Informatik
Refine
Document Type
- Conference proceeding (498)
- Journal article (186)
- Book chapter (48)
- Doctoral Thesis (10)
- Book (3)
- Report (1)
- Working Paper (1)
Has full text
- yes (747) (remove)
Is part of the Bibliography
- yes (747)
Institute
- Informatik (747)
- Technik (2)
Publisher
- Springer (160)
- Hochschule Reutlingen (95)
- IEEE (89)
- Gesellschaft für Informatik e.V (59)
- Elsevier (42)
- Association for Computing Machinery (37)
- IARIA (26)
- RWTH Aachen (14)
- De Gruyter (13)
- SciTePress (11)
Literature reviews are essential for any scientific work, both as part of a dissertation or as a stand-alone work. Scientists benefit from the fact that more and more literature is available in electronic form, and finding and accessing relevant literature has become more accessible through scientific databases. However, a traditional literature review method is characterized by a highly manual process, while technologies and methods in big data, machine learning, and text mining have advanced. Especially in areas where research streams are rapidly evolving, and topics are becoming more comprehensive, complex, and heterogeneous, it is challenging to provide a holistic overview and identify research gaps manually. Therefore, we have developed a framework that supports the traditional approach of conducting a literature review using machine learning and text mining methods. The framework is particularly suitable in cases where a large amount of literature is available, and a holistic understanding of the research area is needed. The framework consists of several steps in which the critical mind of the scientist is supported by machine learning. The unstructured text data is transformed into a structured form through data preparation realized with text mining, making it applicable for various machine learning techniques. A concrete example in the field of smart cities makes the framework tangible.
The shift of populations to cities is creating challenges in many respects, thus leading to increasing demand for smart solutions of urbanization problems. Smart city applications range from technical and social to economic and ecological. The main focus of this work is to provide a systematic literature review of smart city research to answer two main questions: (1) How is current research on smart cities structured? and (2) What directions are relevant for future research on smart cities? To answer these research questions, a text-mining approach is applied to a large number of publications. This provides an overview and gives insights into relevant dimensions of smart city research. Although the main dimensions of research are already described in the literature, an evaluation of the relevance of such dimensions is missing. Findings suggest that the dimensions of environment and governance are popular, while the dimension of economy has received only limited attention.
The shift of populations to cities is creating challenges in many respects, thus leading to increasing demand for smart solutions of urbanization problems. Smart city applications range from technical and social to economic and ecological. The main focus of this work is to provide a systematic literature review of smart city research to answer two main questions: (1) How is current research on smart cities structured? And (2) What directions are relevant for future research on smart cities? To answer these research questions, a text-mining approach is applied to a large number of publications. This provides an overview and gives insights into relevant dimensions of smart city research. Although the main dimensions of research are already described in the literature, an evaluation of the relevance of such dimensions is missing. Findings suggest that the dimensions of environment and governance are popular, while the dimension of economy has received only limited attention.
The increasing urban population growth leads to challenges in cities in many aspects: Urbanisation problems such as excessive environmental pollution or increasing urban traffic demand new and innovative solutions. In this context, the concept of smart cities is discussed. An enabling element of the smart city concept is applying information technology (IT) to improve administrative efficiency and quality of life while reducing costs and resource consumption and ensuring greater citizen participation in administrative and urban development issues. While these smart city services are technologically studied and implemented, government officials, citizens or businesses are often unaware of the large variety of smart city service solutions. Therefore, this work deals with developing a smart city services catalogue that documents best practice services to create a platform that brings citizens, city government, and businesses together. Although the concept of IT service catalogues is not new and guidelines and recommendations for the design and development of service catalogues already exist in the corporate context, there is little work on smart city service catalogues. Therefore, approaches from agile software development and pattern research were adapted to develop the smart city service catalogue platform in this work.
Smart cities are considered data factories that generate an enormous amount of data from various sources. In fact data is the backbone of any smart services. Therefore, the strategic beneficial handling of this digital capital is crucial for cities. Some smart city pioneers have already written down their approach to data in the form of data strategies, but what should a city's data strategy include, and how can the goals and measures defined in the strategies be operationalized? This paper addresses these questions by looking closely at the data strategies of cities in Germany and the top three countries in the EU Digital Economy and Society Index. The in-depth analysis of 8 city data strategies has yielded 11 dimensions that cities should consider in their data strategy. These are relevance of data, principles, methods, data sharing, technology, data culture, data ethics, organizational structure, data security and privacy, collaborations, data literacy. In addition, data governance is a concept to put these 11 strategic dimensions into practice through standardization measures, training programs, and defining roles and responsibilities by developing a data catalog.
The benefits of urban data cannot be realized without a political and strategic view of data use. A core concept within this view is data governance, which aligns strategy in data-relevant structures and entities with data processes, actors, architectures, and overall data management. Data governance is not a new concept and has long been addressed by scientists and practitioners from an enterprise perspective. In the urban context, however, data governance has only recently attracted increased attention, despite the unprecedented relevance of data in the advent of smart cities. Urban data governance can create semantic compatibility between heterogeneous technologies and data silos and connect stakeholders by standardizing data models, processes, and policies. This research provides a foundation for developing a reference model for urban data governance, identifies challenges in dealing with data in cities, and defines factors for the successful implementation of urban data governance. To obtain the best possible insights, the study carries out qualitative research following the design science research paradigm, conducting semi-structured expert interviews with 27 municipalities from Austria, Germany, Denmark, Finland, Sweden, and the Netherlands. The subsequent data analysis based on cognitive maps provides valuable insights into urban data governance. The interview transcripts were transferred and synthesized into comprehensive urban data governance maps to analyze entities and complex relationships with respect to the current state, challenges, and success factors of urban data governance. The findings show that each municipal department defines data governance separately, with no uniform approach. Given cultural factors, siloed data architectures have emerged in cities, leading to interoperability and integrability issues. A city-wide data governance entity in a cross-cutting function can be instrumental in breaking down silos in cities and creating a unified view of the city’s data landscape. The further identified concepts and their mutual interaction offer a powerful tool for developing a reference model for urban data governance and for the strategic orientation of cities on their way to data-driven organizations.
