Refine
Year of publication
Document Type
- Conference Proceeding (642) (remove)
Language
- English (492)
- German (149)
- Multiple languages (1)
Keywords
- 360-degree coverage (1)
- 3D Extended Object Tracking (1)
- 3D Extended Object Tracking (EOT) (2)
- 3D shape tracking (1)
- 3D ship detection (1)
- AAL (1)
- ADAM (1)
- AHI (1)
- Abrasive grain material (1)
- Abtragsprinzip (1)
Institute
- Fakultät Bauingenieurwesen (9)
- Fakultät Elektrotechnik und Informationstechnik (10)
- Fakultät Informatik (50)
- Fakultät Maschinenbau (9)
- Fakultät Wirtschafts-, Kultur- und Rechtswissenschaften (8)
- Institut für Angewandte Forschung - IAF (53)
- Institut für Optische Systeme - IOS (19)
- Institut für Strategische Innovation und Technologiemanagement - IST (29)
- Institut für Systemdynamik - ISD (64)
- Institut für Werkstoffsystemtechnik Konstanz - WIK (5)
Foil-air bearings (FABs) are predominantly used for high-speed, oil-free applications. Offering many advantages such as friction loss at high speeds, stability and price, they lack, however, load capacity as well as start-up and coast-down friction wear resistance.
The friction losses of FABs have been studied experimentally by many authors. In order to predict the friction and, consequently, the lifespan of a FAB, the start-up and coast-down regimes are modelled in such a way that allows for accurate, efficient simulation and later optimisation of lift-off speed and wear characteristics. The proposed simulation method applies the Kirchhoff-Love plate theory to the top foil mapping [20]. This system of differential equations is coupled with the underlying compliant foil to simulate the displacement due to the pressure buildup. Consequently, this coupled system allows for simulation from almost zero rounds per minute (rpm) to full speed. The underlying simulation model uses the finite difference method for spatial discretisation and a temporal explicit Runge-Kutta method.
Difficulties to overcome are the smooth combination of various friction regimes across the sliding surfaces as well as the synchronous coupling of Reynolds, deformation and kinematic equations with highly non-linear terms. Introducing an exponential pressure component based on Greenwood and Tripp’s theory avoids impingement between the rotor and foil.
Incremental one-class learning using regularized null-space training for industrial defect detection
(2024)
One-class incremental learning is a special case of class-incremental learning, where only a single novel class is incrementally added to an existing classifier instead of multiple classes. This case is relevant in industrial defect detection scenarios, where novel defects usually appear during operation. Existing rolled-out classifiers must be updated incrementally in this scenario with only a few novel examples. In addition, it is often required that the base classifier must not be altered due to approval and warranty restrictions. While simple finetuning often gives the best performance across old and new classes, it comes with the drawback of potentially losing performance on the base classes (catastrophic forgetting [1]). Simple prototype approaches [2] work without changing existing weights and perform very well when the classes are well separated but fail dramatically when not. In theory, null-space training (NSCL) [3] should retain the basis classifier entirely, as parameter updates are restricted to the null space of the network with respect to existing classes. However, as we show, this technique promotes overfitting in the case of one-class incremental learning. In our experiments, we found that unconstrained weight growth in null space is the underlying issue, leading us to propose a regularization term (R-NSCL) that penalizes the magnitude of amplification. The regularization term is added to the standard classification loss and stabilizes null-space training in the one-class scenario by counteracting overfitting. We test the method’s capabilities on two industrial datasets, namely AITEX and MVTec, and compare the performance to state-of-the-art algorithms for class-incremental learning.
We are interested in computing a mini-batch-capable end-to-end algorithm to identify statistically independent components (ICA) in large scale and high-dimensional datasets. Current algorithms typically rely on pre-whitened data and do not integrate the two procedures of whitening and ICA estimation. Our online approach estimates a whitening and a rotation matrix with stochastic gradient descent on centered or uncentered data. We show that this can be done efficiently by combining Batch Karhunen-Löwe-Transformation [1] with Lie group techniques. Our algorithm is recursion-free and can be organized as feed-forward neural network which makes the use of GPU acceleration straight-forward. Because of the very fast convergence of Batch KLT, the gradient descent in the Lie group of orthogonal matrices stabilizes quickly. The optimization is further enhanced by integrating ADAM [2], an improved stochastic gradient descent (SGD) technique from the field of deep learning. We test the scaling capabilities by computing the independent components of the well-known ImageNet challenge (144 GB). Due to its robustness with respect to batch and step size, our approach can be used as a drop-in replacement for standard ICA algorithms where memory is a limiting factor.
The proposed approach applies current unsupervised clustering approaches in a different dynamic manner. Instead of taking all the data as input and finding clusters among them, the given approach clusters Holter ECG data (long-term electrocardiography data from a holter monitor) on a given interval which enables a dynamic clustering approach (DCA). Therefore advanced clustering techniques based on the well known Dynamic Time Warping algorithm are used. Having clusters e.g. on a daily basis, clusters can be compared by defining cluster shape properties. Doing this gives a measure for variation in unsupervised cluster shapes and may reveal unknown changes in healthiness. Embedding this approach into wearable devices offers advantages over the current techniques. On the one hand users get feedback if their ECG data characteristic changes unforeseeable over time which makes early detection possible. On the other hand cluster properties like biggest or smallest cluster may help a doctor in making diagnoses or observing several patients. Further, on found clusters known processing techniques like stress detection or arrhythmia classification may be applied.
