Refine
Year of publication
- 2022 (341) (remove)
Document Type
- Conference Proceeding (106)
- Article (reviewed) (72)
- Part of a Book (63)
- Article (unreviewed) (23)
- Other (20)
- Book (18)
- Contribution to a Periodical (11)
- Working Paper (8)
- Report (7)
- Patent (5)
Conference Type
- Konferenzartikel (87)
- Konferenz-Abstract (13)
- Konferenz-Poster (3)
- Sonstiges (3)
Is part of the Bibliography
- yes (341) (remove)
Keywords
- COVID-19 (10)
- injury (10)
- Digitalisierung (8)
- biomechanics (7)
- running (7)
- 3D printing (6)
- Machine Learning (6)
- Entrepreneurship (5)
- ACL (4)
- Robustness (4)
Institute
- Fakultät Maschinenbau und Verfahrenstechnik (M+V) (108)
- Fakultät Elektrotechnik, Medizintechnik und Informatik (EMI) (ab 04/2019) (87)
- Fakultät Medien (M) (ab 22.04.2021) (87)
- Fakultät Wirtschaft (W) (64)
- INES - Institut für nachhaltige Energiesysteme (33)
- POIM - Peter Osypka Institute of Medical Engineering (21)
- ivESK - Institut für verlässliche Embedded Systems und Kommunikationselektronik (16)
- IMLA - Institute for Machine Learning and Analytics (14)
- ACI - Affective and Cognitive Institute (7)
- CRT - Campus Research & Transfer (4)
Open Access
- Closed (165)
- Open Access (143)
- Bronze (51)
- Closed Access (33)
- Diamond (29)
- Gold (23)
- Hybrid (15)
- Grün (7)
Positioning mobile systems with high accuracy is a prerequisite for intelligent autonomous behavior, both in industrial environments and in field robotics. This paper describes the setup of a robotic platform and its use for the evaluation of simultaneous localization and mapping (SLAM) algorithms. A configuration using a mobile robot Husky A200, and a LiDAR (light detection and ranging) sensor was used to implement the setup. For verification of the proposed setup, different scan matching methods for odometry determination in indoor and outdoor environments are tested. An assessment of the accuracy of the baseline 3D-SLAM system and the selected evaluation system is presented by comparing different scenarios and test situations. It was shown that the hdl_graph_slam in combination with the LiDAR OS1 and the scan matching algorithms FAST_GICP and FAST_VGICP achieves good mapping results with accuracies up to 2 cm.
Seismic data has often missing traces due to technical acquisition or economical constraints. A compete dataset is crucial in several processing and inversion techniques. Deep learning algorithms, based on convolutional neural networks (CNNs), have shown alternative solutions that overcome limitation of traditional interpolation methods e.g. data regularity, linearity assumption, etc. There are two different paradigms of CNN methods for seismic interpolation. The first one, so-called deep prior interpolation (DPI), trains a CNN to map random noise to a complete seismic image using only the decimated image itself. The second one, referred as standard deep learning method, trains a CNN to map a decimated seismic image into a complete one using a dataset of complete and artificially decimated images. Within this research, we systematically compare the performance of both methods for different quantities of regular and irregular missing traces using 4 datasets. We evaluate the results of both methods using 5 well-known metrics. We found that DPI method performs better than the standard method if the percentage of missing traces is low (10%) and otherwise if the level of decimation is high (50%).
In this paper, the influence of the material hardening behavior on plasticity-induced fatigue crack closure is investigated for strain-controlled loading and fully plastic, large-scale yielding conditions by means of the finite element method. The strain amplitude and the strain ratio are varied for given Ramberg–Osgood material properties representing materials with different hardening behavior. The results show a pronounced influence of the hardening behavior on crack closure, while no significant effect is found from the considered strain amplitude and strain ratio. The effect of the hardening behavior on the crack opening stress cannot be described by existing crack opening stress equations.
The desire to connect more and more devices and to make them more intelligent and more reliable, is driving the needs for the Internet of Things more than ever. Such IoT edge systems require sound security measures against cyber-attacks, since they are interconnected, spatially distributed, and operational for an extended period of time. One of the most important requirements for the security in many industrial IoT applications is the authentication of the devices. In this paper, we present a mutual authentication protocol based on Physical Unclonable Functions, where challenge-response pairs are used for both device and server authentication. Moreover, a session key can be derived by the protocol in order to secure the communication channel. We show that our protocol is secure against machine learning, replay, man-in-the-middle, cloning, and physical attacks. Moreover, it is shown that the protocol benefits from a smaller computational, communication, storage, and hardware overhead, compared to similar works.