Data governance have been relevant for companies for a long time. Yet, in the broad discussion on smart cities, research on data governance in particular is scant, even though data governance plays an essential role in an environment with multiple stakeholders, complex IT structures and heterogeneous processes. Indeed, not only can a city benefit from the existing body of knowledge on data governance, but it can also make the appropriate adjustments for its digital transformation. Therefore, this literature review aims to spark research on urban data governance by providing an initial perspective for future studies. It provides a comprehensive overview of data governance and the relevant facets embedded in this strand of research. Furthermore, it provides a fundamental basis for future research on the development of an urban data governance framework.
Human pose estimation (HPE) is integral to scene understanding in numerous safety-critical domains involving human-machine interaction, such as autonomous driving or semi-automated work environments. Avoiding costly mistakes is synonymous with anticipating failure in model predictions, which necessitates meta-judgments on the accuracy of the applied models. Here, we propose a straightforward human pose regression framework to examine the behavior of two established methods for simultaneous aleatoric and epistemic uncertainty estimation: maximum a-posteriori (MAP) estimation with Monte-Carlo variational inference and deep evidential regression (DER). First, we evaluate both approaches on the quality of their predicted variances and whether these truly capture the expected model error. The initial assessment indicates that both methods exhibit the overconfidence issue common in deep probabilistic models. This observation motivates our implementation of an additional recalibration step to extract reliable confidence intervals. We then take a closer look at deep evidential regression, which, to our knowledge, is applied comprehensively for the first time to the HPE problem. Experimental results indicate that DER behaves as expected in challenging and adverse conditions commonly occurring in HPE and that the predicted uncertainties match their purported aleatoric and epistemic sources. Notably, DER achieves smooth uncertainty estimates without the need for a costly sampling step, making it an attractive candidate for uncertainty estimation on resource-limited platforms.
Die Bereitstellung klinischer Informationen im Operationssaal ist ein wichtiger Aspekt zur Unterstützung des chirurgischen Teams. Die roboter-assistierte Ösophagusresektion ist ein besonders komplexer Eingriff, der Potenzial zur workflowbasierten Unterstützung bietet. Wir präsentieren erste Ergebnisse der Entwicklung eines Checklisten-Tools mit der zugrundeliegenden Modellierung des chirurgischen Workflows und Informationsbedarf der Chirurgen. Das Checklisten-Tool zeigt hierfür die durchzuführenden Schritte chronologisch an und stellt zusätzliche Informationen kontextadaptiert bereit. Eine automatische Dokumentation von Start- und Endzeiten einzelner OP-Phasen und Schritte soll zukünftige Prozessanalysen der Operation ermöglichen.
Zero or plus energy office buildings must have very high building standards and require highly efficient energy supply systems due to space limitations for renewable installations. Conventional solar cooling systems use photovoltaic electricity or thermal energy to run either a compression cooling machine or an absorption-cooling machine in order to produce cooling energy during daytime, while they use electricity from the grid for the nightly cooling energy demand. With a hybrid photovoltaic-thermal collector, electricity as well as thermal energy can be produced at the same time. These collectors can produce also cooling energy at nighttime by longwave radiation exchange with the night sky and convection losses to the ambient air. Such a renewable trigeneration system offers new fields of applications. However, the technical, ecological and economical aspects of such systems are still largely unexplored.
In this work, the potential of a PVT system to heat and cool office buildings in three different climate zones is investigated. In the investigated system, PVT collectors act as a heat source and heat sink for a reversible heat pump. Due to the reduced electricity consumption (from the grid) for heat rejection, the overall efficiency and economics improve compared to a conventional solar cooling system using a reversible air-to-water heat pump as heat and cold source.
A parametric simulation study was carried out to evaluate the system design with different PVT surface areas and storage tank volumes to optimize the system for three different climate zones and for two different building standards. It is shown such systems are technically feasible today. With a maximum utilization of PV electricity for heating, ventilation, air conditioning and other electricity demand such as lighting and plug loads, high solar fractions and primary energy savings can be achieved.
Annual costs for such a system are comparable to conventional solar thermal and solar electrical cooling systems. Nevertheless, the economic feasibility strongly depends on country specific energy prices and energy policy. However, even in countries without compensation schemes for energy produced by renewables, this system can still be economically viable today. It could be shown, that a specific system dimensioning can be found at each of the investigated locations worldwide for a valuable economic and ecological operation of an office building with PVT technologies in different system designs.