The ageing infrastructure in ports requires regular inspection. This inspection is currently carried out manually by divers who sense by hand the entire underwater infrastructure. This process is cost-intensive as it involves a lot of time and human resources. To overcome these difficulties, we propose to scan the above and underwater port structure with a Multi-SensorSystem, and -by a fully automated processto classify the obtained point cloud into damaged and undamaged zones. We make use of simulated training data to test our approach since not enough training data with corresponding class labels are available yet. To that aim, we build a rasterised heightfield of a point cloud of a sheet pile wall by cutting it into verticall slices. The distance from each slice to the corresponding line generates the heightfield. This latter is propagated through a convolutional neural network which detects anomalies. We use the VGG19 Deep Neural Network model pretrained on natural images. This neural network has 19 layers and it is often used for image recognition tasks. We showed that our approach can achieve a fully automated, reproducible, quality-controlled damage detection which is able to analyse the whole structure instead of the sample wise manual method with divers. The mean true positive rate is 0.98 which means that we detected 98 % of the damages in the simulated environment.
Targetless Lidar-camera registration is a repeating task in many computer vision and robotics applications and requires computing the extrinsic pose of a point cloud with respect to a camera or vice-versa. Existing methods based on learning or optimization lack either generalization capabilities or accuracy. Here, we propose a combination of pre-training and optimization using a neural network-based mutual information estimation technique (MINE [1]). This construction allows back-propagating the gradient to the calibration parameters and enables stochastic gradient descent. To ensure orthogonality constraints with respect to the rotation matrix we incorporate Lie-group techniques. Furthermore, instead of optimizing on entire images, we operate on local patches that are extracted from the temporally synchronized projected Lidar points and camera frames. Our experiments show that this technique not only improves over existing techniques in terms of accuracy, but also shows considerable generalization capabilities towards new Lidar-camera configurations.
The detection of anomalous or novel images given a training dataset of only clean reference data (inliers) is an important task in computer vision. We propose a new shallow approach that represents both inlier and outlier images as ensembles of patches, which allows us to effectively detect novelties as mean shifts between reference data and outliers with the Hotelling T2 test. Since mean-shift can only be detected when the outlier ensemble is sufficiently separate from the typical set of the inlier distribution, this typical set acts as a blind spot for novelty detection. We therefore minimize its estimated size as our selection rule for critical hyperparameters, such as, e.g., the size of the patches is crucial. To showcase the capabilities of our approach, we compare results with classical and deep learning methods on the popular datasets MNIST and CIFAR-10, and demonstrate its real-world applicability in a large-scale industrial inspection scenario.
The Role of Support-Activities for the successful Implementation of Internal Corporate Accelerators
(2018)
Uncertainty about the future requires companies to create discontinuous innovations. Established companies, however, struggle to do so; whereas independent startups seem to better cope with this. Consequently, established companies set up entrepreneurial initiatives to make use of startups' benefits. Consequently, this led-amongst others-to great interest in socalled corporate entrepreneurship (CE) programs and to the development and characterization of several different forms. Their processes to achieve certain objectives, yet, are still rather ineffective. Thus, considerations of the actions performed in preparation for and during CE programs could be one approach to improve this but are still absent today. Furthermore, the increasing use of several CE programs in parallel seems to bear the potential for synergies and, thus, more efficient use of resources. Aiming to provide insights to both issues, this study analyzes actions of CE programs, by looking at interviews with managers of seven corporate incubators and accelerator programs of five established German tech-companies.
Nowadays established companies use Corporate Entrepreneurship (CE) as a means to create discontinuous innovations. Many companies thereby even implement multiple CE units that typically involve several entrepreneurial activities. This explorative study aimed to identify the reasons why established companies implement multiple CE units concurrently. In conducting a comparative case study with eight companies from different industries, valuable insights for science and practice were gained. We provide an overview of different 11 reasons for implementing multiple CE units. This shows that the combination of CE units used by companies differs depending on the reason. It further allowed to derive general approaches of established companies to the implementation of CE units. Last, we identify the concept of co-specialization to be a central driver explaining the creation of the need to set up multiple units. We conclude by indicating implications and subjects for future research.
Corporate Entrepreneurship (CE) units have become an increasingly important part of established companies’ development activities enabling them to also create more discontinuous innovations. As a result, companies have developed and implemented different forms of CE units, such as corporate accelerators, incubators, startup supplier programs, and corporate venture capital. Driven by the need to innovate, companies have even begun to use multiple CE units simultaneously. However, this has not been empirically investigated yet. Thus, with this study, we aim to shed some light on this by investigating the parallel use of multiple CE units in the German business landscape. We conducted an extensive desk research, combining, coding, and analyzing different sources. We found that 55 out of 165 large established companies have multiple CE units, which allowed us to characterize the parallel use and identify differences and similarities, e.g., in terms of industry, company size, and CE forms implemented. We conclude by presenting different implications for both practice and research and by pointing out directions for future research.
Corporate Entrepreneurship (CE) became the new paradigm for organizations to cope with the accelerated development of innovations. Therefore, especially established organizations increasingly implement CE activities, even in combination. Scholars point out that a coordinated portfolio of CE activities could yield synergies and thus higher value for the organization and further call for more scientific examinations. This literature review aims to better the understanding of the combined and coordinated use of CE activities as well as about resulting synergies. Results show that there are only very few studies that addressed a combination and/or coordination of CE activities with respect to the creation of additional value, however, without empirical analyses. Yet, five categories of direct and indirect synergies could be derived. Discussing the results as well as the heterogenous use of terminology and concepts, this paper concludes with a research agenda for future analyses.
We present an approach to reduce the complexity of adjusting privacy preferences for multiple online social networks. To achieve this, we quantify the effect on privacy for choices that users make, and simplify configuration by introducing privacy configuration as a service. We present an algorithm that effectively measures privacy and adjusts privacy settings across social networks. The aim is to configure privacy with one click.
