Master Theses

Acevedo, Carmela

Distributed Digital Desk: Studying the Effects and Influences of Ubiquitous Instrumental Interaction upon Creative Sensemaking (2016)

Distributed User Interfaces (DUIs) place different parts of applications across different dimensions. Research has been concerned on developing frameworks to achieve DUIs. Ubiquitous Instrumental Interaction (UII) is one of the models that has been created to do so. It is based on Instrumental Interaction and it stipulates the creation of applications out of categorized components, documents and instruments. Other models have been extensively evaluated, however, UII still lacks in-depth studies from users’ perspective.

This thesis seeks to make a first step into filling this gap. The Distributed Digital Desk (DDD) is a system that enables the creation of applications with DUIs. It provides several instruments that can be combined with documents to perform creative sensemaking work. These components can be placed in any kind of device that can open web pages. Hence, users can combine mobile phones into their own distributed application.

An explorative within-subject study was designed to better understand the influences of UII upon creative sensemaking. Participants used three applications to perform creative sensemaking tasks. All applications contained a document, while the interface to create and edit content in this document changed. The first one offered a toolbar with settings, much like classical WIMP related interfaces. The other two were made up of instruments. In one, the instruments were all in the same surface as the document, while in the other one they were in individual physical devices. This study showed that participants changed their behavior when presented with applications based on UII. They switched from mainly focusing on the objects they were interacting with, to focusing on the actions they were performing.


Barreras Aragón, Elena

Measuring Cognitive Load using Eye-Tracking in visual search tasks (2017)

The evolution of technology into the era of ubiquitous computing, where computers’ actions are mostly based on information about the human state (cognitive, affective and motivational states) is a motivation to design interfaces which provide as much real-time information about the human state as possible. In this regard, it is relevant to study how cognitive load can be applied in the field of Human-Computer Interaction. Over the past few years, the rise of eye tracker manufacturers has produced a diverse range of analysis software and scientific tools to investigate how the human gaze can provide information about the level of cognition based on single gaze events such as fixations or pupillary response. There have been several studies in the fields of eye tracking and cognition states based on visual and auditory tasks; e-learning, visualization of interfaces, etc. But little has been investigated in relation to cognitive load and visual search tasks.

This master thesis deals with the topic of eye tracking as a tool to measure cognitive load in visual search tasks. The design and conduction of an experiment, the posterior analysis based on the combination of gaze events; and the results obtained, are essential to get a deeper understanding of the level of cognition presented in visual search tasks, in contrast to researches where only a relation between single gaze events and cognitive load is provided.



Baube, Tobias

Interactive Reading of Digital Documents in the Mobile Context (2013)

The increasing amount of available, digital information and correlated documents indicates their importance and an enrichment of their physical counterparts. Yet, there is an evolving offer of a manifold of digital and mobile devices, supporting users to access and manipulate digital information, allows the user accessing these information objects at any place and any time.

Still, users have to cope with limitations compared to the conventional paper-based complement’s interaction approaches. Goal of this work is thus the limitation of interaction barriers by blending the digital and physical interaction traits of either world, to offer well-known behavioral types and to make use advantages of both worlds.



Bieg, Joachim

Laserpointer and Eye Gaze Interaction - Design and Evaluation (2008)

