arXiv subjects

Niklas Kühl

Publications and source records attributed to Niklas Kühl.

At least 91 records · Page 5Linked to original sources

arXiv2022

Factors that influence the adoption of human-AI collaboration in clinical decision-making

Recent developments in Artificial Intelligence (AI) have fueled the emergence of human-AI collaboration, a setting where AI is a coequal partner. Especially in clinical decision-making, it has the potential to improve treatment quality by assisting overworked medical professionals. Even though research has started to investigate the utilization of AI for clinical decision-making, its potential benefits do not imply its adoption by medical professionals. While several studies have started to analyze adoption criteria from a technical perspective, research providing a human-centered perspective with a focus on AI's potential for becoming a coequal team member in the decision-making process remains limited. Therefore, in this work, we identify factors for the adoption of human-AI collaboration by conducting a series of semi-structured interviews with experts in the healthcare domain. We identify six relevant adoption factors and highlight existing tensions between them and effective human-AI collaboration.

cs.HC↗

arXiv2022

Should I Follow AI-based Advice? Measuring Appropriate Reliance in Human-AI Decision-Making

Many important decisions in daily life are made with the help of advisors, e.g., decisions about medical treatments or financial investments. Whereas in the past, advice has often been received from human experts, friends, or family, advisors based on artificial intelligence (AI) have become more and more present nowadays. Typically, the advice generated by AI is judged by a human and either deemed reliable or rejected. However, recent work has shown that AI advice is not always beneficial, as humans have shown to be unable to ignore incorrect AI advice, essentially representing an over-reliance on AI. Therefore, the aspired goal should be to enable humans not to rely on AI advice blindly but rather to distinguish its quality and act upon it to make better decisions. Specifically, that means that humans should rely on the AI in the presence of correct advice and self-rely when confronted with incorrect advice, i.e., establish appropriate reliance (AR) on AI advice on a case-by-case basis. Current research lacks a metric for AR. This prevents a rigorous evaluation of factors impacting AR and hinders further development of human-AI decision-making. Therefore, based on the literature, we derive a measurement concept of AR. We propose to view AR as a two-dimensional construct that measures the ability to discriminate advice quality and behave accordingly. In this article, we derive the measurement concept, illustrate its application and outline potential future research.

cs.HC↗

arXiv2022

Discrete Adjoint Momentum-Weighted Interpolation Strategies

This Technical Note outlines an adjoint complement to a critical building block of pressure-based Finite-Volume (FV) flow solvers that employ a collocated variable arrangement to simulate virtually incompressible fluids. The focal point is to strengthen the adjoint pressure-velocity coupling by using an adjoint Momentum-Weighted Interpolation (MWI) strategy. To this end, analogies of established primal MWI techniques are derived and investigated. The study reveals the merits of an adjoint MWI but also highlights the importance of its careful implementation.

math.NA↗

arXiv2022

Adjoint Node-Based Shape Optimization of Free Floating Vessels

The paper is concerned with a node-based, gradient-driven, continuous adjoint two-phase flow procedure to optimize the shapes of free-floating vessels and discusses three topics. First, we aim to convey that elements of a Cahn-Hilliard formulation should augment the frequently employed Volume-of-Fluid two-phase flow model to maintain dual consistency. It is seen that such consistency serves as the basis for a robust primal/adjoint coupling in practical applications at huge Reynolds and Froude numbers. The second topic covers different adjoint coupling strategies. A central aspect of the application is the floating position, particularly the trim and the sinkage, that interact with a variation of hydrodynamic loads induced by the shape updates. Other topics addressed refer to the required level of density coupling and a more straightforward -- yet non-frozen -- adjoint treatment of turbulence. The third part discusses the computation of a descent direction within a node-based environment. We will illustrate means to deform both the volume mesh and the hull shape simultaneously and at the same time obey technical constraints on the vessel's displacement and its extensions. The Hilbert-space approach provides smooth shape updates using the established coding infrastructure of a computational fluid dynamics algorithm and provides access to managing additional technical constraints. Verification and validation follow from a submerged 2D cylinder case. The application includes a full-scale offshore supply vessel at Re=3E+08 and Fn=0.37. Results illustrate that the fully parallel procedure can automatically reduce the drag of an already pre-optimized shape by 9-13 within approximately O(10.000-30.000) CPUh depending on the considered couplings and floatation aspects.

physics.flu-dyn↗

arXiv2021

Utilizing Active Machine Learning for Quality Assurance: A Case Study of Virtual Car Renderings in the Automotive Industry