Dissipation of heat can be a major challenge when applying sensor systems outdoors under varying environmental conditions. Typically, complex software and knowledge is needed to optimize thermal management. In this paper it is shown how the thermal optimization of a LiDAR (light detection and ranging) sensor can be performed efficiently. This approach uses standard CAD (computer aided design) software, which is readily available, and saves time and cost as the thermal design can be optimized before experimental realisation. A four-step process was developed and realized: (i) Measurement of the thermal energy distribution of the current sensor design; (ii) Simulation of the time-dependant thermal behaviour using standard CAD software; (iii) Simulation of a thermally optimized design. This was compared quantitatively with the original design and was also used for verification of sufficient increase in heat dissipation; (iv) Experimental realisation and verification of the optimized design. It could be shown that the optimized prototype shows significantly improved thermal behaviour in accordance with the predictions from the simulations. The new LiDAR sensor shows lower heat generation and optimized dissipation of thermal energy which proofs the applicability of the approach to complex sensors.
Measuring the natural frequency of buildings and bridges is a possibility to get information about the stiffness of the construction. Decreasing stiffness can be detected be repeatedly measurements. Damaged parts of the construction or too high wood moisture can be reasons for decreasing stiffness. The earlier the failure is detected the better is the chance to repair it with low costs. The method of monitoring by repeatedly measuring the natural frequency is applied at timber bridges, especially on footbridges. As damages due to high wood moisture cannot be seen easily, measuring the natural frequency is a good possibility to detect them and then to repair them. Equations to calculate the natural frequencies regarding the damaged parts are shown and applied to a simple supported beam.
Sleep is essential to existence, much like air, water, and food, as we spend nearly one-third of our time sleeping. Poor sleep quality or disturbed sleep causes daytime solemnity, which worsens daytime activities' mental and physical qualities and raises the risk of accidents. With advancements in sensor and communication technology, sleep monitoring is moving out of specialized clinics and into our everyday homes. It is possible to extract data from traditional overnight polysomnographic recordings using more basic tools and straightforward techniques. Ballistocardiogram is an unobtrusive, non-invasive, simple, and low-cost technique for measuring cardiorespiratory parameters. In this work, we present a sensor board interface to facilitate the communication between force sensitive resistor sensor and an embedded system to provide a high-performing prototype with an efficient signal-to-noise ratio. We have utilized a multi-physical-layer approach to locate each layer on top of another, yet supporting a low-cost, compact design with easy deployment under the bed frame.
Simulationsmethoden bei der Entwicklung von Extremleichtbaukomponenten in Faserverbundbauweise
(2018)
The detection of differences between images of a printed reference and a reprinted wood decor often requires an initial image registration step. Depending on the digitalization method, the reprint will be displaced and rotated with respect to the reference. The aim of registration is to match the images as precisely as possible. In our approach, images are first matched globally by extracting feature points from both images and finding corresponding point pairs using the RANSAC algorithm. From these correspondences, we compute a global projective transformation between both images. In order to get a pixel-wise registration, we train a learning machine on the point correspondences found by RANSAC. The learning algorithm (in our case Gaussian process regression) is used to nonlinearly interpolate between the feature points which results in a high precision image registration method on wood decors.
Optical surface inspection: A novelty detection approach based on CNN-encoded texture features
(2018)
In inspection systems for textured surfaces, a reference texture is typically known before novel examples are inspected. Mostly, the reference is only available in a digital format. As a consequence, there is no dataset of defective examples available that could be used to train a classifier. We propose a texture model approach to novelty detection. The texture model uses features encoded by a convolutional neural network (CNN) trained on natural image data. The CNN activations represent the specific characteristics of the digital reference texture which are learned by a one-class classifier. We evaluate our novelty detector in a digital print inspection scenario. The inspection unit is based on a camera array and a flashing light illumination which allows for inline capturing of multichannel images at a high rate. In order to compare our results to manual inspection, we integrated our inspection unit into an industrial single-pass printing system.
Digital cameras are used in a large variety of scientific and industrial applications. For most applications the acquired data should represent the real light intensity per pixel as accurately as possible. However, digital cameras are subject to different sources of noise which distort the resulting image. Noise includes photon noise, fixed pattern noise and read noise. The aim of the radiometric calibration is to improve the quality of the resulting images by reducing the influence of the different types of noise on the measured data. In this paper, a new approach for the radiometric calibration of digital cameras using sparse Gaussian process regression is presented. Gaussian process regression is a kernel based supervised machine learning technique. It is used to learn the response of a camera system from a set of training images to allow for the calibration of new images. Compared to the standard Gaussian process method or flat field correction our sparse approach allows for faster calibration and higher reconstruction quality.
Digital bedruckte Oberflächen müssen strengen funktionalen und ästhetischen Anforderungen genügen. Diese Eigenschaften werden im Rahmen der Qualitätsprüfung kontrolliert. Hierbei wirken sich Oberflächendefekte oftmals erst dann aus, wenn diese auch vom Menschen wahrgenommen werden. Aufgrund der hohen Produktionsgeschwindigkeit kann eine solche Bewertung der Sichtbarkeit von Defekten bisher nur außerhalb des Produktionsflusses durch manuelle - subjektiv geprägte - Inspektion erfolgen. Ziel des Projektes ist (1) die Modellierung von Texturen in einer Form, die an das menschliche visuelle System angepasst ist und (2) die automatisierte Beurteilung der Wahrnehmung von Texturfehlern. Im Rahmen des Projekts wurde ein prototypisches System zur Inline-Erfassung von texturierten Oberflächen entwickelt. Auf Basis von realen Aufnahmen industriell produzierter Holzdekore wurde eine repräsentative Texturdatenbank erstellt. Gezeigt werden erste Resultate im Bereich der Defektdetektion auf Basis von statistischen Merkmalen. Diese Ergebnisse dienen als Grundlage für die spätere wahrnehmungsorientierte Bewertung. Letztlich sollen die im Rahmen des Projekts erlangten Ergebnisse in einen prototypischen Aufbau zur Inspektion von digital bedruckten Dekoren einfließen.
