Universität Stuttgart
Permanent URI for this communityhttps://elib.uni-stuttgart.de/handle/11682/1
Browse
Search Results
Item Open Access A systematic mapping study on development and use of AI planning tools(2021) Philippsohn, RobertArtificial intelligence (AI) planning is a big area in the AI field with many needs and special problems. Therefore, it needs tools to suit these special problems and request, as well as for trends in the AI planning community. Since 1971 there has been an influx of many tools that assist insolving planning problems and making plans. To give a better overview of the available landscape of AI planning tools this systematic mapping study was conducted and try also to shows what software engineering principles are used in creating the tools. We also try to depict in which industry domains the AI planning tools are used and how many papers mention the tools being used in the industry. In the end, we conclude that there are at least 106 different tools out there, with only a fraction being used in the industry. While only a small part of the tools are talked about being used in the industry, this small part is covering a wide array of industry domains.Item Open Access Development of an infrastructure for creating a behavioral model of hardware of measurable parameters in dependency of executed software(2021) Schwachhofer, DenisSystem-Level Test (SLT) gains traction not only in the industry but as of recently also in academia. It is used to detect manufacturing defects not caught by previous test steps. The idea behind SLT is to embed the Design Under Test (DUT) in an environment and running software on it that corresponds to its end-user application. But even though it is increasingly used in manufacturing since a decade there are still many open challenges to solve. For example, there is no coverage metric for SLT. Also, tests are not automatically generated but manually composed using existing operating systems and programs. This master thesis introduces the foundation for the AutoGen project, that will tackle the aforementioned challenges in the future. This foundation contains a platform for experiments and a workflow to generate Systems-on-Chip (SoCs). A case study is conducted to show an example on how on-chip sensors can be used in SLT applications to replace missing detailed technology-information. For the case study a “power devil” application has been developed that aims to keep the temperature of the Field Programmable Gate Array (FPGA) it runs on in a target range. The study shows an example on how software and parameters influence the extra-functional behavior of hardware.Item Open Access Ereignisbasierte Architektur für Quantenanwendungen(2021) Basaric, StefanIm Vergleich zu herkömmlichen Rechnern können mithilfe von Quantencomputern zum ersten Mal komplexe Probleme mit akzeptablen Berechnungszeiten gelöst werden. Diese werden heutzutage durch eine Vielzahl von öffentlichen Cloud-Diensten wie IBM Quantum, Amazon Braket oder Azure Quantum registrierten Nutzern verfügbar gemacht. Um ihre Experimente auf Quantencomputern durchführen zu können, müssen Nutzer Quantenschaltungen schreiben und an die von den Cloud-Diensten bereitgestellten Schnittstellen schicken. Die Quantenschaltungen kommen dabei zunächst in eine Warteschlange, bevor sie schließlich auf dem Quantencomputer ausgeführt werden. Das hat zur Folge, dass die Ausführung im Vergleich zur reinen Berechnungszeit auf dem Quantencomputer sehr lange dauern kann. Die aktuell verfügbaren Cloud-Dienste bieten derzeit keine Möglichkeit, die Quantenanwendungen ihrer Nutzer zu hosten und sie beim Eintritt von Ereignissen automatisch auszuführen. In dieser Arbeit wird ein Konzept für eine ereignisbasierte Architektur vorgestellt, welches die automatisierte Ausführung von Quantenanwendungen beim Eintritt von beliebigen Ereignissen ermöglicht. Zusätzlich wird ein anhand des Konzepts umgesetzter Prototyp präsentiert, welcher mithilfe von IBM Quantum und OpenWhisk die ereignisbasierte Ausführung von Quantenanwendungen trotz einiger Limitationen ermöglicht.Item Open Access A framework for distributed training of physics-informed neural networks using JAX(2021) Braun, JohannesThe intention of this thesis is to evaluate the high-performance machine learning framework JAX. In the course of this work, a physics-informed neural network that solves the Burgers’ equation is implemented. This problem is chosen, as it is a well known and researched numerical problem and thus allows for great comparability. Here, a basic version of the physics-informed neural network with Flax is first created, which is an ecosystem for JAX that allows to implement neural networks. This version was then first improved with the tools offered via JAX. Afterwards, a SPMD version of this physics-informed neural network is also implemented, where multiple graphics processor units are utilized in the training. Additionally, the physics-informed neural network is extended to predict the parameters of the partial differential equation that describes the Burgers’ equation. This was done by the physics-informed neural network, while still learning to estimate the Burgers’ equation. For the optimized basic physics-informed neural network and the physics-informed neural network that also estimates the parameters of the partial differential equation promising results with JAX were achieved. The outcome of the SPMD physics-informed neural network was dissatisfactory, as it did not yield any improvements compared to the basic version. Although, this might stem from the small amount of data points used for each iteration and further points discussed