Refine
H-BRS Bibliography
- yes (75) (remove)
Departments, institutes and facilities
- Fachbereich Informatik (37)
- Fachbereich Ingenieurwissenschaften und Kommunikation (19)
- Institut für Technik, Ressourcenschonung und Energieeffizienz (TREE) (19)
- Fachbereich Angewandte Naturwissenschaften (10)
- Institute of Visual Computing (IVC) (10)
- Fachbereich Wirtschaftswissenschaften (9)
- Institut für Verbraucherinformatik (IVI) (8)
- Internationales Zentrum für Nachhaltige Entwicklung (IZNE) (5)
- Institut für Sicherheitsforschung (ISF) (4)
- Fachbereich Sozialpolitik und Soziale Sicherung (3)
Document Type
- Conference Object (75) (remove)
Year of publication
- 2019 (75) (remove)
Keywords
- FPGA (2)
- Hyperspectral image (2)
- Raman microscopy (2)
- evaluation (2)
- image fusion (2)
- learning outcomes (2)
- modeling of complex systems (2)
- pansharpening (2)
- 802.11 (1)
- Augmented Reality (1)
This work presents the preliminary research towards developing an adaptive tool for fault detection and diagnosis of distributed robotic systems, using explainable machine learning methods. Autonomous robots are complex systems that require high reliability in order to operate in different environments. Even more so, when considering distributed robotic systems, the task of fault detection and diagnosis becomes exponentially difficult.
To diagnose systems, models representing the behaviour under investigation need to be developed, and with distributed robotic systems generating large amount of data, machine learning becomes an attractive method of modelling especially because of its high performance. However, with current day methods such as artificial neural networks (ANNs), the issue of explainability arises where learnt models lack the ability to give explainable reasons behind their decisions.
This paper presents current trends in methods for data collection from distributed systems, inductive logic programming (ILP); an explainable machine learning method, and fault detection and diagnosis.
Die Globalisierung führt zu immer komplexeren, für die Einzelnen kaum nachvollziehbaren Wertschöpfungsketten in der Lebensmittelindustrie. Zugleich eröffnet die Digitalisierung neue Möglichkeiten, Informationen entlang der Kette zu sammeln, und so mehr Transparenz und Vertrauen für den Verbraucherbeziehungsweise die Verbraucherin zu schaffen. Jedoch finden Verbraucherinformations-Apps wie fTRACE bisher nur eine geringe Verbreitung. Daher haben wir in einer qualitativen Studie mit 16 Teilnehmer/-innen Bedürfnisse und Nutzungshürden von Verbraucher/-innen im Zusammenhang mit Verbraucherinformations-Apps analysiert. Es zeigt sich, dass das Vertrauen in die Informationen, sowie der einfache Zugang dazu für Verbraucher/-innen zentral sind. Durch die gut sichtbare Bereitstellung der Informationen am Point-of-Sale, sowie der automatisierten Informationsversorgung z. B. mittels digitaler Kassenzettel in Kombination mit weiteren Verbraucher-Services kann die Bekanntheit und Akzeptanz von Rückverfolgbarkeitssystemen weiter gesteigert werden.
Stakeholder-Analyse zum Einsatz IIoT-basierter Frischeinformationen in der Lebensmittelindustrie
(2019)
Eine Herausforderung bei der Implementierung des industriellen Internet of Things (IIoT) besteht darin, Mehrwerte in Wertschöpfungsketten zu identifizieren, um darauf aufbauend Lösungen nutzerzentriert zu gestalten. Dieser Beitrag stellt das Forschungsprojekt FreshIndex vor, bei dem diese Herausforderung durch eine Kombination aus Stakeholder-Analyse und User-Centered-Design-Methoden adressiert wurde. Ziel des Projekts ist es, eine IIoT-basierte Lösung zum Monitoring der Kühlkette in der Lebensmittelindustrie zu entwickeln. Hierzu ist es wichtig zu wissen, welche Nutzer/-innen mit den Daten in Berührung kommen und welche Erfahrungen, Fähigkeiten, Anforderungen und Wünsche sie mitbringen. Die Berücksichtigung dieser Aspekte ist relevant für den Erfolg der Konzeption, Implementierung und des Betriebs eines IIoT-Systems. So können nützliche und handhabbare Produktideen generiert und Anwendungen gestaltet werden, die von Mitarbeiter/-innen und Konsument/-innen angenommen werden. IIoT schließt somit die lokale Verwendbarkeit von Daten entlang der Wertschöpfungskette ein und beschränkt sich nicht auf zentrale Verfügbarkeit von Daten.