The global demand for resources such as energy, land, or water is constantly increasing. It is therefore not sur- prising that research on the Food-Energy-Water (FEW) nexus has become a scientific as well as a general focus in recent years. A significant increase in publications since 2015 can be observed, and it can be expected that this trend will continue. A multilevel (macro, meso, and micro) perspective is essential, as the FEW nexus has cross- sectoral interdependencies. Several review studies on the FEW nexus can be found in the literature, in general, it can be concluded that the FEW nexus is a multi-disciplinary and complex topic. The studies examined identify essential fields of action for research, policy, and society. However, questions such as what are the main research fields at each level? Is it possible to divide the research into specific clusters? and do the clusters correlate with the levels, and what are the methods of modeling used in the clusters and levels? are still not fully discussed in the literature. An extensive literature review was conducted to get insight into the existing research areas. Especially in such fields as the FEW nexus, the amount of literature can get huge, and a human could get lost analyzing the literature manually. For that, we created word clouds and performed a cluster- and network-analysis to support the selection of most relevant papers for a detailed reading. In 2021, the most publications were published, with 173 publications, which corresponds to a share of 26.6 %. There has been a significant increase since 2015, and it can be expected that this trend will continue in the coming years. Most of the first authors come from the USA (25.4 %), followed by China with 22.4 %. From the word cloud and the top 20 words, which appear in the title and abstract, it can be deduced that the topic water is the most represented. However, the terms system, resource, model, study, change, development, and management also appear to be very important, which indi- cates the importance of a holistic approach to the topic. In total 9 clusters could be identified at the different levels. It can be seen that three clusters form well. For the others, a rather diffuse picture can be observed. In order to find out which topics are hidden behind the individual clusters, 6 publications from each cluster were subjected to a more detailed examination. With these steps, a number of 54 publications were identified for de- tailed consideration. The modeling approaches that are currently being applied in research can be classified into domain-specific tools (e. g. global water models, crop models or global climate models) and into more general tools to perform for example a life cycle analysis, spatial analysis using geographic information system, or system dynamics for a general understanding of the links between the domains. With the domain-specific tools, detailed research questions can be addressed to answer questions for a specific domain. However, these tools have the disadvantage that especially the links between the sectors food, energy, and water are not fully considered. Many implementations that are made today are at lowest level (micro) relate to bounded spatial areas and are derived from macro and meso level goals.
The paper explains a workflow to simulate the food energy water (FEW) nexus for an urban district combining various data sources like 3D city models, particularly the City Geography Markup Language (CityGML) data model from the Open Geospatial Consortium, Open StreetMap and Census data. A long term vision is to extend the CityGML data model by developing a FEW Application Domain Extension (FEW ADE) to support future FEW simulation workflows such as the one explained in this paper. Together with the mentioned simulation workflow, this paper also identifies some necessary FEW related parameters for the future development of a FEW ADE. Furthermore, relevant key performance indicators are investigated, and the relevant datasets necessary to calculate these indicators are studied. Finally, different calculations are performed for the downtown borough Ville-Marie in the city of Montréal (Canada) for the domains of food waste (FW) and wastewater (WW) generation. For this study, a workflow is developed to calculate the energy generation from anaerobic digestion of FW and WW. In the first step, the data collection and preparation was done. Here relevant data for georeferencing, data for model set-up, and data for creating the required usage libraries, like food waste and wastewater generation per person, were collected. The next step was the data integration and calculation of the relevant parameters, and lastly, the results were visualized for analysis purposes. As a use case to support such calculations, the CityGML level of detail two model of Montréal is enriched with information such as building functions and building usages from OpenStreetMap. The calculation of the total residents based on the CityGML model as the main input for Ville-Marie results in a population of 72,606. The statistical value for 2016 was 89,170, which corresponds to a deviation of 15.3%. The energy recovery potential of FW is about 24,024 GJ/year, and that of wastewater is about 1,629 GJ/year, adding up to 25,653 GJ/year. Relating values to the calculated number of inhabitants in Ville-Marie results in 330.9 kWh/year for FW and 22.4 kWh/year for wastewater, respectively.
Reliable and accurate car driver head pose estimation is an important function for the next generation of advanced driver assistance systems that need to consider the driver state in their analysis. For optimal performance, head pose estimation needs to be non-invasive, calibration-free and accurate for varying driving and illumination conditions. In this pilot study we investigate a 3D head pose estimation system that automatically fits a statistical 3D face model to measurements of a driver’s face, acquired with a low-cost depth sensor on challenging real-world data. We evaluate the results of our sensor-independent, driver-adaptive approach to those of a state-of-the-art camera-based 2D face tracking system as well as a non-adaptive 3D model relative to own ground-truth data, and compare to other 3D benchmarks. We find large accuracy benefits of the adaptive 3D approach.
EAM ist ein holistischer Ansatz, um komplexe IT- und Unternehmensstrukturen darzustellen. Dabei ist es von zentraler Bedeutung, diese Strukturen möglichst komplett und übersichtlich zu visualisieren. Ein Ansatz, dies zu erreichen, ist eine multiperspektivische Darstellung von mehreren Views in einem Architekturcockpit. Dabei können mehrere Views simultan betrachtet und analysiert werden. Dadurch ist es möglich, die Auswirkungen einer Analyse des Views eines Stakeholders simultan aus den Views anderer Stakeholder betrachten zu können, um eventuelle Wechselwirkungen zu erkennen und einen allgemeinen Überblick über die Unternehmensarchitektur zu behalten. In dieser Arbeit zeigen wir, von der Konzeption über die Umsetzung bis zu einem Anwendungsbeispiel, wie ein solches Architekturcockpit realisiert werden kann.
In Folge der gegenwärtigen Digitalisierung in der produzierenden Industrie werden Anwendungen oder Services mit potentiell positiven Auswirkungen auf Faktoren wie Effektivität und Arbeitsqualität entwickelt. Ein geeigneter Ansatz zur Stärkung motivierender Aspekte im Arbeitskontext kann Gamification darstellen. In dieser Arbeit ist die initiale Konzeption und Evaluation eines Gamification-Ansatzes für Anwender eines KI-Service zur Maschinenoptimierung dargestellt und möglichen Anforderungen an ein Konzept zur Motivationssteigerung extrahiert.