Was in Kommunen im benachbarten Ausland, bspw. der Schweiz, Österreich oder den Niederlanden offenbar seit vielen Jahren Stand der Technik ist, ist auf Deutschlands Kommunalstraßen eine „Sonderbauweise“: Betonfahrbahnen. Obwohl, gerade in den neuen Bundesländern diese Bauweise häufig zum Einsatz kam, wird sie heute auf größeren Flächen wie gesamte Straßenzüge, bspw. in Wohngebieten, kaum noch angewendet (Bild 1). Die Gründe liegen offenbar in den schlechten Erfahrungen hinsichtlich einer komfortablen Nutzungsdauer sowie der Notwendigkeit, einen „leichteren“ Zugriff auf die unter der Straßenoberfläche liegenden Ver- und Entsorgungseinrichtungen zu haben. Infolge vermehrt auftretender Schäden wie Spurrinnen, Verdrückungen und weiteren Schäden werden hochbelastete Verkehrsflächen wie Bushaltestellen, Busspuren und Kreisverkehre immer häufiger anstatt in Asphalt- oder Pflaster in Betonbauweise ausgeführt. Die Gründe für die Vernachlässigung der Betonbauweise im kommunalen Umfeld liegen neben einem geringen Erfahrungsschatz sicherlich auch in einer aufwändigeren Planung, höheren Ausführungskosten, einem komplexeren Einbau, gerade in Zusammenhang mit Einbauten, Instandhaltungsmaßnahmen und Maßnahmen Dritter an den Ver- und Entsorgungseinrichtungen. Offen ist auch der Nachweis, dass die Betonbauweise im Lebenszyklus wirtschaftlicher als beispielsweise die Asphaltbauweise ist. Die Thematik „Einsatz von Betonflächen in Kommunen“ ist sehr umfangreich und weitläufig und lässt sich mit einem Vortrag generell nicht behandeln. Mit nachfolgenden Ausführungen soll demnach grundsätzlich auf die Belange der Planung, dem Bau und der Wirtschaftlichkeit kommunaler Verkehrsflächen in Betonbauweise eingegangen werden. Es können leider nicht alle Besonderheiten und Einzelheiten wie Baustoffe (Glasfaser) Berücksichtigung finden. Ziel des Vortrages ist, generelle Möglichkeiten hinsichtlich dem Einsatz von Betonflächen im kommunalen Bereich aufzuzeigen. Besonderer Dank gilt dem Straßenbauamt Böblingen sowie Herrn Baudirektor Andreas Klein, dessen persönliche Erfahrungen hier einfließen durften.
Fatigue and drowsiness are responsible for a significant percentage of road traffic accidents. There are several approaches to monitor the driver’s drowsiness, ranging from the driver’s steering behavior to analysis of the driver, e.g. eye tracking, blinking, yawning or electrocardiogram (ECG). This paper describes the development of a low-cost ECG sensor to derive heart rate variability (HRV) data for the drowsiness detection. The work includes the hardware and the software design. The hardware has been implemented on a printed circuit board (PCB) designed so that the board can be used as an extension shield for an Arduino. The PCB contains a double, inverted ECG channel including low-pass filtering and provides two analog outputs to the Arduino, that combined them and performs the analog-to-digital conversion. The digital ECG signal is transferred to an NVidia embedded PC where the processing takes place, including QRS-complex, heart rate and HRV detection as well as visualization features. The compact resulting sensor provides good results in the extraction of the main ECG parameters. The sensor is being used in a larger frame, where facial-recognition-based drowsiness detection is combined with ECG-based detection to improve the recognition rate under unfavorable light or occlusion conditions.
Acoustic Echo Cancellation (AEC) plays a crucial role in speech communication devices to enable full-duplex communication. AEC algorithms have been studied extensively in the literature. However, device specific details like microphone or loudspeaker configurations are often neglected, despite their impact on the echo attenuation or near-end speech quality. In this work, we propose a method to investigate different loudspeaker-microphone configurations with respect to their contribution to the overall AEC performance. A generic AEC system consisting of an adaptive filter and a Wiener post filter is used for a fair comparison between different setups. We propose the near-end-to-residual-echo ratio (NRER) and the attenuation-of-near-end (AON) as quality measures for the full-duplex AEC performance.
This paper studies suitable models for the identification of nonlinear acoustic systems. A cascaded structure of nonlinear filters is proposed that contains several parallel branches, consisting of polynomial functions followed by a linear filter for each order of nonlinearity. The second order of nonlinearity is additionally modelled with a parallel branch, containing a Volterra filter. These are followed by a long linear FIR filter that is able to model the room acoustics. The model is applied to the identification of a tube power amplifier feeding a guitar loudspeaker cabinet in an acoustic room. The adaptive identification is performed by the normalized least mean square (NLMS) algorithm. Compared with a generalized polynomial Hammerstein (GPH) model, the accuracy in modelling the dedicated real world system can be improved to a greater extend than increasing the order of nonlinearity in the GPH model.
Motion estimation is an essential element for autonomous vessels. It is used e.g. for lidar motion compensation as well as mapping and detection tasks in a maritime environment. Because the use of gyroscopes is not reliable and a high performance inertial measurement unit is quite expensive, we present an approach for visual pitch and roll estimation that utilizes a convolutional neural network for water segmentation, a stereo system for reconstruction and simple geometry to estimate pitch and roll. The algorithm is validated on a novel, publicly available dataset recorded at Lake Constance. Our experiments show that the pitch and roll estimator provides accurate results in comparison to an Xsens IMU sensor. We can further improve the pitch and roll estimation by sensor fusion with a gyroscope. The algorithm is available in its implementation as a ROS node.
Three-dimensional ship localization with only one camera is a challenging task due to the loss of depth information caused by perspective projection. In this paper, we propose a method to measure distances based on the assumption that ships lie on a flat surface. This assumption allows to recover depth from a single image using the principle of inverse perspective. For the 3D ship detection task, we use a hybrid approach that combines image detection with a convolutional neural network, camera geometry and inverse perspective. Furthermore, a novel calculation of object height is introduced. Experiments show that the monocular distance computation works well in comparison to a Velodyne lidar. Due to its robustness, this could be an easy-to-use baseline method for detection tasks in navigation systems.