The design of self-driving cars is one of the most exciting and ambitious challenges of our days and every day, new research work is published. In order to give an orientation, this article will present an overview of various methods used to study the human side of autonomous driving. Simplifying roughly, you can distinguish between design science-oriented methods (such as Research through Design, Wizard of Oz or driving simulator ) and behavioral science methods (such as survey, interview, and observation). We show how these methods are adopted in the context of autonomous driving research and dis-cuss their strengths and weaknesses. Due to the complexity of the topic, we will show that mixed method approaches will be suitable to explore the impact of autonomous driving on different levels: the individual, the social interaction and society.
The alternative use of travel time is one of the widely discussed benefits of driverless cars. We therefore conducted 14 co-design sessions to examine how people manage their time, to determine how they perceive the value of time in driverless cars and to derive design implications. Our findings suggest that driverless mobility will affect both people’s use of travel time as well as their time management in general. The participants repeatedly stated the desire of completing tasks while traveling to save time for activities that are normally neglected in their everyday life. Using travel time efficiently requires using car space efficiently, too. We found out that the design concept of tiny houses could serve as common design pattern to deal with the limited space within cars and support diverse needs.
Energy Profiles of the Ring Puckering of Cyclopentane, Methylcyclopentane and Ethylcyclopentane
(2019)
Application developers constitute an important part of a digital platform’s ecosystem. Knowledge about psychological processes that drive developer behavior in platform ecosystems is scarce. We build on the lead userness construct which comprises two dimensions, trend leadership and high expected benefits from a solution, to explain how developers’ innovative work behavior (IWB) is stimulated. We employ an efficiencyoriented and a social-political perspective to investigate the relationship between lead userness and IWB. The efficiency-oriented view resonates well with the expected benefit dimension of lead userness, while the social-political view might be interpreted as a reflection of trend leadership. Using structural equation modeling, we test our model with a sample of over 400 developers from three platform ecosystems. We find that lead userness is indirectly associated with IWB and the performance-enhancing view to be the stronger predictor of IWB. Finally, we unravel differences between paid and unpaid app developers in platform ecosystems.
Modern Monte-Carlo-based rendering systems still suffer from the computational complexity involved in the generation of noise-free images, making it challenging to synthesize interactive previews. We present a framework suited for rendering such previews ofstatic scenes using a caching technique that builds upon a linkless octree. Our approach allows for memory-efficient storage and constant-time lookup to cache diffuse illumination at multiple hitpoints along the traced paths. Non-diffuse surfaces are dealt with in a hybrid way in order to reconstruct view-dependent illumination while maintaining interactive frame rates. By evaluating the visual fidelity against ground truth sequences and by benchmarking, we show that our approach compares well to low-noise path traced results, but with a greatly reduced computational complexity allowing for interactive frame rates. This way, our caching technique provides a useful tool for global illumination previews and multi-view rendering.
Tell Your Robot What To Do: Evaluation of Natural Language Models for Robot Command Processing
(2019)
The use of natural language to indicate robot tasks is a convenient way to command robots. As a result, several models and approaches capable of understanding robot commands have been developed, which however complicates the choice of a suitable model for a given scenario. In this work, we present a comparative analysis and benchmarking of four natural language understanding models - Mbot, Rasa, LU4R, and ECG. We particularly evaluate the performance of the models to understand domestic service robot commands by recognizing the actions and any complementary information in them in three use cases: the RoboCup@Home General Purpose Service Robot (GPSR) category 1 contest, GPSR category 2, and hospital logistics in the context of the ROPOD project.