Interacting with modern graphical user interfaces usually requires a pointing device like for instance the mouse. The pointing speed and accuracy that can be achieved with a pointing device has a considerable influence on the usability of these systems. In this thesis a comprehensive introduction is given to the methods for the evaluation of speed and precision of pointing devices based on the specifications of the international standard ISO 9241-9. Furthermore, several improvements of these methods are proposed and discussed, that should improve the comparability of future evaluations. Effects of practice are one particular issue in the assessment of pointing device performance. Due to time and cost restrictions, only little time is typically scheduled for training with each device in comparative pointing device evaluations. This has the disadvantage that alterations in performance due to practice remain unconsidered. In order to examine the effects of practice on pointing performance, a longitudinal experiment was conducted with four participants. For each participant several sessions were scheduled on five consecutive days. In each of these sessions participants carried out a simple tapping tests with a novel laserpointer interaction system that was developed by the Human-Computer Interaction Group of the University of Konstanz. The experiment was carried out at the Powerwall at the University of Konstanz, a large highresolution display 5.20 x 2.15 m. The results of the experiment showed that practice has a considerable effect on the pointing performance of the laserpointer. Performance increases were discernible until session four. This motivates a more careful consideration of practice effects in forthcoming comparative evaluations. Furthermore, a novel type of test based on a discrete tapping task is presented that was employed in the experiment. The playful character of the test should maintain participants motivation across several sessions. In the last section of the thesis, the prospects of gaze interaction are discussed by giving an overview of the core issues and existing systems. Furthermore, three novel gaze interaction techniques are presented. Finally, first experiences and observations on gaze interaction for large high-resolution displays with a mobile eye-tracker are reported.



Bogenrieder, Daniel 

Providing Automatic Instance Segmentation and Classification of Meal Images using Deep Neural Networks (2021)

With obesity, diabetes, and in general, with the trend of having a healthier lifestyle, tracking food intake becomes more relevant. With the advance in machine learning and, especially, the improvements in deep learning, it is possible to analyze the food intake with a single image of the meal. This could help to automatically track different kinds of meals and also the amount of food. Even though this fully automatic detection might not be feasible yet, it is possible to provide a list of possible foods to the user and guess the amount of food. Various works have already tried to find a solution for this food detection, but food detection in images seems to be a particularly challenging problem. Pictures of the same food type can vary a lot depending on the used ingredients, the image quality, or the way of serving the food. Even more demanding is the problem that the different food items can occlude each other, making it hard to give a correct volume prediction.

Although there are many works that already cover similar tasks, this work focuses on getting the detection ready to be used with European dishes, provide food image segmentation in addition to classification and using the power of a state of the art Neural Network architecture. Therefore, this work introduces a pipeline, that fulfils the steps necessary to get from unlabeled food images to a trained neural network model, that can predict the food classes and the location of the food parts in the image. These steps consist of the creation of ground truth values, cleaning up and validating these values, training a neural network and deploying the result onto a server. Moreover, this automatic detection was integrated in an existing food tracking app. To evaluate the usefulness of this automation, a usability study with fifteen participants was conducted. The automatic detection speeds up the tracking time of the participants and has the advantage, that the food locations in the images are tracked as well. Based on the results of this study, further research directions and additional improvements of the system are presented.


Borowski, Marcel

Codestrate Packages: Design and Evaluation of a Package-Based Development Environment (2018)

Inspired by instrumental interaction and the concept of shareable dynamic media, the document-centric model describes a future where computation is (i) embedded within documents, (ii) shareable, malleable, and distributable among other documents and users, and (iii) independent from applications. Codestrate Packages implements this model by providing a package-based system to create user-extensible software. Its documents are shareable and collaboratively editable. It builds on Codestrates, a web-based computational notebook platform following the literate computing approach.

While it is a promising platform, Codestrates’ or Codestrate Packages’ influence on users and collaboration is still unexplored. To understand how collaboration unfolds in such a system, how computational notebooks affect programming, and how malleability and extensibility influence the development of applications, Codestrate Packages was deployed for 13 weeks during an introductory course on application development. During the course, pairs of students solved weekly programming assignments.

Data from weekly questionnaires, three focus groups consisting of students and teaching assistants, and keystroke-level log data were analyzed to facilitate the understanding of the subtleties of collaborative pair-based programming with computational notebooks. The findings reveal that there are distinct collaboration patterns. The preferred collaboration pattern varied between pairs and even varied within pairs throughout the 13 weeks. Further, the findings show that the linear structure of computational notebooks is beneficial for novices, however, can be restrictive for more experienced users. The reprogrammable nature and extensibility of the platform proved to be a double-edged sword as they, on the one hand, give users great expressive freedom, but on the other hand, bear the risk of users accidentally breaking the system.

Recognizing these benefits and barriers can help to guide the design of future computational notebooks. 


