Universität Stuttgart
Permanent URI for this communityhttps://elib.uni-stuttgart.de/handle/11682/1
Browse
11 results
Search Results
Item Open Access Distributed cooperative deep transfer learning for industrial image recognition(2020) Maschler, Benjamin; Kamm, Simon; Nasser, Jazdi; Weyrich, MichaelIn this paper, a novel light-weight incremental class learning algorithm for live image recognition is presented. It features a dual memory architecture and is capable of learning formerly unknown classes as well as conducting its learning across multiple instances at multiple locations without storing any images. In addition to tests on the ImageNet dataset, a prototype based upon a Raspberry Pi and a webcam is used for further evaluation: The proposed algorithm successfully allows for the performant execution of image classification tasks while learning new classes at several sites simultaneously, thereby enabling its application to various industry use cases, e.g. predictive maintenance or self-optimization.Item Open Access A non-intrusive nonlinear model reduction method for structural dynamical problems based on machine learning(2020) Kneifl, Jonas; Grunert, Dennis; Fehr, JörgThe paper uses a nonlinear non-intrusive model reduction approach, to derive efficient and accurate surrogate models for structural dynamical problems. Therefore, a combination of proper orthogonal decomposition along with regression algorithms from the field of machine learning is utilized to capture the dynamics in a reduced representation. This allows highly performant approximations of the original system. In this context, we provide a comparison of several regression algorithms based on crash simulations of a structural dynamic frame.Item Open Access Deep learning based soft sensors for industrial machinery(2020) Maschler, Benjamin; Ganssloser, Sören; Hablizel, Andreas; Weyrich, MichaelA multitude of high quality, high-resolution data is a cornerstone of the digital services associated with Industry 4.0. However, a great fraction of industrial machinery in use today features only a bare minimum of sensors and retrofitting new ones is expensive if possible at all. Instead, already existing sensors’ data streams could be utilized to virtually ‘measure’ new parameters. In this paper, a deep learning based virtual sensor for estimating a combustion parameter on a large gas engine using only the rotational speed as input is developed and evaluated. The evaluation focusses on the influence of data preprocessing compared to network type and structure regarding the estimation quality.Item Open Access Realization of AI-enhanced industrial automation systems using intelligent Digital Twins(2020) Nasser, Jazdi; Ashtari Talkhestani, Behrang; Maschler, Benjamin; Weyrich, MichaelA requirement of future industrial automation systems is the application of intelligence in the context of their optimization, adaptation and reconfiguration. This paper begins with an introduction of the definition of (artificial) intelligence to derive a framework for artificial intelligence enhanced industrial automation systems: An artificial intelligence component is connected with the industrial automation system’s control unit and other entities through a series of standardized interfaces for data and information exchange. This framework is then put into context of the intelligent Digital Twin architecture, highlight the latter as a possible implementation of such systems. Concluding, a prototypical implementation on the basis of a modular cyber-physical production system is described. The intelligent Digital Twin realized this way provides the four fundamental sub-processes of intelligence, namely observation, analysis, reasoning and action. A detailed description of all technologies used is given.Item Open Access On the solution of forward and inverse problems in possibilistic uncertainty quantification for dynamical systems(2020) Hose, Dominik; Hanss, MichaelIn this contribution, we adress an apparent lack of methods for the robust analysis of dynamical systems when neither a precise statistical nor an entirely epistemic description of the present uncertainties is possible. Relying on recent results of possibilistic calculus, we revisit standard prediction and filtering problems and show how these may be solved in a numerically exact way.Item Open Access Interdisziplinärer Informatikunterricht - zwischen Chance und Herausforderung(2024) Bahr, TobiasIn Europa wird Informatikunterricht disziplinär und interdisziplinär in der Sekundarstufe vermittelt. Seit 2018 gibt es das Profilfach Informatik, Mathematik, Physik (IMP) an Gymnasien in Baden-Württemberg. Bislang gibt es keine Evidenz zur Umsetzung des interdisziplinären Profilfaches. In einer Interviewstudie (N = 21) wurden IMP-Lehrpersonen zur ihrer Motivation, Qualifikation, Fachvernetzung und Umsetzung befragt. Die explorativen Ergebnisse zeigen, dass die unterrichtliche Umsetzung von fachgetrenntem Unterricht ohne interdisziplinäre Abstimmung bis hin zu einer fächerverbindenden und fächerüberschreitenden Umsetzung reicht. Interdisziplinäre Umsetzung wird als Chance gesehen. Mangelnde Abstimmung zwischen den Lehrpersonen führt zur Disziplinarität. Geschlechterungleichheit, Mängel