Computer-generated imagery of car models has become an indispensable part of car manufacturers' advertisement concepts. They are for instance used in car configurators to offer customers the possibility to configure their car online according to their personal preferences. However, human-led quality assurance faces the challenge to keep up with high-volume visual inspections due to the car models' increasing complexity. Even though the application of machine learning to many visual inspection tasks has demonstrated great success, its need for large labeled data sets remains a central barrier to using such systems in practice. In this paper, we propose an active machine learning-based quality assurance system that requires significantly fewer labeled instances to identify defective virtual car renderings without compromising performance. By employing our system at a German automotive manufacturer, start-up difficulties can be overcome, the inspection process efficiency can be increased, and thus economic advantages can be realized.

cs.LG↗

arXiv2021

What to Prioritize? Natural Language Processing for the Development of a Modern Bug Tracking Solution in Hardware Development

Managing large numbers of incoming bug reports and finding the most critical issues in hardware development is time consuming, but crucial in order to reduce development costs. In this paper, we present an approach to predict the time to fix, the risk and the complexity of debugging and resolution of a bug report using different supervised machine learning algorithms, namely Random Forest, Naive Bayes, SVM, MLP and XGBoost. Further, we investigate the effect of the application of active learning and we evaluate the impact of different text representation techniques, namely TF-IDF, Word2Vec, Universal Sentence Encoder and XLNet on the model's performance. The evaluation shows that a combination of text embeddings generated through the Universal Sentence Encoder and MLP as classifier outperforms all other methods, and is well suited to predict the risk and complexity of bug tickets.

cs.LG↗

arXiv2021

Intelligent Decision Assistance Versus Automated Decision-Making: Enhancing Knowledge Work Through Explainable Artificial Intelligence

While recent advances in AI-based automated decision-making have shown many benefits for businesses and society, they also come at a cost. It has for long been known that a high level of automation of decisions can lead to various drawbacks, such as automation bias and deskilling. In particular, the deskilling of knowledge workers is a major issue, as they are the same people who should also train, challenge and evolve AI. To address this issue, we conceptualize a new class of DSS, namely Intelligent Decision Assistance (IDA) based on a literature review of two different research streams -- DSS and automation. IDA supports knowledge workers without influencing them through automated decision-making. Specifically, we propose to use techniques of Explainable AI (XAI) while withholding concrete AI recommendations. To test this conceptualization, we develop hypotheses on the impacts of IDA and provide first evidence for their validity based on empirical studies in the literature.

cs.HC↗

arXiv2021

Deep Learning Strategies for Industrial Surface Defect Detection Systems

Deep learning methods have proven to outperform traditional computer vision methods in various areas of image processing. However, the application of deep learning in industrial surface defect detection systems is challenging due to the insufficient amount of training data, the expensive data generation process, the small size, and the rare occurrence of surface defects. From literature and a polymer products manufacturing use case, we identify design requirements which reflect the aforementioned challenges. Addressing these, we conceptualize design principles and features informed by deep learning research. Finally, we instantiate and evaluate the gained design knowledge in the form of actionable guidelines and strategies based on an industrial surface defect detection use case. This article, therefore, contributes to academia as well as practice by (1) systematically identifying challenges for the industrial application of deep learning-based surface defect detection, (2) strategies to overcome these, and (3) an experimental case study assessing the strategies' applicability and usefulness.

cs.CV↗

arXiv2021

Fake or Credible? Towards Designing Services to Support Users' Credibility Assessment of News Content

Fake news has become omnipresent in digitalized areas such as social media platforms. While being disseminated online, it also poses a threat to individuals and societies offline, for example, in the context of democratic elections. Research and practice have investigated the detection of fake news with behavioral science or method-related perspectives. However, to date, we lack design knowledge on presenting fake news warnings to users to support their individual news credibility assessment. We present the journey through the first design cycle on developing a fake news detection service focusing on the user interface design. The design is grounded in concepts from the field of source credibility theory and instantiated in a prototype that was qualitatively evaluated. The 13 participants communicated their interest in a lightweight application that aids in the news credibility assessment and rated the design features as useful as well as desirable.

cs.HC↗

arXiv2021

Towards a Reference Architecture for Future Industrial Internet of Things Networks