Entrepreneurial software engineering: towards a hybrid development method for early-stage startups
(2021)
A considerable share of innovative software-intensive products is developed by startups. However, product development in an early-stage startup is not a sequential process. A business idea is usually based on a number of assumptions. The riskiest assumptions need to be tested. Depending on the test results, a product strategy may change several times. This raises the question of how to create sufficiently stable software using engineering principles despite a dynamic product strategy that is subject to many uncertainties. Hybrid development methods that combine agile aspects with classical engineering methods seem to be a good choice in such a start-up context. This paper proposes a lightweight hybrid development method that provides early-stage startups with a framework to support the development of single-feature minimum viable products. The method was derived from a start-up company's founding case and evaluated in expert interviews. The proposed method is intended to provide a basis for discussion between practitioners and scientists with the aim of better understanding the application of software engineering principles in software start-ups.
In dieser Ausarbeitung wird auf Visualisierungsmöglichkeiten von neuronalen Netzen eingegangen. Ein neuronales Netz scheint zuerst nicht von außen einsehbar und ist somit für viele eine Blackbox. Häufig genutzte Python-Bibliotheken, zum Beispiel TensorFlow, werden vorgestellt und deren Stärken wie auch Schwächen präsentiert. Anhand dieser werden bereits bestehende Visualisierungen gezeigt und ihr derzeitiger Einsatz wird erläutert. Durch einen Vergleich soll ersichtlich werden, welche Bibliothek am meisten Daten während des Trainings liefert, damit diese Informationen weiter verarbeitet werden. Diese Daten sollen so visualisiert werden, dass sie bei der Entwicklung eines neuronalen Netzes unterstützend sind. Ziel ist es, auf die Möglichkeiten einzugehen, welche geboten werden können. Durch eine Vereinfachung des Debuggings neuronaler Netze sollen weiterführende Entwicklungen in diese Richtung unterstützt werden.
In diesem Beitrag wird ein neuer Ansatz vorgestellt, welcher eine schwerkraftreduzierte Navigation innerhalb einer VR-Umgebung erlaubt, wie beispielsweise ein simulierter Mondspaziergang. Zur Navigation in der VR-Umgebung wird der Cyberith Virtualizer ein-gesetzt. Die Schwerkraftsimulation erfolgt mittels eines einstellbaren Gurtsystems, das anelastischen Seilen aufgehängt wird und abgestufte Schwerkraftkompensationen erlaubt. Als Umgebung wurde ein Raumschiffszenario sowie eine Mondoberfläche generiert. Hier sind in der aktuellen Anwendung einfache Interaktionen möglich. In Anlehnung an existierende Gravity Offload Systeme wird die Lösung ViRGOS bezeichnet. ViRGOS wurde bereits bei verschiedenen Besuchsterminen und Hochschulevents eingesetzt, so dass erste Rückmeldungen von Nutzern eingeholt werden konnten.
Lots of movies are produced every year, too many to watch all of them and in particular, to get an overview about the evolution of typical movie genres and actors playing in them. Moreover, it is a challenging problem to detect correlations among the movies and the actors in those movies, in particular, if we are interested in time-varying data patterns like trends, countertrends, or anomalies and outliers. Those correlations are specifically interesting if they can be inspected on different levels of granularity, e.g., temporal, but also hierarchical in form of country- or continent-based correlations. In this paper we describe the IMDb Explorer, a webbased visualization tool that consists of two major views denoted by the movie cosmos and the career lines. Both views are linked and interactively manipulable while a list of user-defined metrics are explorable. We illustrate the usefulness of the visualization tool by applying it to the entire movie database provided by IMDb.
Painting galleries typically provide a wealth of data composed of several data types. Those multivariate data are too complex for laymen like museum visitors to first, get an overview about all paintings and to look for specific categories. Finally, the goal is to guide the visitor to a specific painting that he wishes to have a more closer look on. In this paper we describe an interactive visualization tool that first provides such an overview and lets people experiment with the more than 41,000 paintings collected in the web gallery of art. To generate such an interactive tool, our technique is composed of different steps like data handling, algorithmic transformations, visualizations, interactions, and the human user working with the tool with the goal to detect insights in the provided data. We illustrate the usefulness of the visualization tool by applying it to such characteristic data and show how one can get from an overview about all paintings to specific paintings.
Public transport maps are typically designed in a way to support route finding tasks for passengers, while they also provide an overview about stations, metro lines, and city-specific attractions. Most of those maps are designed as a static representation, maybe placed in a metro station or printed in a travel guide. In this paper, we describe a dynamic, interactive public transport map visualization enhanced by additional views for the dynamic passenger data on different levels of temporal granularity. Moreover, we also allow extra statistical information in form of density plots, calendar-based visualizations, and line graphs. All this information is linked to the contextual metro map to give a viewer insights into the relations between time points and typical routes taken by the passengers. We also integrated a graph-based view on user-selected routes, a way to interactively compare those routes, an attribute- and property-driven automatic computation of specific routes for one map as well as for all available maps in our repertoire, and finally, also the most important sights in each city are included as extra information to include in a user-selected route. We illustrate the usefulness of our interactive visualization and map navigation system by applying it to the railway system of Hamburg in Germany while also taking into account the extra passenger data. As another indication for the usefulness of the interactively enhanced metro maps we conducted a controlled user experiment with 20 participants.