Böhnemann, Tobias

HomePlanAR: Smart Dynamic Guides for Space Planning in Handheld Augmented Reality (2022)

Space planning is the activity of creating layouts for living spaces. It is a domain well-suited for augmented reality (AR) and some commercial applications like IKEA Place exist that implement this. However, they do not support the precise creation of layouts well and do not incorporate the pre-existing environment beyond detecting a ground level. Research into the support of space planning and 3D layout creation is sparse. This thesis introduces two techniques to assist space planning in AR: Smart dynamic guides, a technique for making alignments and measurements well-known from presentation software such as Microsoft PowerPoint, and a scaleable grid system. For this purpose, it presents a general overview of both foundations of 3D manipulation and current research into AR layout planning. It then documents the design and implementation of an application that uses both of these techniques and presents a study that was conducted as a usability test for the application. The data collected in this study was subjected to the method of thematic analysis. From the results of this study, it derives ideas that could improve the design of both techniques, such as adding transparency to avoid visual obstruction issues, and places them in the context of existing research. Finally, it gives an outlook on potential future research.


Budzinski, Jochen

Hybrid Brainsketching - Interaction Concepts for Sketching Activities Based on Digital Pen (2012)

This master thesis research considers the analysis, the design and the evaluation of a digital tool, which aims at offering support for collaborative brainsketching sessions. The domain, the tasks and the activities that relate to social creativity have been studied and two user-roles have been considered during the phases of the system-design, namely the creative workers and the facilitator. Guidelines and requirements were either established by means of an observational user study or by means of theoretical investigations. The inferred design-implications, fit to the embodied practice of brainsketching activities, are then implemented by means of a demonstrative prototype - the hyb-system (hybrid brainsketching) - which is usable in collaborative brainsketching scenarios. The software has been developed by means of the ZOIL framework and Anoto´s digital pen and paper technology. Subsequently, a case study was conducted with seven design students and with two professional moderators, in order to reveal the drawbacks and the strengths of the hyb-system. To conclude, the system offers support for divergent thinking (the duplication of and the work with design-artifacts by multiple persons at the same time) and for idea communication, as this system promotes the role of a co-moderator. The user study also points out the opportunities that emerge, when those meetings are augmented with computational-support. Finally, on the basis of the results of the user study, suggestions were made for leveraging hyb´s potential by facilitating clustering-activities of the groupís output.



Butscher, Simon

Development of a visualization and interaction concept for monitoring and diagnosis in the context of control rooms: An examination of different multi-focus views using the exemple of a traffic control center (2012)

The work at hand deals with the development and the evaluation of a holistic visualization and interaction concept in the context of control centers. Thereby holistic means that based on the new interaction paradigm of Blended Interaction the personal interaction, the social interaction and communication as well as the physical environment and work processes are taken into consideration. The focus of the visualization and interaction concept lies on the primary control room operator tasks, such as the surveillance of complex processes for example in power plants or traffic monitoring. The operator navigates within a graphic information space, keeps the necessary process variables under surveillance and if necessary reacts to error reports and warnings. The developed concept which is intended to support these primary tasks is based on a combination of a public horizontal orientated display and a private vertical orientated multi-touch display and will follow the Overview Detail design pattern. The key aspect of the concept is, however, the additional integration of a multi-focus view using a Focus Context technique. In this work, two versions of a multi-focus view were developed, prototypically implemented and evaluated within a predominantly qualitative expert assessment with volunteers from a variety of traffic control centers. In the evaluation, an additional control condition was offered to the experts. Within this additional control condition the process was visualized without the use of a Focus Context technique. All experts claim to prefer one of two developed multi-focus views to the control condition.



Demarmels, Mischa

Enhancing Collaborative Information–Seeking through Surface and Tangible Computing (2010)

This master thesis discusses the topic of social interaction and collaboration in digital information-seeking systems. As context of the thesis the “Blended Library”, a vision of the library of the future is introduced. Novel interaction paradigms like surface and tangible computing are adopted to approach the flaws of today’s information systems in supporting users during collaborative search activities. The main contribution of the thesis is the concept of the “Search Tokens”, a tangible user interface for co-located, collaborative information-seeking. After establishing the design rationales and the basic visualization and interaction principles behind the concept an evaluation is presented that explores whether users benefit from the Search Token concept during search activities in groups. The research presented in this thesis shows promising and successful solutions to the problem of co-located, collaborative information-seeking.



