.)Die Variabilität in der kundenindividuellen Massenproduktion stellt eine enorme Herausforderung für die industrielle Fertigung dar. Die komplexe Werkstattfertigung als Produktionsprinzip eignet sich aufgrund der inhärenten Flexibilität besonders für die kundenindividuelle Massenproduktion. Allerdings sind die bestehenden Methodiken für die Produktionssteuerung einer Werkstattfertigung für die Einmal- oder Wiederholproduktion ausgelegt, was zu Defiziten in der Massenproduktion führt. Entweder ist die globale Qualität der Ergebnisse suboptimal oder die notwendige Echtzeitfähigkeit in der Entscheidungsfindung kann nicht bereitgestellt werden. Zudem entsteht durch Veränderungen und Anpassungen der Produktionssteuerung einer komplexen Werkstattfertigung ein hoher manueller Aufwand. In der vorliegenden Arbeit wird eine Methodik für eine dezentrale, selbstorganisierte und autonome Produktionssteuerung für eine Werkstattfertigung entwickelt, die dazu beiträgt, mit der zunehmenden Komplexität und dem Produktionsvolumen umzugehen. Dabei wird die Produktion als Reinforcement-Learning-Modell formalisiert, das die Grundlage für das autonome Lernen einer Strategie zur Optimierung der Abarbeitungsreihenfolge bildet. Mehrere kooperative Deep-Q-Network-Agenten werden in diesem Modell darauf trainiert, eine Strategie zu finden, die eine gegebene Bewertungsfunktion - meist ein Key Performance Indicator aus der Produktion - maximiert. Die Neuronalen Netze, in denen die erlernte Entscheidungslogik der Deep-Q-Network-Agenten abgebildet ist, werden nach der Trainingsphase in die Produktion übertragen. Der Multi-Agenten-Ansatz trägt dazu bei, dass der Lernvorgang beschleunigt wird und im produktiven Einsatz durch die Dezentralität Entscheidungen schneller bestimmt werden können. Die Erprobung der Methodik in zwei praxisnahen Fallbeispielen aus der Halbleiterindustrie zeigt ihre Leistungsfähigkeit. In beiden Fallbeispielen konnten Strategien zur Optimierung der Abarbeitungsreihenfolge auf oder über Expertenniveau autonom erlernt werden. Konkret konnte dadurch im zweiten Fallbeispiel der Anteil verspäteter Aufträge in einer Technologieklasse von 17, 0 % auf 1, 3 % reduziert werden. Abgerundet wird die Arbeit durch eine Einordnung in das soziotechnische System „Fabrik“, in der die Umsetzung der Reihenfolgeentscheidungen durch die Werker betrachtet wird. Dabei wird offensichtlich, dass die Optimierung der Produktionssteuerung ganzheitlich unter Einbeziehung der Werker in einem kontinuierlichen Verbesserungsprozess erfolgen muss.Item Open Access Temporally dense exploration of moving and deforming shapes(2020) Frey, SteffenWe present our approach for the dense visualization and temporal exploration of moving and deforming shapes from scientific experiments and simulations. Our image space representation is created by convolving a noise texture along shape contours (akin to LIC). Beyond indicating spatial structure via luminosity, we additionally use colour to depict time or classes of shapes via automatically customized maps. This representation summarizes temporal evolution, and provides the basis for interactive user navigation in the spatial and temporal domain in combination with traditional renderings. Our efficient implementation supports the quick and progressive generation of our representation in parallel as well as adaptive temporal splits to reduce overlap. We discuss and demonstrate the utility of our approach using 2D and 3D scalar fields from experiments and simulations.Item Open Access Stochastic model for energy propagation in disordered granular chains(2021) Taghizadeh, Kianoosh; Shrivastava, Rohit; Luding, StefanEnergy transfer is one of the essentials of mechanical wave propagation (along with momentum transport). Here, it is studied in disordered one-dimensional model systems mimicking force-chains in real systems. The pre-stressed random masses (other types of disorder lead to qualitatively similar behavior) interact through (linearized) Hertzian repulsive forces, which allows solving the deterministic problem analytically. The main goal, a simpler, faster stochastic model for energy propagation, is presented in the second part, after the basic equations are re-visited and the phenomenology of pulse propagation in disordered granular chains is reviewed. First, the propagation of energy in space is studied. With increasing disorder (quantified by the standard deviation of the random mass distribution), the attenuation of pulsed signals increases, transiting from ballistic propagation (in ordered systems) towards diffusive-like characteristics, due to energy localization at the source. Second, the evolution of energy in time by transfer across wavenumbers is examined, using the standing wave initial conditions of all wavenumbers. Again, the decay of energy (both the rate and amount) increases with disorder, as well as with the wavenumber. The dispersive ballistic transport in ordered systems transits to low-pass filtering, due to disorder, where localization of energy occurs at the lowest masses in the chain. Instead of dealing with the too many degrees of freedom or only with the lowest of all the many eigenmodes of the system, we propose a stochastic master equation approach with reduced complexity, where all frequencies/energies are grouped into bands. The mean field stochastic model, the matrix of energy-transfer probabilities between bands, is calibrated from the deterministic analytical solutions by ensemble averaging various band-to-band transfer situations for short times, as well as considering the basis energy levels (decaying with the wavenumber increasing) that are not transferred. Finally, the propagation of energy in the wavenumber space at transient times validates the stochastic model, suggesting applications in wave analysis for non-destructive testing, underground resource exploration, etc.Item Open Access Context scenarios