Human recognition is an important part of perception systems, such as those used in autonomous vehicles or robots. These systems often use deep neural networks for this purpose, which rely on large amounts of data that ideally cover various situations, movements, visual appearances, and interactions. However, obtaining such data is typically complex and expensive. In addition to raw data, labels are required to create training data for supervised learning. Thus, manual annotation of bounding boxes, keypoints, orientations, or actions performed is frequently necessary. This work addresses whether the laborious acquisition and creation of data can be simplified through targeted simulation. If data are generated in a simulation, information such as positions, dimensions, orientations, surfaces, and occlusions are already known, and appropriate labels can be generated automatically. A key question is whether deep neural networks, trained with simulated data, can be applied to real data. This work explores the use of simulated training data using examples from the field of pedestrian detection for autonomous vehicles. On the one hand, it is shown how existing systems can be improved by targeted retraining with simulation data, for example to better recognize corner cases. On the other hand, the work focuses on the generation of data that hardly or not occur at all in real standard datasets. It will be demonstrated how training data can be generated by targeted acquisition and combination of motion data and 3D models, which contain finely graded action labels to recognize even complex pedestrian situations. Through the diverse annotation data that simulations provide, it becomes possible to train deep neural networks for a wide variety of tasks with one dataset. In this work, such simulated data is used to train a novel deep multitask network that brings together diverse, previously mostly independently considered but related, tasks such as 2D and 3D human pose recognition and body and orientation estimation.
We present a multitask network that supports various deep neural network based pedestrian detection functions. Besides 2D and 3D human pose, it also supports body and head orientation estimation based on full body bounding box input. This eliminates the need for explicit face recognition. We show that the performance of 3D human pose estimation and orientation estimation is comparable to the state-of-the-art. Since very few data sets exist for 3D human pose and in particular body and head orientation estimation based on full body data, we further show the benefit of particular simulation data to train the network. The network architecture is relatively simple, yet powerful, and easily adaptable for further research and applications.
Model-guided Therapy and Surgical Workflow Systems are two interrelated research fields, which have been developed separately in the last years. To make full use of both technologies, it is necessary to integrate them and connect them to Hospital Information Systems. We propose a framework for integration of Model-guided Therapy in Hospital Information Systems based on the Electronic Medical Record, and a taskbased Workflow Management System, which is suitable for clinical end users. Two prototypes - one based on Business Process Modeling Language, one based on the serum-board - are presented. From the experience with these prototypes, we developed a novel personalized visualization system for Surgical Workflows and Model-guided Therapy. Key challenges for further development are automated situation detection and a common communication infrastructure.
Multi-dimensional patient data, such as time varying volume data, data of different imaging modalities, surface segmentations etc. are of growing importance in the clinical routine. For many use cases, it is of major importance to replicate a certain visualization of a data set created on one machine on a different computer using different software tools. Up until now, there exists no standardized methodology for this consistent presentation. We propose an extension of the Digital Imaging und Communications in Medicine (DICOM) called “Multi dimensional Presentation State” and outline scope and first results of the standardization process.
A configuration-management-database driven approach for fabric-process specification and automation
(2014)
In this paper we describe an approach that integrates a Configuration- Management-Database into fabric-process specification and automation in order to consider different conditions regarding to cloud-services. By implementing our approach, the complexity of fabric processes gets reduced. We developed a prototype by using formal prototyping principles as research methods and integrated the Configuration-Management-Database Command into the Workflow- Management-System Activiti. We used this prototype to evaluate our approach. We implemented three different fabric-processes and show that by using our approach the complexity of these three fabric-processes gets reduced.
Motivation: Aim of this project is the automatic classification of total hip endoprosthesis (THEP) components in 2D Xray images. Revision surgeries of total hip arthroplasty (THA) are common procedures in orthopedics and trauma surgery. Currently, around 400.000 procedures per year are performed in the United States (US) alone. To achieve the best possible result, preoperative planning is crucial. Especially if parts of the current THEP system are to be retained.
Methods: First, a ground truth based on 76 X-ray images was created: We used an image processing pipeline consisting of a segmentation step performed by a convolutional neural network and a classification step performed by a support vector machine (SVM). In total, 11 classes (5 pans and 6 shafts) shall be classified.
Results: The ground truth generated was of good quality even though the initial segmentation was performed by technicians. The best segmentation results were achieved using a U-net architecture. For classification, SVM architectures performed much better than additional neural networks.
Conclusions: The overall image processing pipeline performed well, but the ground truth needs to be extended to include a broader variability of implant types and more examples per training class.
Die digitale Transformation ist die heute vorherrschende geschäftliche Transformation, die einen starken Einfluss darauf hat, wie digitale Dienstleistungen und Produkte dienstleistungsdominant gestaltet werden. Eine beliebte zugrundeliegende Theorie der Wertschöpfung und des wirtschaftlichen Austauschs, die als dienstleistungsdominante Logik (S-D) bekannt ist, kann mit vielen erfolgreichen digitalen Geschäftsmodellen verbunden werden. Allerdings ist die S-D-Logik an sich abstrakt. Unternehmen können sie nicht ohne Weiteres als Instrument für die Innovation und Gestaltung von Geschäftsmodellen nutzen. Um dies zu ändern, wird eine umfassende Ideenfindungsmethode auf der Grundlage der S-D-Logik vorgeschlagen, die als service-dominantes Design (SDD) bezeichnet wird. SDD zielt darauf ab, Unternehmen beim Übergang zu einer service- und wertorientierten Perspektive zu unterstützen. Die Methode bietet eine vereinfachte Möglichkeit, den Ideenfindungsprozess auf der Grundlage von vier Modellkomponenten zu strukturieren. Jede Komponente besteht aus praktischen Implikationen, Hilfsfragen und Visualisierungstechniken, die aus einer Literaturrecherche, einer Anwendungsfallbewertung der digitalen Mobilität und einer Fokusgruppendiskussion abgeleitet wurden. SDD ist ein erster Schritt zu einem Toolset, das etablierte Unternehmen bei der Service- und Werteorientierung im Rahmen ihrer digitalen Transformation unterstützen kann.
Purpose
In recognising the key role of business intelligence and big data analytics in influencing companies’ decision-making processes, this paper aims to codify the main phases through which companies can approach, develop and manage big data analytics.