This paper presents a method for supporting the application of Additive Tooling (AT)-based validation environments in integrated product development. Based on a case study, relevant process steps, activities and possible barriers in the realisation of an injection-moulded product are identified and analysed. The aim of the method is to support the target-oriented application of Additive Tooling to obtain physical prototypes at an early stage and to shorten validation cycles.
Linear acceleration is a key performance determinant and major training component of many sports. Although extensive research about lower limb kinetics and kinematics is available, consistent definitions of distinctive key body positions, the underlying mechanisms and their related movement strategies are lacking. The aim of this ‘Method and Theoretical Perspective’ article is to introduce a conceptual framework which classifies the sagittal plane ‘shin roll’ motion during accelerated sprinting. By emphasising the importance of the shin segment’s orientation in space, four distinctive key positions are presented (‘shin block’, ‘touchdown’, ‘heel lock’ and ‘propulsion pose’), which are linked by a progressive ‘shin roll’ motion during swing-stance transition. The shin’s downward tilt is driven by three different movement strategies (‘shin alignment’, ‘horizontal ankle rocker’ and ‘shin drop’). The tilt’s optimal amount and timing will contribute to a mechanically efficient acceleration via timely staggered proximal-to-distal power output. Empirical data obtained from athletes of different performance levels and sporting backgrounds are required to verify the feasibility of this concept. The framework presented here should facilitate future biomechanical analyses and may enable coaches and practitioners to develop specific training programs and feedback strategies to provide athletes with a more efficient acceleration technique.
In recent years, the topic of embedded machine learning has become very popular in AI research. With the help of various compression techniques such as pruning, quantization and others compression techniques, it became possible to run neural networks on embedded devices. These techniques have opened up a whole new application area for machine learning. They range from smart products such as voice assistants to smart sensors that are needed in robotics. Despite the achievements in embedded machine learning, efficient algorithms for training neural networks in constrained domains are still lacking. Training on embedded devices will open up further fields of applications. Efficient training algorithms would enable federated learning on embedded devices, in which the data remains where it was collected, or retraining of neural networks in different domains. In this paper, we summarize techniques that make training on embedded devices possible. We first describe the need and requirements for such algorithms. Then we examine existing techniques that address training in resource-constrained environments as well as techniques that are also suitable for training on embedded devices, such as incremental learning. At the end, we also discuss which problems and open questions still need to be solved in these areas.
Nowadays decarbonisation of the energy system is one of the main concerns for most governments. Renewable energy technologies, such as rooftop photovoltaic systems and home battery storage systems, are changing the energy system to be more decentralised. As a consequence, new ways of energy business models are emerging, e.g., peer-to-peer energy trading. This new concept provides an online marketplace where direct energy exchange can occur between its participants. The purpose of this study is to conduct a content analysis of the existing literature, ongoing research projects, and companies related to peer-to-peer energy trading. From this review, a summary of the most important aspects and journal papers is assessed, discussed, and classified. It was found that the different energy market types were named in various ways and a proposal for standard language for the several peer-to-peer market types and the different actors involved is suggested. Additionally, by grouping the most important attributes from peer-to-peer energy trading projects, an assessment of the entry barrier and scalability potential is performed by using a characterisation matrix.
The increase of the Internet of Things (IoT) calls for secure solutions for industrial applications. The security of IoT can be potentially improved by blockchain. However, blockchain technology suffers scalability issues which hinders integration with IoT. Solutions to blockchain’s scalability issues, such as minimizing the computational complexity of consensus algorithms or blockchain storage requirements, have received attention. However, to realize the full potential of blockchain in IoT, the inefficiencies of its inter-peer communication must also be addressed. For example, blockchain uses a flooding technique to share blocks, resulting in duplicates and inefficient bandwidth usage. Moreover, blockchain peers use a random neighbor selection (RNS) technique to decide on other peers with whom to exchange blockchain data. As a result, the peer-to-peer (P2P) topology formation limits the effective achievable throughput. This paper provides a survey on the state-of-the-art network structures and communication mechanisms used in blockchain and establishes the need for network-based optimization. Additionally, it discusses the blockchain architecture and its layers categorizes existing literature into the layers and provides a survey on the state-of-the-art optimization frameworks, analyzing their effectiveness and ability to scale. Finally, this paper presents recommendations for future work.
Voice user interfaces (VUIs) offer an intuitive, fast and convenient way for humans to interact with machines and computers. Yet, whether they’ll be truly successful and find widespread uptake in the near future depends on the user experience (UX) they offer. With this survey-based study (n = 108), we aim to identify the major annoyances German voice assistant users are facing in voice-driven human-computer interactions. The results of our questionnaire show that irritations appear in six categories: privacy issues, unwanted activation, comprehensibility, response quality, conversational design and voice characteristics. Our findings can help identify key areas of work to optimize voice user experience in order to achieve greater adaptation of the technology. In addition, they can provide valuable information for the further development and standardization of voice user experience (VUX) research.