Dierdorf, Stefan

Requirements Specification and Interaction-Design of a Multi-Modal Diary Application for Field-Studies (2011)

As a result of a rapidly evolving information landscape, which is more than ever being considered as ubiquitous and mobile, traditional data gathering methods from human-computer interaction and other disciplines have at times reached the limits of their usefulness. Researchers of various domains, however, are interested in investigating situations, processes, behavior, emotions, or interactions from within their natural environments. Controlled experiments with their artificial emulation of reality cannot meet this demand. Likewise, observations turn out inadequate when one wants to investigate subjective experiences of people. Moreover, they can only be seen as a snap-shot since they do not cover time. Hence, we will look into diary studies as a research method in this thesis. Diaries can be implemented when other methods cannot provide the required level of insight: for the investigation of phenomena from within their natural, spontaneous context over a larger period of time. Therefore, they contribute to other research methods as a valuable addition. As an introduction into diary studies as a research method we will start off with a disquisition on the method. On the one hand this will cover the point of view of human-computer interaction research and on the other hand describe a psychological perspective. This inspection examines the diary method as such and is independent of technology. Due to this we form a theoretical basis that will be used as the foundation for specifying requirements for a multimodal diary application for field studies. Moreover, we introduce an extension of diary designs that allow for taking advantage of technologic progress. The specification on its part will be used as a guideline for the development of a diary tool that aims at improving the data gathering process, which will be presented afterwards. The implemented diary application has been used in two case studies with 10 participants each to examine on the one hand its user interface for usability. Results show that its concept had been understood and was applicable for the participants. Moreover, providing multimodal recording capabilities can improve the data gathering for the participants. Findings of a generally very high participant satisfaction with the tool outweighed marginal criticism towards little issues. Furthermore, all participants emphasized the usefulness of the tool. On the other hand we investigated our specification of requirements and added some missing findings to the list. Lastly, this thesis is concluded with a summary of the results as well as an outlook on promising future work. Latter is based on the specification of requirements and the insight gained from the case studies.



Dürr, Maximilian

Navigating Virtuality in Reality, InformationSense: A system facilitating direct interactions with digital information spaces in the real world (2016)

The given work introduces the development and implementation of a system which supports direct interactions with digital information spaces in reality. The focus resides on the use of deformable cloth surfaces. Besides a brief introduction in regard to the navigation of virtual spaces in the physical world, the framework of Reality-Based-Interaction, as well as a model of design spaces is taken into consideration. The benefits of a system which facilitates the deformation of a textile in reality are highlighted. Nine existing interfaces are verified, against predefined requirements, in context to their support for the navigation of electronic spaces in the real world. Interaction techniques for different tasks are extracted. They are surveyed in regard to their tradeoff between computational power and reality. Based on the analysis results, an interface which utilizes a flexible cloth surface, imprinted with invisible markers, is introduced. Digital content is projected. The system is enhanced with rigid lenses to facilitated detailed explorations of electronic information spaces. In course of a user study the developed cloth interface was compared with two other systems which simulate a deformable surface on a multi-touch table. A search & compare task was used. The results indicate advantageous properties of the touch interfaces in regard to pragmatic qualities and suggest that users preferred the real world textile in terms of hedonic qualities. The evaluation of participants interactions and comments provides insights into their strategies and reveals that they frequently conducted interactions based on their pre-existing knowledge of the physical world when they worked with the real textile. Participants appreciated the high degrees of freedom supplied by the flexible real world cloth object. In conclusion, based on the studies results, recommendations for the future design of deformable textile displays are proposed. 