Digital transformation in Higher Education and Science is a mission-critical demand to prepare educational institutions for their future competition on the international market. In many cases, the digitization goes along with the search for and acquisition of new software. For easily exchangeable software, wrong product decisions, in the worst case, lead to calculable financial losses. However, if a planned software requires a lot of technological adjustments and is to be applied as central component of a business- and/or security-critical environment, wrong decisions during the software acquisition process might lead to hardly calculable damage. Questions arising are how to decide for a product and how many resources should be invested for the acquisition process.
We planned to apply a commercial Business Support System, which should replace the currently used in-house developed software. Our goals were the increase of our university’s level of data security, to ease the interaction between stakeholders, to eliminate media discontinuities, to improve the process management and transparency, and to reduce the execution time of automated processes. Alongside with the introduction of the electronic case file, our agenda stipulates the digitization (and automation) of administrative university processes, especially, but not limited to, the student self-service and the administrative student life cycle. Usual tools and practices, commonly applied to (simple) software acquisition, failed in our scenario.
With the case study introduced in this paper, we address all persons, involved within software acquisition processes: From our experiences, we strongly recommend to place greater value on an exhaustively completed acquisition process, than on short-termed economic advantages.
The Learning Culture Survey (LCS) is a questionnaire-based research, investigating students’ perceptions of and expectations towards Higher Education (HE). The aim of this survey is to improve our understanding about the sources of cultural conflicts in educational scenarios. This understanding, shell help us to predict potential conflict situations and develop supportive measures.
After three years of development, the LCS was initialized in 2010 in South Korea and Germany. During the following years, the investigations were extended to further countries. The results, on the one hand, provided insights about the cultural context of HE in general and on the other hand, about specific (national / regional) characteristics of learners in HE. Most issues targeted with the questionnaire were directly linked to value systems. Thus, we expected from the beginning that the collected data would keep valid over longer periods of time. However, we had no evidence regarding the actual persistence of learning culture. For a study, designed to being implemented on a global scope and providing input for further applications, persistence is a basic condition to justify related investigations.
To answer the question on persistence, we repeated the LCS in our university every four years, between 2010 to 2018/19. Besides a small number of slight changes, explainable out of their situational context, the overall results kept consistent over the investigated years. In this paper, after an introduction of the LCS’ concept, setting and its general results from the past years, we present the insights from our most recently finalized longitudinal study on learning culture.
Quantifying Interference in WiLD Networks using Topography Data and Realistic Antenna Patterns
(2019)
Avoiding possible interference is a key aspect to maximize the performance in Wi-Fi based Long Distance networks. In this paper we quantify self-induced interference based on data derived from our testbed and match the findings against simulations. By enhancing current simulation models with two key elements we significantly reduce the deviation between testbed and simulation: the usage of detailed antenna patterns compared to the cone model and propagation modeling enhanced by license-free topography data. Based on the gathered data we discuss several possible optimization approaches such as physical separation of local radios, tuning the sensitivity of the transmitter and using centralized compared to distributed channel assignment algorithms. While our testbed is based on 5 GHz Wi-Fi, we briefly discuss the possible impact of our results to other frequency bands.
Data-Driven Robot Fault Detection and Diagnosis Using Generative Models: A Modified SFDD Algorithm
(2019)
This paper presents a modification of the data-driven sensor-based fault detection and diagnosis (SFDD) algorithm for online robot monitoring. Our version of the algorithm uses a collection of generative models, in particular restricted Boltzmann machines, each of which represents the distribution of sliding window correlations between a pair of correlated measurements. We use such models in a residual generation scheme, where high residuals generate conflict sets that are then used in a subsequent diagnosis step. As a proof of concept, the framework is evaluated on a mobile logistics robot for the problem of recognising disconnected wheels, such that the evaluation demonstrates the feasibility of the framework (on the faulty data set, the models obtained 88.6% precision and 75.6% recall rates), but also shows that the monitoring results are influenced by the choice of distribution model and the model parameters as a whole.