Featherweight Generic Go (FGG) is a minimal core calculus modeling the essential features of the programming language Go. It includes support for overloaded methods, interface types, structural subtyping and generics. The most straightforward semantic description of the dynamic behavior of FGG programs is to resolve method calls based on runtime type information of the receiver.
This article shows a different approach by defining a type-directed translation from FGG to an untyped lambda-calculus. The translation of an FGG program provides evidence for the availability of methods as additional dictionary parameters, similar to the dictionary-passing approach known from Haskell type classes. Then, method calls can be resolved by a simple lookup of the method definition in the dictionary.
Every program in the image of the translation has the same dynamic semantics as its source FGG program. The proof of this result is based on a syntactic, step-indexed logical relation. The step-index ensures a well-founded definition of the relation in the presence of recursive interface types and recursive methods.
In dem Projekt BioMeth wurde der Ansatz der Membranbegasung zur Erhöhung der Verfügbarkeit von gelöstem Wasserstoff für die biologische Methanisierung im Sinn der Etablierung eines Power-to-Gas-Konzeptes zur Energiespeicherung verfolgt. Übergeordnetes Ziel war die Entwicklung eines skalierbaren Verfahrenskonzeptes, dass sich zur Nutzung CO2-haltiger Gasvolumenström eignet. Geplant war es, das Verfahren am Beispiel der Biogasanlage der Biokäserei Monte-Ziego in Teningen zu demonstrieren und dort das bestehende Konzept der parallelen Abwasseraufbereitung und Energieerzeugung zu erweitern. Die ursprüngliche Struktur des Arbeitspaketplanes ist in nachfolgender Abbildung gezeigt.
Subspace clustering aims to find all clusters in all subspaces of a high-dimensional data space. We present a massively data-parallel approach that can be run on graphics processing units. It extends a previous density-based method that scales well with the number of dimensions. Its main computational bottleneck consists of (sequentially) generating a large number of minimal cluster candidates in each dimension and using hash collisions in order to find matches of such candidates across multiple dimensions. Our approach parallelizes this process by removing previous interdependencies between consecutive steps in the sequential generation process and by applying a very efficient parallel hashing scheme optimized for GPUs. This massive parallelization gives up to 70x speedup for
the bottleneck computation when it is replaced by our approach and run on current GPU hardware. We note that depending on data size and choice of parameters, the parallelized part of the algorithm can take different percentages of the overall runtime of the clustering process, and thus, the overall clustering speedup may vary significantly between different cases. However, even
in our ”worst-case” test, a small dataset where the computation makes up only a small fraction of the overall clustering time, our parallel approach still yields a speedup of more than 3x for the complete run of the clustering process. Our method could also be combined with parallelization of other parts of the clustering algorithm, with an even higher potential gain in processing speed.
In pandemic times, the possibilities for conventional sports activities are severely limited; many sports facilities are closed or can only be used with restrictions. To counteract this lack of health activities and social exchange, people are increasingly adopting new digital sports solutions—a behavior change that had already started with the trend towards fitness apps and activity trackers. Existing research suggests that digital solutions increase the motivation to move and stay active. This work further investigates the potentials of digital sports incorporating the dimensions gender and preference for team sports versus individual sports. The study focuses on potential users, who were mostly younger professionals and academics. The results show that the SARS-CoV-19 pandemic had a significant negative impact on sports activity, particularly on persons preferring team sports. To compensate, most participants use more digital sports than before, and there is a positive correlation between the time spent physically active during the pandemic and the increase in motivation through digital sports. Nevertheless, there is still considerable skepticism regarding the potential of digital sports solutions to increase the motivation to do sports, increase performance, or raise a sense of team spirit when done in groups.
Deep learning models are intrinsically sensitive to distribution shifts in the input data. In particular, small, barely perceivable perturbations to the input data can force models to make wrong predictions with high confidence. An common defense mechanism is regularization through adversarial training which injects worst-case perturbations back into training to strengthen the decision boundaries, and to reduce overfitting. In this context, we perform an investigation of 3 × 3 convolution filters that form in adversarially- trained models. Filters are extracted from 71 public models of the ℓ ∞ -RobustBench CIFAR-10/100 and ImageNet1k leaderboard and compared to filters extracted from models built on the same architectures but trained without robust regularization. We observe that adversarially-robust models appear to form more diverse, less sparse, and more orthogonal convolution filters than their normal counterparts. The largest differences between robust and normal models are found in the deepest layers, and the very first convolution layer, which consistently and predominantly forms filters that can partially eliminate perturbations, irrespective of the architecture.