For European space missions the importance of electric propulsion is strongly growing and has recently experienced a real burst in the telecom market. The initial drivers of this development were programs of the European Space Agency and projects of the European national space agencies. In addition, electric propulsion is now on the priority list of European commercial satellite manufacturers. Actual programs target orbit raising and station keeping with full electric propulsion for telecom satellites. European space industry, represented by individual companies, has developed specific and generic solutions for the electronics dedicated to powering and controlling electric propulsion systems. The European Space Agency and the European Commission providing support for enabling technology related to Power Processing Units (PPUs) and increasing competitiveness.
Technology commercialization is described as the most dreadful challenge for technology-based entrepreneurs. The scarcity of resources and limited managerial experience make it a daunting task, putting in danger the whole firm emergence. Prior research has often build upon the resource-based view to propose that the new firms' performance is dependent on their initial resource endowments and configurations. Nevertheless, little is known on how the early-stage decisions of the entrepreneur might influence on the growth of the firm. Scholars have suggested that both technology and market orientation actions could influence the performance and growth of firms in this context; nevertheless, there is limited empirical evidence of the influence of these different orientations in the context of new technology-based firms (NTBFs). In this study we propose to explore the influence of technology and demand creation actions adopting a demand-side view. We use a longitudinal study on a panel dataset (2004-2007) with 249 U.S. new high-technology firms to test our hypothesis. The results point towards a rather limited influence of initial resource configurations, as well as an unexpected influence of market and technology orientation in the growth dimensions of an NTBF. The research holds implications for the management of new technology-based firms and for those interested in supporting the development of technology entrepreneurship.
Why does a teacher use drama elements in her language classes? This article critically reflects on the author's experiences with employing drama elements in a compulsory class of German as a Foreign Language at college level. Motives for using drama are discussed, with regard to possible positive effects for the teacher-such as rapport with learners and personal confidence-and potential benefits for the language learners-such as promoting vocabulary and grammar learning, fostering oral skills and increasing motivation. After that the drama activities used in class are briefly described. However, two years of experimentation have also shown up a number of problems with implementing drama elements in class and still leave questions open, requiring further iterations of the drama-based approach in this context. The paper was developed from a talk given at the Drama in Education Days in Konstanz 2018.
Ingenieure auf die Bühne
(2018)
Steps to the stage
(2017)
Domain-specific modelling is increasingly adopted in the software development industry. While open source metamodels like Ecore have a wide impact, they still have some problems. The independent storage of nodes (classes) and edges (references) is currently only possible with complex, specific solutions. Furthermore the developed models are stored in the extensible markup language (XML) data format, which leads to problems with large models in terms of scaling. In this paper we describe an approach that solves the problem of independent classes and references in metamodels and we store the models in the JavaScript Object Notation (JSON) data format to support high scalability. First results of our tests show that the developed approach works and classes and references can be defined independently. In addition, our approach reduces the amount of characters per model by a factor of approximately two compared to Ecore. The entire project is made available as open source under the name MoDiGen. This paper focuses on the description of the metamodel definition in terms of scaling.
Three-level inverters are used in electrical drive systems, as grid infeed inverter in PV power plants or as active power line filters. Up to now so called hard switching topologies have been used. A new 'Soft Switching Three Level Inverter (S3L Inverter)' which is now available provides reduced switching losses and higher efficiency. In this paper the S3L inverter is compared with a hard switching T-type inverter topology (H3L inverter). S3L inverters provide higher efficiency and additionally advantages in electromagnetic compatibility due to the soft switching performance, especially when using the 'Super Soft Switching Three Level Inverter (SS3L Inverter)'.
This paper focuses on the multivariable control of a drawing tower process. The nature of the process together with the differences in measurement noise levels that affect the variables to be controlled motivated the development of a new MPC algorithm. An extension of a multivariable predictive control algorithm with separated prediction horizons is proposed. The obtained experimental results show the usefulness of the proposed algorithm..
Das hier beschriebene und auf einem FPGA vom Typ Spartan-3A DSP realisierte System dient dazu, auf besonders effiziente Weise die Häufigkeitsverteilung nicht erkannter fehlerhafter Nachrichten mit verschiedenen CRCPolynomen
zu berechnen. Damit die Berechnung in möglichst kurzer Zeit stattfindet, wurde das System aus 64 parallel arbeitenden Instanzen von CRC-Findern in mehrstufiger Fließbandorganisation aufgebaut. In der hier beschriebenen Ausbaustufe erreicht das System eine Gesamtleistung von 6,4 ·109 Operationen in der Sekunde.
A method is investigated by which tight bounds on the range of a multivariate rational function over a box can be computed. The approach relies on the expansion of the numerator and denominator polynomials in Bernstein polynomials. Convergence of the bounds to the range with respect to degree elevation of the Bernstein expansion, to the width of the box and to subdivision are proven and the inclusion isotonicity of the related enclosure function is shown.
Corporate venturing is one way for corporations to
introduce strategic renewal into their business portfolios, which is
imperative for ongoing success in innovation-driven industries.
Prior research finds that corporate ventures should be separated
from the mainstream business in loosely coupled sub-units, but
scholars continue to discuss how loose or tight the ventures should
be to balance exploration and exploitation. Hence, the antecedents
for successful venture management are yet to be fully explored and
our study contributes to this effort. The study shows that
corporate venture success is enhanced when corporate
management grants job and strategic autonomy to the venture
managers. This is further amplified when corporate management
simultaneously imposes an exploitative policy that forces venture
managers to prioritize extensions to and improvements of existing
competences and product-market offerings.