When developing robot functionalities, finite state machines are commonly used due to their straightforward semantics and simple implementation. State machines are also a natural implementation choice when designing robot experiments, as they generally lead to reproducible program execution. In practice, the implementation of state machines can lead to significant code repetition and may necessitate unnecessary code interaction when reparameterisation is required. In this paper, we present a small Python library that allows state machines to be specified, configured, and dynamically created using a minimal domain-specific language. We illustrate the use of the library in three different use cases - scenario definition in the context of the RoboCup@Home competition, experiment design in the context of the ROPOD project, as well as specification transfer between robots.
For robots acting - and failing - in everyday environments, a predictable behaviour representation is important so that it can be utilised for failure analysis, recovery, and subsequent improvement. Learning from demonstration combined with dynamic motion primitives is one commonly used technique for creating models that are easy to analyse and interpret; however, mobile manipulators complicate such models since they need the ability to synchronise arm and base motions for performing purposeful tasks. In this paper, we analyse dynamic motion primitives in the context of a mobile manipulator - a Toyota Human Support Robot (HSR)- and introduce a small extension of dynamic motion primitives that makes it possible to perform whole body motion with a mobile manipulator. We then present an extensive set of experiments in which our robot was grasping various everyday objects in a domestic environment, where a sequence of object detection, pose estimation, and manipulation was required for successfully completing the task. Our experiments demonstrate the feasibility of the proposed whole body motion framework for everyday object manipulation, but also illustrate the necessity for highly adaptive manipulation strategies that make better use of a robot's perceptual capabilities.
Opportunities for Sustainable Mobility: Re-thinking Eco-feedback from a Citizen's Perspective
(2019)
In developed nations, a growing emphasis is being placed on the promotion of sustainable behaviours amongst individuals, or ‘citizen-consumers’. In HCI, various eco-feedback tools have been designed as persuasive instruments, with a strong normative appeal geared to encouraging citizens to conduct a more sustainable mobility. However, many critiques have been formulated regarding this ‘paternalistic’ stance. In this paper, we switched the perspective from a designer’s to a citizen’s point of view and explored how people would use eco-feedback tools to support sustainable mobility in their city. In the study, we conducted 14 interviews with citizens who had used eco-feedback previously. The findings indicate new starting points that could inform future eco-feedback tools. These encompass: (1) better information regarding how sustainable mobility is measured and monitored; (2) respect for individual mobility situations and preferences; and (3) the scope for participation and the sharing of responsibility between citizens and municipal city services.
Background & Objective: Due to the policy goals for sustainable energy production, renewable energy plants such as photovoltaics are increasingly in use. The energy production from solar radiation depends strongly on atmospheric conditions. As the weather mostly changes, electrical power generation fluctuates, making technical planning and control of power grids to a complex problem. Due to used materials (semiconductors e.g. silicon, gallium arsenide, cadmium telluride) the photovoltaic cells are spectrally selective. It means that only radiation of certain wavelengths converts into electrical energy. A material property called spectral response characterizes a certain degree of conversion of solar radiation into the electric current for each wavelength of solar light.
Bisherige Versuche der HCI-Community die Lebensmittelverschwendung oder den CO2-Fußabdruck zu reduzieren, basierten meist auf Persuasive Design Ansätzen. Diese werden jedoch kritisiert, die Alltagswelten und Konsumpraktiken für eine Langzeitwirkung nur unzureichend zu berücksichtigen. Das Problem aufgreifend, untersucht dieser Beitrag die Rolle (digitaler) Medien im Übergang zu einer veganen Ernährungspraktik. Hierfür wurden semi-strukturierte Interviews mit 9 VeganerInnen geführt und vor dem Hintergrund der Praxistheorie analysiert. Die Ergebnisse deuten dabei auf eine intensive Nutzung (digitaler) Medien, insbesondere in der frühen Phase der Änderung der Konsumpraktik. Statt Gamification oder Persuasive Design, zeigt sich Mediennutzung in Form von Irritation, Informationsbereitstellung zur Ausbildung eines vegan-spezifischen Konsumwissens sowie als Vermittler zwischen Gleichgesinnten.