The significant market growth of stationary electrical energy storage systems both for private and commercial applications has raised the question of battery lifetime under practical operation conditions. Here, we present a study of two 8 kWh lithium-ion battery (LIB) systems, each equipped with 14 lithium iron phosphate/graphite (LFP) single cells in different cell configurations. One system was based on a standard configuration with cells connected in series, including a cell-balancing system and a 48 V inverter. The other system featured a novel configuration of two stacks with a parallel connection of seven cells each, no cell-balancing system, and a 4 V inverter. The two systems were operated as part of a microgrid both in continuous cycling mode between 30% and 100% state of charge, and in solar-storage mode with day–night cycling. The aging characteristics in terms of capacity loss and internal resistance change in the cells were determined by disassembling the systems for regular checkups and characterizing the individual cells under well-defined laboratory conditions. As a main result, the two systems showed cell-averaged capacity losses of 18.6% and 21.4% for the serial and parallel configurations, respectively, after 2.5 years of operation with 810 (serial operation) and 881 (parallel operation) cumulated equivalent full cycles. This is significantly higher than the aging of a reference single cell cycled under laboratory conditions at 20 °C, which showed a capacity loss of only 10% after 1000 continuous full cycles.
Many commonly well-performing convolutional neural network models have shown to be susceptible to input data perturbations, indicating a low model robustness. To reveal model weaknesses, adversarial attacks are specifically optimized to generate small, barely perceivable image perturbations that flip the model prediction. Robustness against attacks can be gained by using adversarial examples during training, which in most cases reduces the measurable model attackability. Unfortunately, this technique can lead to robust overfitting, which results in non-robust models. In this paper, we analyze adversarially trained, robust models in the context of a specific network operation, the downsampling layer, and provide evidence that robust models have learned to downsample more accurately and suffer significantly less from downsampling artifacts, aka. aliasing, than baseline models. In the case of robust overfitting, we observe a strong increase in aliasing and propose a novel early stopping approach based on the measurement of aliasing.
Many commonly well-performing convolutional neural network models have shown to be susceptible to input data perturbations, indicating a low model robustness. Adversarial attacks are thereby specifically optimized to reveal model weaknesses, by generating small, barely perceivable image perturbations that flip the model prediction. Robustness against attacks can be gained for example by using adversarial examples during training, which effectively reduces the measurable model attackability. In contrast, research on analyzing the source of a model’s vulnerability is scarce. In this paper, we analyze adversarially trained, robust models in the context of a specifically suspicious network operation, the downsampling layer, and provide evidence that robust models have learned to downsample more accurately and suffer significantly less from aliasing than baseline models.
In the literature, many studies have described the 3D printing of ceramic-based scaffolds (e.g., printing with calcium phosphate cement) in the form of linear structures with layer rotations of 90°, although no right angles can be found in the human body. Therefore, this work focuses on the adaptation of biological shapes, including a layer rotation of only 1°. Sample shapes were printed with calcium phosphate cement using a 3D Bioplotter from EnvisionTec. Both straight and wavy spokes were printed in a round structure with 12 layers. Depending on the strand diameter (200 and 250 µm needle inner diameter) and strand arrangement, maximum failure loads of 444.86 ± 169.39 N for samples without subsequent setting in PBS up to 1280.88 ± 538.66 N after setting in PBS could be achieved.
Running footwear is continuously being modified and improved; however, running-related overuse injury rates remain high. Nevertheless, novel manufacturing processes enable the production of individualized running shoes that can fit the individual needs of runners, with the potential to reduce injury risk. For this reason, it is essential to investigate functional groups of runners, a collective of runners who respond similarly to a footwear intervention. Therefore, the objective of this study was to develop a framework to identify functional groups based on their individual footwear response regarding injury-specific running-related risk factors for Achilles tendinopathy, Tibial stress fractures, Medial tibial stress syndrome, and Patellofemoral pain syndrome. In this work, we quantified the footwear response patterns of 73 female and male participants when running in three different footwear conditions using unsupervised learning (k-means clustering). For each functional group, we identified the footwear conditions minimizing the injury-specific risk factors. We described differences in the functional groups regarding their running style, anthropometric, footwear perception, and demographics. The results implied that most functional groups showed a tendency for a single footwear condition to reduce most biomechanical risk factors for a specific overuse injury. Functional groups often differed in their hip and pelvis kinematics as well as their subjective rating of the footwear conditions. The footwear intervention only partially affected biomechanical risk factors attributed to more proximal joints. Due to its adaptive nature, the framework could be applied to other footwear interventions or performance-related biomechanical variables.