of the German Energy Transition : a data collection for the analysis of the socio-political framework of a socio-technical transformation(2020) Weimer-Jehle, Wolfgang; Prehofer, Sigrid; Hauser, Wolfgang; Bräutigam, Klaus-Rainer (Translator); Buchgeister, Jens (Translator); Kopfmüller, Jürgen (Translator)An expert survey about the socio-technical context of the German Energy Transformation is described and selected results are reported. Major socio-technical drivers of the energy system and its evolution were identified, alternative futures for each driver were derived based on literature review and expert questioning. Using the framework of Cross-Impact Balance Analysis, the interrelations between the possible futures of the drivers were estimated by a series of expert interviews.Item Open Access SyKonaS - Projektbericht. Nr. 5, Systemische Analyse der Wechselwirkungen zwischen Konfliktlinien und Rahmenbedingungen der Energiewende: Weiterentwicklung soziotechnischer Energieszenarien(Stuttgart : Verbundvorhaben SyKonaS, Zentrum für interdisziplinäre Risiko- und Innovationsforschung der Universität Stuttgart (ZIRIUS), 2024) Hauser, Wolfgang; Wassermann, Sandra; Oviedo, Patricia; León, Christian D.; Weimer-Jehle, Wolfgang; Jaschek, Carolin (Mitwirkende); Prehofer, Sigrid (Mitwirkende)Im Teilvorhaben SyKonaS/iKonS („Systemische Konfliktanalyse mittels Szenariotechnik“) wurden die technoökonomischen Szenarien um soziale Größen ergänzt und zu soziotechnischen Szenarien weiterentwickelt, um ihre jeweilige Konflikthaftigkeit abzuschätzen. Hierfür wurden zwölf durch die technoökonomischen Energieszenarien vorgegebene Größen (wie z.B. die installierte Leistung von Wind onshore im Jahr 2050) und ihre Wirkungen auf zwölf sozio-politische Größen abgeschätzt. Ebenso wurden die Interdependenzen der soziopolitischen Größen erhoben. Es wurde dann die Frage gestellt, welche Konflikte von den verschiedenen Energieszenarien und ihren gesellschaftlichen Wirkungen ausgelöst werden können.Item Open Access Linking qualitative scenarios with quantitative energy models: knowledge integration in different methodological designs(2021) Prehofer, Sigrid; Kosow, Hannah; Naegler, Tobias; Pregger, Thomas; Vögele, Stefan; Weimer-Jehle, WolfgangLinking qualitative scenarios with quantitative models is a common approach to integrate assumptions on possible future societal contexts into modeling. But reflection on how and to what degree knowledge is effectively integrated during this endeavor does not generally take place. In this paper, we reflect on the performance of a specific hybrid scenario approach (qualitative Cross-Impact Balance analysis, CIB, linked with quantitative energy models) concerning knowledge integration through eleven different process steps. In order to guide the scenario community in applying this approach, we reflect on general methodological features as well as different design options. We conceptualize different forms of interdisciplinary knowledge integration (compiling, combining and synthesizing) and analyze how and to what degree knowledge about society and uncertainty are integrated into scenario process and products. In addition, we discuss trade-offs regarding design choices and forms of knowledge integration. On the basis of three case studies we identify two general designs of linking which build on each other (basic and extended design) and which differ in essence regarding the balance of power between the CIB and the energy modeling. Ex-post assessment of the form of interdisciplinary knowledge integration in each step revealed that specific method properties of CIB as well as the interaction with additional quantitative as well as specific qualitative methods foster distinct forms of knowledge integration. The specific roles assigned to CIB in the hybrid scenario process can also influence the form of knowledge integration. In this study, we use a joint process scheme linking qualitative context scenarios with energy modeling. By applying our conceptualization of different forms of knowledge integration we analyze the designs´ respective potential for and respective effects on knowledge integration. Consequently, our findings can give guidance to those who are designing their own hybrid scenario processes. As this is an explorative study, it would be useful to further test our hypotheses in different hybrid scenario designs. Finally, we note that at some points in the process a more precise differentiation of three forms of knowledge integration would have been useful and propose to further differentiate and detail them in future research.Item Open Access Simulation model for digital twins of pneumatic vacuum ejectors(2022) Stegmaier, Valentin; Schaaf, Walter; Jazdi, Nasser; Weyrich, MichaelIncreasing productivity, as well as flexibility, is required for the industrial production sector. To meet these challenges, concepts in the field of “Industry 4.0” are arising, such as the concept of Digital Twins. Vacuum handling systems are a widespread technology for material handling in industry and face the same challenges and opportunities. In this field, a key issue is the lack of Digital Twins containing behavior models for vacuum handling systems and their components in different applications and use cases. A novel concept for modeling and simulating the fluidic behavior of pneumatic vacuum ejectors as key components of vacuum handling systems is proposed. In order to increase the simulation accuracy, the concept can access instance‐specific data of the used asset instead of object‐specific data. The model and the data are part of the Digital Twins of pneumatic vacuum ejectors, which shall be able to be combined with other components to represent a Digital Twin of entire vacuum handling systems. The proposed model is validated in an experimental test setup and in an industrial application delivering sufficiently accurate results.