With the continuing decrease of sensor technology prices as well as the increase of communication and analytical capabilities of modern internet of things devices, the continuously generated amount of data is constantly growing. Various use cases show the untapped potential of this data for new business models. However, conventional industrial IT networks of traditional manufacturing companies can hardly meet the modern requirements emerging with today's and future industrial internet of things applications. Outdated and rigid network infrastructures are one of the main reasons for hesitant innovation efforts and cross-organizational collaborations as well as the slow adoption of modern business models by traditional manufacturing companies. Following the design science research paradigm, our work contributes by elaborating on a comprehensive list of requirements for future industrial internet of things networks from a theoretical and practical perspective as well as a proposed reference architecture acting as a blueprint for future implementations.

cs.NI↗

arXiv2021

Utilizing Concept Drift for Measuring the Effectiveness of Policy Interventions: The Case of the COVID-19 Pandemic

As a reaction to the high infectiousness and lethality of the COVID-19 virus, countries around the world have adopted drastic policy measures to contain the pandemic. However, it remains unclear which effect these measures, so-called non-pharmaceutical interventions (NPIs), have on the spread of the virus. In this article, we use machine learning and apply drift detection methods in a novel way to predict the time lag of policy interventions with respect to the development of daily case numbers of COVID-19 across 9 European countries and 28 US states. Our analysis shows that there are, on average, more than two weeks between NPI enactment and a drift in the case numbers.

cs.CY↗

arXiv2021

A Picture is Worth a Collaboration: Accumulating Design Knowledge for Computer-Vision-based Hybrid Intelligence Systems

Computer vision (CV) techniques try to mimic human capabilities of visual perception to support labor-intensive and time-consuming tasks like the recognition and localization of critical objects. Nowadays, CV increasingly relies on artificial intelligence (AI) to automatically extract useful information from images that can be utilized for decision support and business process automation. However, the focus of extant research is often exclusively on technical aspects when designing AI-based CV systems while neglecting socio-technical facets, such as trust, control, and autonomy. For this purpose, we consider the design of such systems from a hybrid intelligence (HI) perspective and aim to derive prescriptive design knowledge for CV-based HI systems. We apply a reflective, practice-inspired design science approach and accumulate design knowledge from six comprehensive CV projects. As a result, we identify four design-related mechanisms (i.e., automation, signaling, modification, and collaboration) that inform our derived meta-requirements and design principles. This can serve as a basis for further socio-technical research on CV-based HI systems.

cs.CV↗

arXiv2021

A Novel $p$-Harmonic Descent Approach Applied to Fluid Dynamic Shape Optimization

We introduce a novel method for the implementation of shape optimziation in fluid dynamics applications, where we propose to use the shape derivative to determine deformation fields with the help of the $p-$ Laplacian for $p > 2$. This approach is closely related to the computation of steepest descent directions of the shape functional in the $W^{1,\infty}-$ topology. Our approach is demonstrated for shape optimization related to drag-minimal free floating bodies. The method is validated against existing approaches with respect to convergence of the optimization algorithm, the obtained shape, and regarding the quality of the computational grid after large deformations. Our numerical results strongly indicate that shape optimization related to the $W^{1,\infty}$-topology -- though numerically more demanding -- seems to be superior over the classical approaches invoking Hilbert space methods, concerning the convergence, the obtained shapes and the mesh quality after large deformations, in particular when the optimal shape features sharp corners.

math.OC↗

arXiv2021

labelCloud: A Lightweight Domain-Independent Labeling Tool for 3D Object Detection in Point Clouds

Within the past decade, the rise of applications based on artificial intelligence (AI) in general and machine learning (ML) in specific has led to many significant contributions within different domains. The applications range from robotics over medical diagnoses up to autonomous driving. However, nearly all applications rely on trained data. In case this data consists of 3D images, it is of utmost importance that the labeling is as accurate as possible to ensure high-quality outcomes of the ML models. Labeling in the 3D space is mostly manual work performed by expert workers, where they draw 3D bounding boxes around target objects the ML model should later automatically identify, e.g., pedestrians for autonomous driving or cancer cells within radiography. While a small range of recent 3D labeling tools exist, they all share three major shortcomings: (i) they are specified for autonomous driving applications, (ii) they lack convenience and comfort functions, and (iii) they have high dependencies and little flexibility in data format. Therefore, we propose a novel labeling tool for 3D object detection in point clouds to address these shortcomings.

cs.CV↗

arXiv2021

Adjoint-based Shape Optimization for the Minimization of Flow-induced Hemolysis in Biomedical Applications