in der Ausstattung und der volle Stundenverteilungsplan sind Herausforderungen. Die Ergebnisse deuten begrenzten Erfolg der interdisziplinären Umsetzung an.Item Open Access User-friendly, requirement-based assistance for production workforce using an asset administration shell design(2020) Al Assadi, Anwar; Fries, Christian; Fechter, Manuel; Maschler, Benjamin; Ewert, Daniel; Schnauffer, Hans-Georg; Zürn, Michael; Reichenbach, MatthiasFuture production methods like cyber physical production systems (CPPS), flexibly linked assembly structures and the matrix production are characterized by highly flexible and reconfigurable cyber physical work cells. This leads to frequent job changes and shifting work environments. The resulting complexity within production increases the risk of process failures and therefore requires longer job qualification times for workers, challenging the overall efficiency of production. During operation, cyber physical work cells generate data, which are specific to the individual process and worker. Based on the asset administration shell for Industry 4.0, this paper develops an administration shell for the production workforce, which contains personal data (e.g. qualification level, language skills, machine access, preferred display and interaction settings). Using worker and process specific data as well as personal data, allows supporting, training and instating workers according to their individual capabilities. This matching of machine requirements and worker skills serves to optimize the allocation of workers to workstations regarding the ergonomic workplace setup and the machine efficiency. This paper concludes with a user-friendly, intuitive design approach for a personalized machine user interface. The presented use-cases are developed and tested at the ARENA2036 (Active Research Environment for the Next Generation of Automobiles) research campus.Item Open Access Analysis of the SPH interpolation moments matrix with regard to the influences of the discretization error in adaptive simulations(2021) Heinzelmann, Pascal; Spreng, Fabian; Sollich, Daniel; Eberhard, Peter; Williams, John R.Item Open Access Effiziente Modellierung flexibler Robotersysteme zur Echtzeitsimulation am Beispiel eines Leichtbauroboters(2025) Hoschek, Sebastian; Rodegast, Philipp; Gesell, Jakob; Scheid, Jonas; Fehr, JörgDie Echtzeitsimulation mechanischer Systeme und deren digitale Zwillinge gewinnen in der Industrie zunehmend an Bedeutung. Sie ermöglichen unter anderem die Optimierung von Steuerungsalgorithmen, die Vorhersage des Systemverhaltens und die Implementierung von Regelstrategien in der Automatisierungstechnik. Ein Industriepartner entwickelt derzeit einen mobilen Leichtbauroboter für den Einsatz im Logistikbereich, bei dem die hohe Flexibilität der Struktur zu elastischen Durchbiegungen führt. Um die Genauigkeit und Leistungsfähigkeit des Roboters zu verbessern, ist eine präzise Modellierung dieser elastischen Effekte erforderlich. In dieser Arbeit werden zwei verschiedene Modellierungsansätze für die Echtzeitsimulation untersucht. Der erste basiert auf einer physikalischen White-Box-Modellierung als flexibles Mehrkörpersystem, wobei ein klassisches Finite-Elemente-Modell (FEM) durch Modellordnungsreduktion vereinfacht wird, um eine effiziente Berechnung zu ermöglichen. Der zweite Ansatz verwendet ein Finite-Segmente-Modell, das eine Parameteridentifikation erfordert, um eine realitätsgetreue Abbildung des Systemverhaltens zu gewährleisten. Beide Methoden werden auf den Leichtbauroboter angewendet und hinsichtlich ihrer Vor- und Nachteile verglichen. Wesentliche Kriterien sind dabei der Modellierungsaufwand, die Berechnungsgeschwindigkeit und die Genauigkeit der Simulationsergebnisse. Die Ergebnisse liefern eine Entscheidungsgrundlage zur Auswahl geeigneter Modellierungsmethoden in Echtzeitanwendungen.Item Open Access The size of the functional base of support decreases with age(2025) Sloot, Lizeth H.; Gerhardy, Thomas; Mombaur, Katja; Millard, MatthewFalls occur more often as we age. To identify people at risk of falling, balance analysis requires an accurate base-of-support model. We previously developed a functional base-of-support (fBOS) model for standing young adults and showed that its area is smaller than the footprint area. Our fBOS model is a polygon that contains centre-of-pressure trajectories recorded as standing participants move their bodies in the largest possible loop while keeping their feet flat on the ground. Here we assess how the size of the fBOS area changes with age by comparing 38 younger (YA), 14 middle-aged (MA), and 34 older adults (OA). The fBOS area is smaller in older adults: OA area is 58% of the YA area (p<0.001), and 59% of the MA area (p=0.001), with no difference between YA and MA. The reduction in fBOS area among the OA is primarily caused by a reduction in the length of the fBOS. In addition, among older adults smaller fBOS areas correlated with a lower score on the Short Physical Performance Battery (τ=0.28, p=0.04), a reduced walking speed (τ=0.25, p=0.04), and a higher frailty level (p=0.09). So that others can extend our work, we have made our fBOS models available online.