Polysomnography is a gold standard for a sleep study, and it provides very accurate results, but its cost (both personnel and material) are quite high. Therefore, the development of a low-cost system for overnight breathing and heartbeat monitoring, which provides more comfort while recording the data, is a well-motivated challenge. The system proposed in this manuscript is based on the usage of resistive pressure sensors installed under the mattress. These sensors can measure slight pressure changes provoked during breathing and heartbeat. The captured signal requires advanced processing, like applying filters and amplifiers before the analog signal is ready for the next step. Then, the output signal is digitalized and further processed by an algorithm that performs a custom filtering before it can recognize breathing and heart rate in real-time. The result can be directly visualized. Furthermore, a CSV file is created containing the raw data, timestamps, and unique IDs to facilitate further processing. The achieved results are promising, and the average deviation from a reference device is about 4bpm.
Normal breathing during sleep is essential for people’s health and well-being. Therefore, it is crucial to diagnose apnoea events at an early stage and apply appropriate therapy. Detection of sleep apnoea is a central goal of the system design described in this article. To develop a correctly functioning system, it is first necessary to define the requirements outlined in this manuscript clearly. Furthermore, the selection of appropriate technology for the measurement of respiration is of great importance. Therefore, after performing initial literature research, we have analysed in detail three different methods and made a selection of a proper one according to determined requirements. After considering all the advantages and disadvantages of the three approaches, we decided to use the impedance measurement-based one. As a next step, an initial conceptual design of the algorithm for detecting apnoea events was created. As a result, we developed an activity diagram on which the main system components and data flows are visually represented.
In recent decades, it can be observed that a steady increase in the volume of tourism is a stable trend. To offer travel opportunities to all groups, it is also necessary to prepare offers for people in need of long-term care or people with disabilities. One of the ways to improve accessibility could be digital technologies, which could help in planning as well as in carrying out trips. In the work presented, a study of barriers was first conducted, which led to selecting technologies for a test setup after analysis. The main focus was on a mobile app with travel information and 360° tours. The evaluation results showed that both technologies could increase accessibility, but some essential aspects (such as usability, completeness, relevance, etc.) need to be considered when implementing them.
The process of restoring our body and brain from fatigue is directly depend-ing on the quality of sleep. It can be determined from the report of the sleep study results. Classification of sleep stages is the first step of this study and this includes the measurement of biovital data and its further processing.
In this work, the sleep analysis system is based on a hardware sensor net, namely a grid of 24 pressure sensors, supporting sleep phase recognition. In comparison to the leading standard, which is polysomnography, the proposed approach is a non-invasive system. It recognises respiration and body move-ment with only one type of low-cost pressure sensors forming a mesh archi-tecture. The nodes implement as a series of pressure sensors connected to a low-power and performant microcontroller. All nodes are connected via a system wide bus with address arbitration. The embedded processor is the mesh network endpoint that enables network configuration, storing and pre-processing of the data, external data access and visualization.
The system was tested by executing experiments recording the sleep of different healthy young subjects. The results obtained have indicated the po-tential to detect breathing rate and body movement. A major difference of this system in comparison to other approaches is the innovative way to place the sensors under the mattress. This characteristic facilitates the continuous using of the system without any influence on the common sleep process.
Healthy sleep is one of the prerequisites for a good human body and brain condition, including general well-being. Unfortunately, there are several sleep disorders that can negatively affect this. One of the most common is sleep apnoea, in which breathing is impaired. Studies have shown that this disorder often remains undiagnosed. To avoid this, developing a system that can be widely used in a home environment to detect apnoea and monitor the changes once therapy has been initiated is essential. The conceptualisation of such a system is the main aim of this research. After a thorough analysis of the available literature and state of the art in this area of knowledge, a concept of the system was created, which includes the following main components: data acquisition (including two parts), storage of the data, apnoea detection algorithm, user and device management, data visualisation. The modules are interchangeable, and interfaces have been defined for data transfer, most of which operate using the MQTT protocol. System diagrams and detailed component descriptions, including signal requirements and visualisation mockups, have also been developed. The system's design includes the necessary concepts for the implementation and can be realised in a prototype in the next phase.
The last decades have shown that the volume of tourism, in general, is constantly increasing (with some justified exceptions). To offer a possibility of travel for all groups of people, it is necessary to pay attention to accessibility. One of the possibilities for increasing accessibility is digital technologies, which could assist in planning and the implementation and completion of trips. To make a selection of technologies, first, a study of barriers was conducted, which was then analyzed, and finally, some technologies were made available in a test setup. A focus on two technologies was made: 360°-Tours and mobile app with the travel information. The two technologies were implemented and presented to the test subjects.
The evaluation results showed that both technologies could increase accessibility if some essential aspects (such as usability, completeness, relevance, etc.) are considered during the implementation.
This paper presents a bed system able to analyze a person’s movement, breathing and recognize the positions that the subject is lying on the bed during the night without any additional physical contact. The measurements are performed with sensors placed between the mattress and the bed-frame. An Intel Edison board was used as an endpoint that served as a communication node from the mesh network to external service. Two nodes and Intel Edison are attached to the bottom of the bed frame and they are connected to the sensors. First test results have indicated the potential of the proposed approach for the recognition of sleep positions with 83% of correct recognized positions.
This document presents an algorithm for a non-obtrusive recognition of Sleep/Wake states using signals derived from ECG, respiration, and body movement captured while lying in a bed. As a core mathematical base of system data analytics, multinomial logistic regression techniques were chosen. Derived parameters of the three signals are used as the input for the proposed method. The overall achieved accuracy rate is 84% for Wake/Sleep stages, with Cohen’s kappa value 0.46. The presented algorithm should support experts in analyzing sleep quality in more detail. The results confirm the potential of this method and disclose several ways for its improvement.
Die Schlafapnoe ist eine häufig auftretende Schlafstörung,
die unterschiedliche Auswirkungen auf unseren Alltag hat; so wurde z. B.
über eine Tagesschläfrigkeit von etwa 25 % der Patienten mit obstruktiver
Schlafapnoe (OSA) berichtet. Ziel dieser Arbeit ist die Entwicklung eines
Systems, das eine nichtinvasive Erkennung der Schlafapnoe in häuslicher
Umgebung ermöglichen soll.