Erdös, Paula

Evaluierung des Einflusses von Interaktionstechniken auf kollaborative Designprozesse bei Tabletop-Anwendungen (2014)

Diese Masterarbeit stellt ein Framework vor, das für die Evaluation von Kollaboration am Tabletop ausgelegt ist. Bisher existiert kein ganzheitliches Framework für Evaluationen am Tabletop, da jede Evaluation einen anderen Schwerpunkt setzt und die Studien untereinander daher nur schwer vergleichbar sind. Aus dem Grund, dass die Evaluation von Kollaboration viele Aspekte berücksichtigen muss, ist es notwendig einen systematischen Ansatz zu wählen, um der Evaluation eine Struktur zu geben. Damit wird sichergestellt, dass keine wichtigen Aspekte vergessen werden. Gleichzeitig wird es dadurch möglich, effektiver und effizienter Evaluationsstudien zu planen und durchzuführen.Das hier vorgestellte Framework basiert auf dem etablierten Input-Process-Output Modell, welches den Vorgang der Kollaboration in seine gleichnamigen Teile aufschlüsselt – den Input, den (Kollaborations-) Prozess selbst und den Output. Jeder Teil beinhaltet dabei Faktoren, die für die entsprechende Phase von Bedeutung sind. Zusätzlich zeigt diese Arbeit, welche Methoden verwendet werden können, um die einzelnen Faktoren messbar zu machen.Anhand des Input-Process-Output Modells (IPO) wurden in einer Studie verschiedene Faktoren der Kollaboration evaluiert. Der Input dieser Studie definierte sich dabei aus dem designspezifischen Aufgabentyp und der Tabletop-Technologie, die hierfür verwendet wurde. Die Teilnehmer sollten mit Hilfe einer Tabletop-Anwendung ein Flexible-Office- Raumkonzept entwerfen, wobei zwei verschiedene Varianten der Anwendung zum Einsatz kamen, welche sich im Umfang ihrer Funktionalität unterschieden.Um den Kollaborationsprozess zu evaluieren, wurden verschiedene Methoden verwendet, die zur Betrachtung unterschiedlicher Perspektiven der Zusammenarbeit dienten. Zum einen wurde die Studie per Video aufgezeichnet und anschließend per Video- Kodierungssoftware analysiert, zum anderen wurden alle Interaktionen auf dem Tabletopsystem geloggt, was eine anschließende Visualisierung ermöglichte. Die eigens für diese Studie implementierte RoomVis Visualisierung vereint die Daten der Video- Kodierung und der Tabletopinteraktionen in einer übersichtlichen Zeitachsen Darstellung. Zusätzlich wurde eine weitere Prozessvariable durch einen Teamworkfragebogen gemessen, welcher die subjektive Meinung der Teilnehmer zur Zusammenarbeit festhielt. Die letzte Kategorie des Input-Process-Output Frameworks entspricht dem Output, welcher durch zwei weitere Fragebögen und durch eine Expertenbewertung hinsichtlich der Qualität der Gruppenresultate evaluiert wurde. Einer der Fragebögen maß die Benutzerfreundlichkeit des Tabletopsystems und die damit zusammenhängende Zufriedenheit der Probanden. Die letzte gemessene Output-Variable bilden die geteilten mentalen Modelle, welche ebenfalls durch Fragebögen festgehalten wurden. Die Auswertung der Daten hat ergeben, dass die Homogenisierung von mentalen Modellen durch die erhöhte Interfacefunktionalität in der Experimentalbedingung stärker gefördert wird. Zudem haben Probanden in dieser Bedingung signifikant bessere Ergebnisse produziert und akkuratere mentale Modelle von ihrer Lösung gehabt. Es hat sich auch gezeigt, dass die erhöhte Interfacefunktionalität positiven Einfluss auf die Anzahl der gemachten Interaktionen hatte.Insgesamt wurden viele Methoden zur Messung von Kollaboration angewandt, dabei haben sich vor allem die Fragebögen bezüglich der mentalen Modelle und das Logging der Interaktionen bewährt. Zudem hat die Visualisierung einen guten Überblick über die Vorgehensweise der Probanden gegeben. Die Expertenbewertung hat geholfen Qualitätsunterschiede von Lösungen festzuhalten und diese mit der verwendeten Vorgehensweise in Verbindung zu bringen. Das Hauptfazit dieser Masterarbeit ist, dass das weiterentwickelte Input-Process-Output Framework und die vorgestellten Evaluationsmethoden ein adäquates Mittel zur Erfassung von Kollaboration am Tabletop bilden.