This paper reports on the derivation and implementation of a shape optimization procedure for the minimization of hemolysis induction in biomedical devices. Hemolysis is a blood damaging phenomenon that may occur in mechanical blood-processing applications where large velocity gradients are found. An increased level of damaged blood can lead to deterioration of the immune system and quality of life. It is, thus, important to minimize flow-induced hemolysis by improving the design of next-generation biomedical machinery. Emphasis is given to the formulation of a continuous adjoint complement to a power-law hemolysis prediction model dedicated to efficiently identifying the shape sensitivity to hemolysis. The computational approach is verified against the analytical solutions of a benchmark problem and computed sensitivity derivatives are validated by a finite differences study on a generic 2D stenosed geometry. The application included addresses a 3D ducted geometry which features typical characteristics of biomedical devices. An optimized shape, leading to a potential improvement in hemolysis induction up to 22%, is identified. It is shown, that the improvement persists for different, literature-reported hemolysis-evaluation parameters.

physics.flu-dyn↗

arXiv2021

Needmining: Designing Digital Support to Elicit Needs from Social Media

Today's businesses face a high pressure to innovate in order to succeed in highly competitive markets. Successful innovations, though, typically require the identification and analysis of customer needs. While traditional, established need elicitation methods are time-proven and have demonstrated their capabilities to deliver valuable insights, they lack automation and scalability and, thus, are expensive and time-consuming. In this article, we propose an approach to automatically identify and quantify customer needs by utilizing a novel data source: Users voluntarily and publicly expose information about themselves via social media, as for instance Facebook or Twitter. These posts may contain valuable information about the needs, wants, and demands of their authors. We apply a Design Science Research (DSR) methodology to add design knowledge and artifacts for the digitalization of innovation processes, in particular to provide digital support for the elicitation of customer needs. We want to investigate whether automated, speedy, and scalable need elicitation from social media is feasible. We concentrate on Twitter as a data source and on e-mobility as an application domain. In a first design cycle we conceive, implement and evaluate a method to demonstrate the feasibility of identifying those social media posts that actually express customer needs. In a second cycle, we build on this artifact to additionally quantify the need information elicited, and prove its feasibility. Third, we integrate both developed methods into an end-user software artifact and test usability in an industrial use case. Thus, we add new methods for need elicitation to the body of knowledge, and introduce concrete tooling for innovation management in practice.

cs.CY↗

arXiv2020

Adjoint Complement to the Universal Momentum Law of the Wall

The paper is devoted to an adjoint complement to the universal Law of the Wall (LoW) for fluid dynamic momentum boundary layers. The latter typically follows from a strongly simplified, unidirectional shear flow under a constant stress assumption. We first derive the adjoint companion of the simplified momentum equation, while distinguishing between two strategies. Using mixing-length arguments, we demonstrate that the frozen turbulence strategy and a LoW-consistent (differentiated) approach provide virtually the same adjoint momentum equations, that differ only in a single scalar coefficient controlling the inclination in the logarithmic region. Moreover, it is seen that an adjoint LoW can be derived which resembles its primal counterpart in many aspects. The strategy is also compatible with wall-function assumptions for prominent RANS-type two-equation turbulence models, which ground on the mixing-length hypothesis. As a direct consequence of the frequently employed assumption that all primal flow properties algebraically scale with the friction velocity, it is demonstrated that a simple algebraic expression provides a consistent closure of the adjoint momentum equation in the logarithmic layer. This algebraic adjoint closure might also serve as an approximation for more general adjoint flow optimization studies using standard one- or two-equation Boussinesq-viscosity models for the primal flow. Results obtained from the suggested algebraic closure are verified against the primal/adjoint LoW formulations for both, low- and high-Re settings. Applications included in this paper refer to two- and three-dimensional shape optimizations of internal and external engineering flows. Related results indicate that the proposed adjoint algebraic turbulence closure accelerates the optimization process and provides improved optima in comparison to the frozen turbulence approach.

physics.flu-dyn↗

arXiv2020

Human vs. supervised machine learning: Who learns patterns faster?

The capabilities of supervised machine learning (SML), especially compared to human abilities, are being discussed in scientific research and in the usage of SML. This study provides an answer to how learning performance differs between humans and machines when there is limited training data. We have designed an experiment in which 44 humans and three different machine learning algorithms identify patterns in labeled training data and have to label instances according to the patterns they find. The results show a high dependency between performance and the underlying patterns of the task. Whereas humans perform relatively similarly across all patterns, machines show large performance differences for the various patterns in our experiment. After seeing 20 instances in the experiment, human performance does not improve anymore, which we relate to theories of cognitive overload. Machines learn slower but can reach the same level or may even outperform humans in 2 of the 4 of used patterns. However, machines need more instances compared to humans for the same results. The performance of machines is comparably lower for the other 2 patterns due to the difficulty of combining input features.

cs.AI↗