In the field of service robots, dealing with faults is crucial to promote user acceptance. In this context, this work focuses on some specific faults which arise from the interaction of a robot with its real world environment due to insufficient knowledge for action execution.
In our previous work [1], we have shown that such missing knowledge can be obtained through learning by experimentation. The combination of symbolic and geometric models allows us to represent action execution knowledge effectively. However we did not propose a suitable representation of the symbolic model.
In this work we investigate such symbolic representation and evaluate its learning capability. The experimental analysis is performed on four use cases using four different learning paradigms. As a result, the symbolic representation together with the most suitable learning paradigm are identified.
We present a novel, multilayer interaction approach that enables state transitions between spatially above-screen and 2D on-screen feedback layers. This approach supports the exploration of haptic features that are hard to simulate using rigid 2D screens. We accomplish this by adding a haptic layer above the screen that can be actuated and interacted with (pressed on) while the user interacts with on-screen content using pen input. The haptic layer provides variable firmness and contour feedback, while its membrane functionality affords additional tactile cues like texture feedback. Through two user studies, we look at how users can use the layer in haptic exploration tasks, showing that users can discriminate well between different firmness levels, and can perceive object contour characteristics. Demonstrated also through an art application, the results show the potential of multilayer feedback to extend on-screen feedback with additional widget, tool and surface properties, and for user guidance.
Due to the policy goals for sustainable energy production, renewable energy plants such as photovoltaics are increasingly in use. The energy production from solar radiation depends strongly on atmospheric conditions. As the weather mostly changes, electrical power generation fluctuates, making technical planning and control of power grids to a complex problem.
Surrogate models are used to reduce the burden of expensive-to-evaluate objective functions in optimization. By creating models which map genomes to objective values, these models can estimate the performance of unknown inputs, and so be used in place of expensive objective functions. Evolutionary techniques such as genetic programming or neuroevolution commonly alter the structure of the genome itself. A lack of consistency in the genotype is a fatal blow to data-driven modeling techniques: interpolation between points is impossible without a common input space. However, while the dimensionality of genotypes may differ across individuals, in many domains, such as controllers or classifiers, the dimensionality of the input and output remains constant. In this work we leverage this insight to embed differing neural networks into the same input space. To judge the difference between the behavior of two neural networks, we give them both the same input sequence, and examine the difference in output. This difference, the phenotypic distance, can then be used to situate these networks into a common input space, allowing us to produce surrogate models which can predict the performance of neural networks regardless of topology. In a robotic navigation task, we show that models trained using this phenotypic embedding perform as well or better as those trained on the weight values of a fixed topology neural network. We establish such phenotypic surrogate models as a promising and flexible approach which enables surrogate modeling even for representations that undergo structural changes.
The initial phase in real world engineering optimization and design is a process of discovery in which not all requirements can be made in advance, or are hard to formalize. Quality diversity algorithms, which produce a variety of high performing solutions, provide a unique chance to support engineers and designers in the search for what is possible and high performing. In this work we begin to answer the question how a user can interact with quality diversity and turn it into an interactive innovation aid. By modeling a user's selection it can be determined whether the optimization is drifting away from the user's preferences. The optimization is then constrained by adding a penalty to the objective function. We present an interactive quality diversity algorithm that can take into account the user's selection. The approach is evaluated in a new multimodal optimization benchmark that allows various optimization tasks to be performed. The user selection drift of the approach is compared to a state of the art alternative on both a planning and a neuroevolution control task, thereby showing its limits and possibilities.