Fäh, Simon

Bimanual Pointing Techniques for Cross-Display Interaction (2011)

Multi-display environments can nowadays be found in meeting rooms,control rooms and other locations. Common multi-display environments provide limited input capabilities. Often, some important requirements such as mobility, privacy, awareness or similar cannot be granted. Based on a set of specific tasks for collaborative work, this thesis analyzes several aspects of cross-display interaction. In order to provide the most natural and efficient way of interacting with such an environment we propose bimanual gesture interaction. Using the two-handed approach does not only support cross-display object movement, but also manipulation of objects and more.

We introduce HyPoba Pointing, a novel pointing technique for bimanual crossdisplay interaction. This technique is designed for the special requirements of collaborative multi-display environments. Pointing gestures allow the control of one cursor for each hand and static gestures trigger various actions. MultiDragger, a tool which is designed to support bimanual pointing gestures in multi-display environments allows us to test HyPoba Pointing.

During an experiment with 14 participants, MultiDragger was tested using HyPoba Pointing and Absolute Pointing. The experiment contained two tasks. A single-handed drag & drop task and a bimanual manipulation task. The results show significant faster task completion time for Absolute Pointing and significant better accuracy for HyPoba Pointing. Whatever technique is preferred by the user, the development of MultiDragger has shown the possibility of bimanual gesture interaction for multi-display environments.


Fauth, Anja

Entwicklung und Evaluation eines Konzepts zur interaktiven Analyse von Tagebuchstudien (2013)

Diaries are a research method which aims to collect data in the everyday life of a participant. As they are used in the field, a researcher only has little to no effect on the motivation of the diarist and thereby the quality of the diary entry. Therefore it’s essential to influence the data collection during an ongoing study in order to enable feedback. This is possible with the usage of mobile devices, such as smart phones. To recognize whether feedback is meaningful or not contemporary data analysis is recommended. For this reason this work presents a concept for an interactive diary analysis tool.

Initially the characteristics of diary studies are evaluated and then appropriate analysis processes are presented. The analysis of real-time data plays an important part because it allows the researcher to decide when feedback is appropriate. The temporal progress of the real-time data can be understood by means of visualizations. The explanation of the main issues of time-based visualizations is followed by the definition of requirements for the analysis tool. They provide the basis for the developed concept which is evaluated in a usability study. The findings attest positive assessments and suggestions for tool improvements.



Feyer, Stefan

Ego OR Exo: Comparing Visual Perspectives on Guidance Visualisations for Motor Learning (2021)

Motor learning is a substantial part of life, like in sports, arts or the ergonomic handling of physical load. Motor learning is traditionally performed with the help of a teacher. If a teacher is not available, a digital guidance visualisation in Virtual Reality can be consulted. When motor learning is done together with a human teacher, a learner can watch the teacher’s movements from an exo-centric visual perspective (third-person perspective). In contrast, in Virtual Reality, a guidance visualisation can be seen from the ego-centric visual perspective (first-person perspective), too. The change of the visual perspective on the guidance visualisation influences motor learning. However, the empirical evidence about how the change of the visual perspective influences motor learning is low, especially for full-body movements, for tasks which include the ergonomic handling of a physical load and visual perspectives that utilise ego-centric and exo-centric guidance visualisations simultaneously. Furthermore, the field of ego-centric guidance of locomotion movements is unexplored. This master’s thesis proposes an experiment to close this research gap. The experiment compares an ego-centric visual perspective, an exo-centric visual perspective and a visual perspective which combines both. The experiment utilises a task which consits of elemental tasks of handling physical load. For the evaluation of the experiment, accuracy measurements, ergonomic measurements, the learner’s visual focus and qualitative data is taken. The experiment was evaluated with a pilot study and proved to be suitable to generate the data to close the above-mentioned research gap. First data indicates that the presence of an ego-centric guidance visualisation positively influences accuracy and attracts the visual focus of the learner.