Good sleep is crucial for a healthy life of every person. Unfortunately, its quality often decreases with aging. A common approach to measuring the sleep characteristics is based on interviews with the subjects or letting them fill in a daily questionnaire and afterward evaluating the obtained data. However, this method has time and personal costs for the interviewer and evaluator of responses. Therefore, it would be important to execute the collection and evaluation of sleep characteristics automatically. To do that, it is necessary to investigate the level of agreement between measurements performed in a traditional way using questionnaires and measurements obtained using electronic monitoring devices. The study presented in this manuscript performs this investigation, comparing such sleep characteristics as "time going to bed", "total time in bed", "total sleep time" and "sleep efficiency". A total number of 106 night records of elderly persons (aged 65+) were analyzed. The results achieved so far reveal the fact that the degree of agreement between the two measurement methods varies substantially for different characteristics, from 31 minutes of mean difference for "time going to bed" to 77 minutes for "total sleep time". For this reason, a direct exchange of objective and subjective measuring methods is currently not possible.
Sleep study can be used for detection of sleep quality and in general bed behaviors. These results can helpful for regulating sleep and recognizing different sleeping disorders of human. In comparison to the leading standard measuring system, which is Polysomnography (PSG), the system proposed in this work is a non-invasive sleep monitoring device. For continuous analysis or home use, the PSG or wearable Actigraphy devices tends to be uncomfortable. Besides, these methods not only decrease practicality due to the process of having to put them on, but they are also very expensive. The system proposed in this paper classifies respiration and body movement with only one type of sensor and also in a noninvasive way. The sensor used is a pressure sensor. This sensor is low cost and can be used for commercial proposes. The system was tested by carrying out an experiment that recorded the sleep process of a subject. These recordings showed excellent results in the classification of breathing rate and body movements.
Recognition of sleep and wake states is one of the relevant parts of sleep analysis. Performing this measurement in a contactless way increases comfort for the users. We present an approach evaluating only movement and respiratory signals to achieve recognition, which can be measured non-obtrusively. The algorithm is based on multinomial logistic regression and analyses features extracted out of mentioned above signals. These features were identified and developed after performing fundamental research on characteristics of vital signals during sleep. The achieved accuracy of 87% with the Cohen’s kappa of 0.40 demonstrates the appropriateness of a chosen method and encourages continuing research on this topic.
Das klinische Standardverfahren und Referenz der Schlafmessung und der Klassifizierung der einzelnen Schlafstadien ist die Polysomnographie (PSG). Alternative Ansätze zu diesem aufwändigen Verfahren könnten einige Vorteile bieten, wenn die Messungen auf eine komfortablere Weise durchgeführt werden. Das Hauptziel dieser Forschung Studie ist es, einen Algorithmus für die automatische Klassifizierung von Schlafstadien zu entwickeln, der ausschließlich Bewegungs- und Atmungssignale verwendet.
Identifikation von Schlaf- und Wachzuständen durch die Auswertung von Atem- und Bewegungssignalen
(2021)
The main aim of presented in this manuscript research is to compare the results of objective and subjective measurement of sleep quality for older adults (65+) in the home environment. A total amount of 73 nights was evaluated in this study. Placing under the mattress device was used to obtain objective measurement data, and a common question on perceived sleep quality was asked to collect the subjective sleep quality level. The achieved results confirm the correlation between objective and subjective measurement of sleep quality with the average standard deviation equal to 2 of 10 possible quality points.
Healthy sleep is required for sufficient restoration of the human body and brain. Therefore, in the case of sleep disorders, appropriate therapy should be applied timely, which requires a prompt diagnosis. Traditionally, a sleep diary is a part of diagnosis and therapy monitoring for some sleep disorders, such as cognitive behaviour therapy for insomnia. To automatise sleep monitoring and make it more comfortable for users, substituting a sleep diary with a smartwatch measurement could be considered. With the aim of providing accurate results, a study with a total of 30 night recordings was conducted. Objective sleep measurement with a Samsung Galaxy Watch 4 was compared with a subjective approach (sleep diary), evaluating the four relevant sleep characteristics: time of getting asleep, wake up time, sleep efficiency (SE), and total sleep time (TST). The performed analysis has demonstrated that the median difference between both measurement approaches was equal to 7 and 3 minutes for a time of getting asleep and wake up time correspondingly, which allows substituting a subjective measurement with a smartwatch. The SE was determined with a median difference between the two measurement methods of 5.22%. This result also implicates a possibility of substitution. Some single recordings have indicated a higher variance between the two approaches. Therefore, the conclusion can be made that a substitution provides reliable results primarily in the case of long-term monitoring. The results of the evaluation of the TST measurement do not allow to recommend substitution of the measurement method.
The recovery of our body and brain from fatigue directly depends on the quality of sleep, which can be determined from the results of a sleep study. The classification of sleep stages is the first step of this study and includes the measurement of vital data and their further processing. The non-invasive sleep analysis system is based on a hardware sensor network of 24 pressure sensors providing sleep phase detection. The pressure sensors are connected to an energy-efficient microcontroller via a system-wide bus. A significant difference between this system and other approaches is the innovative way in which the sensors are placed under the mattress. This feature facilitates the continuous use of the system without any noticeable influence on the sleeping person. The system was tested by conducting experiments that recorded the sleep of various healthy young people. Results indicate the potential to capture respiratory rate and body movement.
The importance of sleep for human life is enormous. It affects physical, mental, and psychological health. Therefore, it is vital to recognise sleep disorders in a timely manner in order to be able to initiate therapy. There are two methods for measuring sleep-related parameters - objective and subjective. Whether the substitution of a subjective method for an objective one is possible is investigated in this paper. Such replacement may bring several advantages, including increased comfort for the user. To answer this research question, a study was conducted in which 75 overnight recordings were evaluated. The primary purpose of this study was to compare both ways of measurement for total sleep time and sleep efficiency, which are essential parameters for, e.g., insomnia diagnosis and treatment. The evaluation results demonstrated that, on average, there are 32 minutes of difference between the two measurement methods when total sleep time is analysed. In contrast, on average, both measurement methods differ by 7.5% for sleep efficiency measurement. It should also be noted that people typically overestimate total sleep time and efficiency with the subjective method, where the perceived values are measured.