in this paper. Additionally, a caveat must be voiced, as it often becomes apparent that the documentations of JAX and Flax are a work in progress. Because of this, a lot of crucial features have to be found out by trial and error, while working with these frameworks. Yet still, JAXand hence also Flax are considered a compelling framework to implement high-performance neural networks. Especially because of its potent Autograd and straightforward XLA just in time compilation. Through these components a performant physics-informed neural network can be quickly setup as shown in this thesis. Here, Autograd aids in creating the necessary gradients for the physical loss. Whereas the XLA just in time compilation yields drastic improvements to the run time of the training performed on the physics-informed neural network. These features then lead to previously mentioned promising results for the basic physics-informed neural network and the physics-informed neural network that also estimates the parameters of the partial differential equation.Item Open Access Question answering on knowledge bases : A comparative study(2021) Kanjur, VishnudathaQuestion Answering intends to automatically extract accurate and relevant information as the answer to a particular question. A large amount of data from the Web is stored as Knowledge bases in a structured way. Question answering on Knowledge bases is a research field that involves multiple branches of computer science like natural language processing, information retrieval and artificial intelligence. Knowledge Base Question Answering (KBQA) research involves various challenges to be solved in multiple aspects. This thesis aimed to compare several state-of-the-art methods for single relation KBQA. The widely used standard single relation dataset, SimpleQuestions dataset was used in the study against Freebase Knowledge Base (KB). A comprehensive analysis of the underlying models and their architecture was performed. Furthermore, to identify the drawbacks and possible enhancements, several approaches for evaluating the models were explored. The results show how the models were performed and the suitability of considering them for solving real-world problems in question answering.Item Open Access Models for data-efficient reinforcement learning on real-world applications(2021) Dörr, Andreas; Toussaint, Marc (Prof. Dr.)Large-scale deep Reinforcement Learning is strongly contributing to many recently published success stories of Artificial Intelligence. These techniques enabled computer systems to autonomously learn and master challenging problems, such as playing the game of Go or complex strategy games such as Star-Craft on human levels or above. Naturally, the question arises which problems could be addressed with these Reinforcement Learning technologies in industrial applications. So far, machine learning technologies based on (semi-)supervised learning create the most visible impact in industrial applications. For example, image, video or text understanding are primarily dominated by models trained and derived autonomously from large-scale data sets with modern (deep) machine learning methods. Reinforcement Learning, on the opposite side, however, deals with temporal decision-making problems and is much less commonly found in the industrial context. In these problems, current decisions and actions inevitably influence the outcome and success of a process much further down the road. This work strives to address some of the core problems, which prevent the effective use of Reinforcement Learning in industrial settings. Autonomous learning of new skills is always guided by existing priors that allow for generalization from previous experience. In some scenarios, non-existing or uninformative prior knowledge can be mitigated by vast amounts of experience for a particular task at hand. Typical industrial processes are, however, operated in very restricted, tightly calibrated operating points. Exploring the space of possible actions or changes to the process naively on the search for improved performance tends to be costly or even prohibitively dangerous. Therefore, one reoccurring subject throughout this work is the emergence of priors and model structures that allow for efficient use of all available experience data. A promising direction is Model-Based Reinforcement Learning, which is explored in the first part of this work. This part derives an automatic tuning method for one of themostcommonindustrial control architectures, the PID controller. By leveraging all available data about the system’s behavior in learning a system dynamics model, the derived method can efficiently tune these controllers from scratch. Although we can easily incorporate all data into dynamics models, real systems expose additional problems to the dynamics modeling and learning task. Characteristics such as non-Gaussian noise, latent states, feedback control or non-i.i.d. data regularly prevent using off-the-shelf modeling tools. Therefore, the second part of this work is concerned with the derivation of modeling solutions that are particularly suited for the reinforcement learning problem. Despite the predominant focus on model-based reinforcement learning as a promising, data-efficient learning tool, this work’s final part revisits model assumptions in a separate branch of reinforcement learning algorithms. Again, generalization and, therefore, efficient learning in model-based methods is primarily driven by the incorporated model assumptions (e.g., smooth dynamics), which real, discontinuous processes might heavily violate. To this end, a model-free reinforcement learning is