Design/methodology/approach
By adopting a research strategy based on case studies, this paper depicts the main phases and challenges that companies “live” through in approaching big data analytics as a way to support their decision-making processes. The analysis of case studies has been chosen as the main research method because it offers the possibility for different data sources to describe a phenomenon and subsequently to develop and test theories.
Findings
This paper provides a possible depiction of the main phases and challenges through which the approach(es) to big data analytics can emerge and evolve over time with reference to companies’ decision-making processes.
Research limitations/implications
This paper recalls the attention of researchers in defining clear patterns through which technology-based approaches should be developed. In its depiction of the main phases of the development of big data analytics in companies’ decision-making processes, this paper highlights the possible domains in which to define and renovate approaches to value. The proposed conceptual model derives from the adoption of an inductive approach. Despite its validity, it is discussed and questioned through multiple case studies. In addition, its generalisability requires further discussion and analysis in the light of alternative interpretative perspectives.
Practical implications
The reflections herein offer practitioners interested in company management the possibility to develop performance measurement tools that can evaluate how each phase can contribute to companies’ value creation processes.
Originality/value
This paper contributes to the ongoing debate about the role of digital technologies in influencing managerial and social models. This paper provides a conceptual model that is able to support both researchers and practitioners in understanding through which phases big data analytics can be approached and managed to enhance value processes.
Artificial intelligence (AI) is one of the most promising technologies of the post-pandemic era. Cloud computing technology can simplify the process of developing AI applications by offering a variety of services, including ready-to-use tools to train machine learning (ML) algorithms. However, comparing the vast amount of services offered by different providers and selecting a suitable cloud service can be a major challenge for many firms. Also in academia, suitable criteria to evaluate this type of service remain largely unclear. Therefore, the overall aim of this work has been to develop a framework to evaluate cloud-based ML services. We use Design Science Research as our methodology and conduct a hermeneutic literature review, a vendor analysis, as well as, expert interviews. Based on our research, we present a novel framework for the evaluation of cloud-based ML services consisting of six categories and 22 criteria that are operationalized with the help of various metrics. We believe that our results will help organizations by providing specific guidance on how to compare and select service providers from the vast amount of potential suppliers.
The digitization of factories will be a significant issue for the 2020s. New scenarios are emerging to increase the efficiency of production lines inside the factory, based on a new generation of robots’ collaborative functions. Manufacturers are moving towards data-driven ecosystems by leveraging product lifecycle data from connected goods. Energy-efficient communication schemes, as well as scalable data analytics, will support these various data collection scenarios. With augmented reality, new remote services are emerging that facilitate the efficient sharing of knowledge in the factory. Future communication solutions should generally ensure connectivity between the various production sites spread worldwide and new players in the value chain (e.g., suppliers, logistics) transparent, real-time, and secure. Industry 4.0 brings more intelligence and flexibility to production. Resulting in more lightweight equipment and, thus, offering better ergonomics. 5G will guarantee real-time transmissions with latencies of less than 1 ms. This will provide manufacturers with new possibilities to collect data and trigger actions automatically.
Theory and practice of implementing a successful enterprise IoT strategy in the industry 4.0 era
(2021)
Since the arrival of the internet and affordable access to technologies, digital technologies have occupied a growing place in industries, propelling us towards a 4th industrial revolution: Industry 4.0. In today’s era of digital upheaval, enterprises are increasingly undergoing transformations that are leading to their digitalization. The traditional manufacturing industry is in the throes of a digital transformation that is accelerated by exponentially growing technologies (e.g., intelligent robots, Internet of Things, sensors, 3D printing). Around the world, enterprises are in a frantic race to implement solutions based on IoT to improve their productivity, innovation, and reduce costs and improve their markets on the international scene. Considering the immense transformative potential that IoTs and big data have to bring to the industrial sector, the adoption of IoT in all industrial systems is a challenge to remain competitive and thus transform the industry into a smart factory. This paper presents the description of the innovation and digitalization process, following the Industry 4.0 paradigm to implement a successful enterprise IoT strategy.
Automatic anode rod inspection in aluminum smelters using deep-learning techniques: a case study
(2020)
Automatic fault detection using machine learning has become an exciting and promising area of research. This because it accurate and timely way to manage and classify with minimal human effort. In the computer vision community, deep-learning methods have become the most suitable approaches for this task. Anodes are large carbon blocks that are used to conduct electricity during the aluminum reduction process. The most basic function of anode rod inspection is to prevent a situation where the anode rod will not fit into the stub-holes of a new anode. It would be the case for a rod containing either severe toe-in, missing stubs, or a retained thimble on one or more stubs. In this work, to improve the accuracy of shape defect inspection for an anode rod, we use the Fast Region-based Convolutional Network method (Fast R-CNN), model. To train the detection model, we collect an image dataset composed of multi-class of anode rod defects with annotated labels. Our model is trained using a small number of samples, an essential requirement in the industry where the number of available defective samples is limited. It can simultaneously detect multi-class of defects of the anode rod in nearly real-time.
Perceptual integration of kinematic components in the recognition of emotional facial expressions
(2018)
According to a long-standing hypothesis in motor control, complex body motion is organized in terms of movement primitives, reducing massively the dimensionality of the underlying control problems. For body movements, this low dimensional organization has been convincingly demonstrated by the learning of low-dimensional representations from kinematic and EMG data. In contrast, the effective dimensionality of dynamic facial expressions is unknown, and dominant analysis approaches have been based on heuristically defined facial ‘‘action units,’’ which reflect contributions of individual face muscles. We determined the effective dimensionality of dynamic facial expressions by learning of a low dimensional model from 11 facial expressions. We found an amazingly low dimensionality with only two movement primitives being sufficient to simulate these dynamic expressions with high accuracy. This low dimensionality is confirmed statistically, by Bayesian model comparison of models with different numbers of primitives, and by a psychophysical experiment that demonstrates that expressions, simulated with only two primitives, are indistinguishable from natural ones.