Fink, Daniel

Re-located: Aligning Interaction Spaces to Support Remote Collaboration using Augmented Reality (2020)

This thesis introduces a prototype called Re-located that facilitates synchronous remote collaboration in Augmented Reality independent of room architecture and furnishing. First, in a one-time room calibration process, the positions of screens in a room are calibrated. The purpose of every screen will be predefined in this process. Afterward, by wearing an Augmented Reality headset, remote users, and their movements, are visualized as 3D avatars in front of the screens with the same purpose. Furthermore, the prototype integrates a voice chat supporting spatial sound. The influence of the prototype was investigated through a usability study. The results of the study were that the prototype provides overall high user experience and a high sense of social presence and social richness, while the sense of spatial presence was rated rather neutral. The visual appearance of the avatar was rated as too abstract, and the missing gestures and facial expressions were noted. The spatial sound feature was noticed by over half of the participants and rated positively. Most participants have not noticed that the room settings were different. Overall, results indicate that the concept of Re-locations works and provide a foundation for future work.


Foehrenbach, Stephanie

Hand Gesture Interaction for Large High-Resolution Displays: Design and Evaluation (2009)

Large high-resolution displays are widely used in academic and business contexts. Those kinds of displays offer great advantages for information visualization and can improve user performance. However, they are challenging for human-computer interaction, as they lead to more physical user movements. Therefore flexible input devices and interaction techniques are needed which allow interaction from any point and distance. This thesis investigates hand gesture interaction techniques, which meet the mobility requirement, but also provide a natural and expressive way of interaction. For the two most basic tasks in today’s graphical user interfaces pointing and selecting, we identified suitable hand gesture interaction techniques, based on hand gestures used in human-human communication and previous research. To underline the analogy to real-world interaction, we provided additional tactile feedback to the users? fingertips for target crossing to enhance the selection task. Previous research suggests that different movement directions of input devices, achieved with different physical user movements can influence user performance. With our hand gestures performed in mid-air no external physical support can be given, to guide user movements and compensate for irregularities. Different directions for cursor movements may therefore reveal different user performances. To assess the performance and acceptance of our proposed hand gesture techniques for pointing and selecting, and the influence of additional tactile feedback and movement direction we conducted a comparative evaluation study based on the ISO 9241-9. The 20 participants performed horizontal and vertical one-directional tapping tasks with hand gesture input with and without tactile feedback in front of the Powerwall of the University of Konstanz, a large high-resolution display (5.20 x 2.15 m). To track hand and finger movements and provide tactile feedback, our participants were equipped with a commercial data glove. For fast and robust gesture classification we applied an algorithm based on geometrical gesture models and state dependent threshold comparison. In contrast to previous research we cannot confirm a benefit of tactile feedback on user performance. Furthermore we found a significant difference in favour of the horizontal target alignment compared to the vertical one in terms of the effective index of performance. The non-tactile version of our hand gesture interaction techniques was very well received by our participants, and the effective index of performance with a mean of 2.53 bits/s for vertical and 3 bits/s for horizontal target alignment is promising and suggests that our hand gesture interaction techniques provide an adequate and valuable interaction technique for large high-resolution displays. To navigate within the presented information space on a large-high resolution display and explore it, users can physically move. From a distant position they can gain overview information, while moving closer reveals more details. However, physical navigation may not always be sufficient. Some parts of the display may always stay distant to the user, such as the upper part. To complement physical navigation and compensate for its limitations, additional interaction techniques are needed for virtual navigation. Therefore, we extended our set of gesture techniques to support panning, dragging and zooming tasks too, as those tasks are commonly used for virtual navigation. Based on interaction with physical objects and human-human communication, we identified suitable hand gesture interaction techniques, which fit seamlessly in with our existing gesture set. Limitations of commercially available data glove solutions, such as arising issues of hygienic or fit, and the observed restriction of user movements, motivated the design of Whitey, a novel data glove solution. Whitey combines an optical tracking system with a modified textile glove and a finger classification algorithm. For the classification of fingers this algorithm takes advantage of biomechanical constraints and heuristic knowledge on finger movements. Whitey can be adapted to different hand sizes, and used almost instantly by different users without the need for an individual calibration session after an initial set up. We conducted informal user studies to get initial feedback and first experience on the usability of our extended hand gesture interaction set and Whitey. We found that users could easily learn and apply our techniques. We could further gain valuable insides for fine tuning Whitey and our hand gesture interaction techniques to improve user interaction. For future work we plan to further extend our set of hand gesture interaction techniques to utilize more of the potential hand gesture interaction holds for human-computer interaction, in terms of naturalness and expressiveness.