Health monitoring in a home environment can have broader use since it may provide continuous control of health parameters with relatively minor intrusiveness into regular life. This work aims to verify if it is possible to replace the typical in some sleep medicine areas subjective questioning by an objective measurement using electronic devices. For this purpose, a study was conducted with ten subjects, in which objective and subjective measurement of relevant sleep parameters took place. The results of both measurement methods were evaluated and analyzed. The results showed that while for some measures, such as Total Time in Bed, there is a high agreement between objective and subjective measurements, for others, such as sleep quality, there are significant differences. For this reason, currently, a combination of both measurement methods may be beneficial and provide the most detailed results, while a partial replacement can already reduce the number of questions at the subjective measurement by measurement through electronic devices.
The influence of sleep on human health is enormous. Accordingly, sleep disorders can have a negative impact on it. To avoid this, they should be identified and treated in time. For this purpose, objective (with an appropriate device) or subjective (based on perceived values) measurement methods are used for sleep analysis to understand the problem. The aim of this work is to find out whether an exchange of the two methods is possible and can provide reliable results. In accordance with this goal, a study was conducted with people aged over 65 years old (a total of 154 night-time recordings) in which both measurement methods were compared. Sleep questionnaires and electronic devices for sleep assessment placed under the mattress were applied to achieve the study aims. The obtained results indicated that the correlation between both measurement methods could be observed for sleep characteristics such as total sleep time, total time in bed and sleep efficiency. However, there are also significant differences in absolute values of the two measurement approaches for some subjects/nights, which leads us to conclude that the substitution is more likely to be considered in case of long-term monitoring where the trends are of more importance and not the absolute values for individual nights.
In many cases continuous monitoring of vital signals is required and low intrusiveness is an important requirement. Incorporating monitoring systems in the hospital or home bed could have benefits for patients and caregivers. The objective of this work is the definition of a measurement protocol and the creation of a data set of measurements using commercial and low-cost prototypes devices to estimate heart rate and breathing rate. The experimental data will be used to compare results achieved by the devices and to develop algorithms for feature extraction of vital signals.
This document presents a new complete standalone system for a recognition of sleep apnea using signals from the pressure sensors placed under the mattress. The developed hardware part of the system is tuned to filter and to amplify the signal. Its software part performs more accurate signal filtering and identification of apnea events. The overall achieved accuracy of the recognition of apnea occurrence is 91%, with the average measured recognition delay of about 15 seconds, which confirms the suitability of the proposed method for future employment. The main aim of the presented approach is the support of the healthcare system with the cost-efficient tool for recognition of sleep apnea in the home environment.
Sleep quality and in general, behavior in bed can be detected using a sleep state analysis. These results can help a subject to regulate sleep and recognize different sleeping disorders. In this work, a sensor grid for pressure and movement detection supporting sleep phase analysis is proposed. In comparison to the leading standard measuring system, which is Polysomnography (PSG), the system proposed in this project is a non-invasive sleep monitoring device. For continuous analysis or home use, the PSG or wearable Actigraphy devices tends to be uncomfortable. Besides this fact, they are also very expensive. The system represented in this work classifies respiration and body movement with only one type of sensor and also in a non-invasive way. The sensor used is a pressure sensor. This sensor is low cost and can be used for commercial proposes. The system was tested by carrying out an experiment that recorded the sleep process of a subject. These recordings showed the potential for classification of breathing rate and body movements. Although previous researches show the use of pressure sensors in recognizing posture and breathing, they have been mostly used by positioning the sensors between the mattress and bedsheet. This project however, shows an innovative way to position the sensors under the mattress.
Advanced approaches for analysis and form finding of membrane structures with finite elements
(2018)
Part I deals with material modelling of woven fabric membranes. Due to their structure of crossed yarns embedded in coating, woven fabric membranes are characterised by a highly nonlinear stress-strain behaviour. In order to determine an accurate structural response of membrane structures, a suitable description of the material behaviour is required. A linear elastic orthotropic model approach, which is current practice, only allows a relative coarse approximation of the material behaviour. The present work focuses on two different material approaches: A first approach becomes evident by focusing on the meso-scale. The inhomogeneous, however periodic structure of woven fabrics motivates for microstructural modelling. An established microstructural model is considered and enhanced with regard to the coating stiffness. Secondly, an anisotropic hyperelastic material model for woven fabric membranes is considered. By performing inverse processes of parameter identification, fits of the two different material models w.r.t. measured data from a common biaxial test are shown. The results of the inversely parametrised material models are compared and discussed.
Part II presents an extended approach for a simultaneous form finding and cutting patterning computation of membrane structures. The approach is formulated as an optimisation problem in which both the geometries of the equilibrium and cutting patterning configuration are initially unknown. The design objectives are minimum deviations from prescribed stresses in warp and fill direction along with minimum shear deformation. The equilibrium equations are introduced into the optimisation problem as constraints. Additional design criteria can be formulated (for the geometry of seam lines etc.). Similar to the motivation for the Updated Reference Strategy [4] the described problem is singular in the tangent plane. In both the equilibrium and the cutting patterning configuration finite element nodes can move without changing stresses. Therefore, several approaches are presented to stabilise the algorithm. The overall result of the computation is a stressed equilibrium and an unstressed cutting patterning geometry. The interaction of both configurations is described in Total Lagrangian formulation.
The microstructural model, which is focused in Part I, is applied. Based on this approach, information about fibre orientation as well as the ending of fibres at cutting edges are available. As a result, more accurate results can be computed compared to simpler approaches commonly used in practice.
Generative Design Software - How does digitalization change the professional profile of architects
(2019)
Abstract, Poster und Vortrag