In addition, we find statistically optimal integration of the emotion information specified by these primitives in visual perception. Taken together, our results indicate that facial expressions might be controlled by a very small number of independent control units, permitting very low dimensional parametrization of the associated facial expression.
Die Bedeutung und Stellung der Informationstechnologie erlebte in den letzten 60 Jahren einen fortlaufenden Wandel. Der anfänglich rein unterstützende Charakter entwickelte sich immer mehr zu einem wichtigen Bestandteil der Aufbau- und Ablauforganisation im Unternehmen. Ein definiertes IT-Servicemanagement im Unternehmen sieht sich mittlerweile gleichgeordnet mit den restlichen Fachabteilungen, tritt mit seinen Leistungen als Dienstleister auf und betrachtet die Fachabteilungen als „Kunde“. Neue Technologien und Innovationen und die daraus resultierenden Neudefinitionen bestehender Anforderungen sollen im Rahmen der Digitalisierung in Unternehmen positive Effekte zeigen. IT Infrastructure Library (ITIL) wird als Framework für IT-Servicemanagement in der Industrie und im öffentlichen Dienst genutzt. Der Ansatz von ITIL unterstützte den Kulturwandel und sensibilisierte das Management und die Mitarbeiter darin, serviceorientiert zu denken. Da dieser Ansatz einen vordefinierten, zyklischen Ablauf hat, könnten schnell eintreffende Kundenanforderungen nicht fristgerecht umgesetzt werden, weshalb agile Methoden wie der DevOps-Ansatz in den Vordergrund rücken. Die Herausforderung besteht darin, den Kulturwandel bei der Einführung von DevOps in bestehenden ITIL-Strukturen in Unternehmen zu fördern.
Clinical reading centers provide expertise for consistent, centralized analysis of medical data gathered in a distributed context. Accordingly, appropriate software solutions are required for the involved communication and data management processes. In this work, an analysis of general requirements and essential architectural and software design considerations for reading center information systems is provided. The identified patterns have been applied to the implementation of the reading center platform which is currently operated at the Center of Ophthalmology of the University Hospital of Tübingen.
In a world with rapidly changing customer requirements and the increased role of technology, companies need more flexible systems to adapt their processes and react dynamically to changes. Adaptive Case Management (ACM) comes into consideration by providing a concept to adapt to changing business conditions. Within our research project we did a first foundational evaluation of the potential of ACM in supporting unpredictable sales processes. Based on a set of criteria we tested the concept of ACM with the open source tool Cognoscenti. The evaluation gave us the possibility to experience the concept of ACM. Hence we were able to provide a statement about the potential of ACM within the context of an unpredictable sales process, setting the path to further research and discussion of ACM in the area of sales processes.
The ballistocardiography is a technique that measures the heart rate from the mechanical vibrations of the body due to the heart movement. In this work a novel noninvasive device placed under the mattress of a bed estimates the heart rate using the ballistocardiography. Different algorithms for heart rate estimation have been developed.
Globalization has increased the number of road trips and vehicles. The result has been an intensification of traffic accidents, which are becoming one of the most important causes of death worldwide. Traffic accidents are often due to human error, the probability of which increases when the cognitive ability of the driver decreases. Cognitive capacity is closely related to the driver’s mental state, as well as other external factors such as the CO2 concentration inside the vehicle. The objective of this work is to analyze how these elements affect driving. We have conducted an experiment with 50 drivers who have driven for 25 min using a driving simulator. These drivers completed a survey at the start and end of the experiment to obtain information about their mental state. In addition, during the test, their stress level was monitored using biometric sensors and the state of the environment (temperature, humidity and CO2 level) was recorded. The results of the experiment show that the initial level of stress and tiredness of the driver can have a strong impact on stress, driving behavior and fatigue produced by the driving test. Other elements such as sadness and the conditions of the interior of the vehicle also cause impaired driving and affect compliance with traffic regulations.
Introduction to the special issue on self‑managing and hardware‑optimized database systems 2022
(2023)
Data management systems have evolved in terms of functionality, performance characteristics, complexity, and variety during the last 40 years. Particularly, the relational database management systems and the big data systems (e.g., Key-Value stores, Document stores, Graph stores and Graph Computation Systems, Spark, MapReduce/Hadoop, or Data Stream Processing Systems) have evolved with novel additions and extensions. However, the systems administration and tasks have become highly complex and expensive, especially given the simultaneous and rapid hardware evolution in processors, memory, storage, or networking. These developments present new open problems and challenges to data management systems as well as new opportunities.
The SMDB (International Workshop on Self-Managing Database Systems) and HardBD&Active (Joint International Workshop on Big Data Management on Emerging Hardware and Data Management on Virtualized Active Systems) workshops organized in conjunction with the IEEE ICDE (International Conference on Data Engineering) offered two distinct platforms for examining the above system-related challenges from different perspectives. The SMDB workshop looks into developing autonomic or self-* features in database and data management systems to tackle complex administrative tasks, while the HardBD&Active workshop focuses on harnessing hardware technologies to enhance efficiency and performance of data processing and management tasks. As a result of these workshops, we are delighted to present the third special issue of DAPD titled “Self-Managing and Hardware-Optimized Database Systems 2022,” which showcases the best contributions from the SMDB 2021/2022 and HardBD&Active 2021/2022 workshops.