presented that carefully reintroduces prior model structure to facilitate efficient learning without the need for strong dynamic model priors. The methods and solutions proposed in this work are grounded in the challenges experienced when operating with real-world hardware systems. With applications on a humanoid upper-body robot or an autonomous model race car, the proposed methods are demonstrated to successfully model and master their complex behavior.Item Open Access Performance measurements for personalizable route planning for uncorrelated edge costs(2021) Bühler, FelixNowadays, ordinary route planners compute paths by choosing the shortest or fastest route. However, there exist additional metrics from which users with varying preferences could benefit. Personalized route planning offers the possibility to combine different metrics with personal preferences. Nevertheless, personalized route planning has mainly been tested with correlated metrics. But when including uncorrelated metrics, the computing time increases significantly. Previous work found that the speedup technique “Customizable Route Planning” can lead to feasible speedups for single metric calculations. Thus, in this work, we investigate how this speedup technique for Dijkstra improves the query performances of “Personalizable Route Planning” compared to “Personalizable Contraction Hierarchies”. Furthermore, we study the performances on uncorrelated metrics. We introduce a graph structure to compare the personalized speedup techniques “Personalizable Contraction Hierarchies”, “Personalizable Customizable Route Planning” and “Personalizable Route Planning”. Three graph partitioning algorithms have been implemented to realize “Customizable Route Planning”: K-means, Gonzales, and Merge. Our experiments show that Merge works well in combination with “Personalizable Contraction Hierarchies” preprocessing. We found that “Personalizable Customizable Route Planning” is a good alternative, as it uses much fewer edges for finding the costs of the shortest path. For uncorrelated metrics, “Personalizable Customizable Route Planning” and “Personalizable Route Planning” achieved speedups higher than “Personalizable Contraction Hierarchies”. Our contribution comprises a novel graph structure for comparing different Dijkstra variants. With our experiments, we provide a deeper understanding of the personalized route planning problem. Additionally, we propose improvements for “Personalizable Contraction Hierarchies” for less contracted graphs with uncorrelated metrics.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 Developing a multimodal feedback motion guidance system in VR for people with motion disabilities(2021) Wennrich, KevinMotion is an important aspect in the area of physiotherapy. The correctness of those motions is even more important, especially in the home exercises. In this thesis, the prototype of a multimodal guidance system in virtual reality, which tracks the movements of the users and compares it to the correct position in the field of physiotherapy exercises was created. The get the requirements for the system, people who needed to go to physiotherapy, because of an injury or a disability (stroke, MS, NPC), were interviewed, as well as a physiotherapist. Based on the results, we have implemented a virtual physiotherapist and the auditory guidance as two modalities. Further modalities have been the ghostarm and the haptic guidance as vibration bands. The prototype in which the user can choose and combine the guidances have been developed. The system, the modalities and its limits have been evaluated in a online study and a pilot study, with the results, that until now the ghostarm and virtual physiotherapist are the most liked guidances. A user study is planned for the future.Item Open Access Evaluating dynamic load balancing of ECM workload pattern employed in cloud environments managed by a Kubernetes/Docker eco-system(2021) Hagemann, PascalThe transition to cloud-based ECM solutions requires adaptation and enhancements for dynamic cloud environment. Cloud computing and containerization offer key concepts to create solutions for this task. They also open new opportunities to benefit from cloud usage by leveraging the pay-as-you-go model. Cloud users and providers both benefit from the increased efficiency of dynamic applications. But legacy applications are not yet able to leverage the benefits provided by cloud orchestration. This thesis, therefore researches the feasibility of a dynamic load balancing approach applied to an ECM application deployed into a cloud environment. To evaluate the approach, a prototype using open-source software is created on a Kubernetes orchestrated cluster. Previous work included the port of the containerized ECM application into the Kubernetes environment. The present prototype builds up on this approach by enhancing the ECM application components with metrics export capabilities. A monitoring system based on Prometheus is introduced to gather these metrics from the ECM application and other system components. Information provided by these metrics are used to add elasticity to application components. The prototype proves that dynamic load balancing of the ECM application in the cloud is feasible. Two major challenges for an efficient deployment of the application were identified, (1) the generation of useful metrics and (2) removing dependencies from individual components. Further research into optimizations of stateful service components is required. This further ensures an efficient usage in cloud based elastic topologies, especially considering stateful database applications.