Data Integration of heterogeneous data sources relies either on periodically transferring large amounts of data to a physical Data Warehouse or retrieving data from the sources on request only. The latter results in the creation of what is referred to as a virtual Data Warehouse, which is preferable when the use of the latest data is paramount. However, the downside is that it adds network traffic and suffers from performance degradation when the amount of data is high. In this paper, we propose the use of a readCheck validator to ensure the timeliness of the queried data and reduced data traffic. It is further shown that the readCheck allows transactions to update data in the data sources obeying full Atomicity, Consistency, Isolation, and Durability (ACID) properties.
At DBKDA 2019, we demonstrated that StrongDBMS with simple but rigorous optimistic algorithms, provides better performance in situations of high concurrency than major commercial database management systems (DBMS). The demonstration was convincing but the reasons for its success were not fully analysed. There is a brief account of the results below. In this short contribution, we wish to discuss the reasons for the results. The analysis leads to a strong criticism of all DBMS algorithms based on locking, and based on these results, it is not fanciful to suggest that it is time to re-engineer existing DBMS.
This paper reviews suggestions for changes to database technology coming from the work of many researchers, particularly those working with evolving big data. We discuss new approaches to remote data access and standards that better provide for durability and auditability in settings including business and scientific computing. We propose ways in which the language standards could evolve, with proof-of-concept implementations on Github.
Recent work on database application development platforms has sought to include a declarative formulation of a conceptual data model in the application code, using annotations or attributes. Some recent work has used metadata to include the details of such formulations in the physical database, and this approach brings significant advantages in that the model can be enforced across a range of applications for a single database. In previous work, we have discussed the advantages for enterprise integration of typed graph data models (TGM), which can play a similar role in graphical databases, leveraging the existing support for the unified modelling language UML. Ideally, the integration of systems designed with different models, for example, graphical and relational database, should also be supported. In this work, we implement this approach, using metadata in a relational database management system (DBMS).
Semi-automated image data labelling using AprilTags as a pre-processing step for machine learning
(2019)
Data labelling is a pre-processing step to prepare data for machine learning. There are many ways to collect and prepare this data, but these are usually associated with a greater effort. This paper presents an approach to semi-automated image data labelling using AprilTags. The AprilTags attached to the object, which contain a unique ID, make it possible to link the object surfaces to a particular class. This approach will be implemented and used to label data of a stackable box.
The data is evaluated by training a You Only Look Once (YOLO) net, with a subsequent evaluation of the detection results. These results show that the semi-automatically collected and labelled data can certainly be used for machine learning. However, if concise features of an object surface are covered by the AprilTag, there is a risk that the concerned class will not be recognized. It can be assumed that the labelled data can not only be used for YOLO, but also for other machine learning approaches.
3D assisted 2D face recognition involves the process of reconstructing 3D faces from 2D images and solving the problem of face recognition in 3D. To facilitate the use of deep neural networks, a 3D face, normally represented as a 3D mesh of vertices and its corresponding surface texture, is remapped to image-like square isomaps by a conformal mapping. Based on previous work, we assume that face recognition benefits more from texture. In this work, we focus on the surface texture and its discriminatory information content for recognition purposes. Our approach is to prepare a 3D mesh, the corresponding surface texture and the original 2D image as triple input for the recognition network, to show that 3D data is useful for face recognition. Texture enhancement methods to control the texture fusion process are introduced and we adapt data augmentation methods. Our results show that texture-map-based face recognition can not only compete with state-of-the-art systems under the same precon ditions but also outperforms standard 2D methods from recent years.
We examine the role of communication from users on dropout from digital learning systems to answer the following questions: (1) how does the sentiment within qualitative signals (user comments) affect dropout rates? (2) does the variance in the proportion of positive and negative sentiments affect dropout rates? (3) how do quantitative signals (e.g. likes) moderate the effect of the qualitative signals? and (4) how does the effect of qualitative signals on dropout rates change across early and late stages of learning? Our hypotheses draws from learning theory and self-regulation theory, and were tested using data of 447 learning videos across 32 series of online tutorials, spanning 12 different fields of learning. The findings indicate a main effect of negative sentiment on dropout rates but no effect of positive sentiment on preventing dropout behaviour. This main effect is stronger in the early stages of learning and weakens at later stages. We also observe an effect of the extent of variance of positive and negative sentiments on dropout behaviour. The effects are negatively moderated by quantitative signals. Overall, making commenting more broad-based rather than polarised can be a useful strategy in managing learning, transferring knowledge, and building consensus.
Sleep is an important aspect in life of every human being. The average sleep duration for an adult is approximately 7 h per day. Sleep is necessary to regenerate physical and psychological state of a human. A bad sleep quality has a major impact on the health status and can lead to different diseases. In this paper an approach will be presented, which uses a long-term monitoring of vital data gathered by a body sensor during the day and the night supported by mobile application connected to an analyzing system, to estimate sleep quality of its user as well as give recommendations to improve it in real-time. Actimetry and historical data will be used to improve the individual recommendations, based on common techniques used in the area of machine learning and big data analysis.
Stress is becoming an important topic in modern life. The influence of stress results in a higher rate of health disorders such as burnout, heart problems, obesity, asthma, diabetes, depressions and many others. Furthermore individual’s behavior and capabilities could be directly affected leading to altered cognition, inappropriate decision making and problem solving skills. In a dynamic and unpredictable environment, such as automotive, this can result in a higher risk for accidents. Different papers faced the estimation as well as prediction of drivers’ stress level during driving. Another important question is not only the stress level of the driver himself, but also the influence on and of a group of other drivers in the near area. This paper proposes a system, which determines a group of drivers in a near area as clusters and it derives the individual stress level. This information will be analyzed to generate a stress map, which represents a graphical view about road section with a higher stress influence. Aggregated data can be used to generate navigation routes with a lower stress influence to decrease stress influenced driving as well as improve road safety.