Gebhardt, Christoph

Integrative Workplace: Employing Reality-based Interaction to join Digital and Analog Media at a Workplace (2013)

This thesis presents the development and the evaluation of an interactive workplace which supports knowledge workers by integrating printed media. The major benefit of the workplace is the possibility to use features of digital devices within paper documents. Features of this thesis’ prototype that demonstrate this benefit are digital full- text search as well as the simple excerpting of contents. Furthermore, digital content and its analog sources can be related to each other. To process digital and printed texts in the same manner, the same stylus and the same interaction techniques are used. The unique characteristic of this system is the combination of aforementioned features and their application in a real working environment.
A case study with participants solving a legal record showed that this system is actually facilitating knowledge work. The study revealed that it is necessary to picture the linear evaluation order of legal issues to support the juristic working methodology. In addition it proofed that users appreciate the way paper was integrated into the prototype and enjoy the experience of using features known from digital devices on paper documents. Most of all, the evaluation discovered that users having problems with the distinction between pen and finger for context-switching in pen and touch user interfaces.



Gerken, Jens

Orientation and Navigation in Zoomable User Interfaces in consideration of Findings in Cognitive Psychology (2006)

Zoomable User Interfaces (ZUIs) allow access to information-objects in a natural and visual-spatial way. Instead of navigating in abstract hierarchies and structures, the user can navigate in an infinite information landscape by zooming and panning. Information-objects are therefore organized in space and scale, which allows the user to navigate analogue to the real world. If someone would like to access more information about an object in the real world they simply step closer to it. In a similar way, the knowledge about the location of objects can be compared to remembering where one has placed a certain book on the table. By transferring these concepts to the design of user interfaces it is made better use of the cognitive abilities of people, which have been specialized for these natural visual-spatial ways of navigation and orientation since many centuries. However this also implies some problems, which have to be considered. As in the real world, there is a certain danger of getting lost in such an infinite information landscape.


Geyer, Florian

A Zoom-Based Specification Tool supporting Interdisciplinary User Interface Design Processes (2008)

Careful design of user interfaces is getting more and more important as computers increasingly determine the capabilities, limitations and organization of our work. Today, interfaces have to meet an extensive number of requirements on performance, functionality and usability. Additionally, the quality of user interfaces is also an economic factor, as it has a dramatical influence on customer satisfaction and brand value. Consequently, structured approaches to user interface design are characterized by a multitude of tasks and techniques that make interface development a complex process. However, current design practice is accompanied with a variety of employed tools and frequent transitions from abstract problem definitions to detailed specifications. Based on a detailed investigation of interface design theory and practice, this thesis introduces a novel interface specification tool that aims on addressing the interdisciplinary nature of the design process and gaps in communication and technology in order to ease work transitions. Nevertheless, it also aims on promoting creativity and innovation in design by offering informal means of expression and by allowing to relate design artifacts in context. It contributes to current design practice by respecting the need for iterations and continuous reviewing as well as collaboration among designersthroughout the design process. This thesis explains the rationale behind the developed tool by a detailed examination of applied theoretical foundations, related research, practical design methods and utilized interface concepts, before presenting a conceptual model that led to the physical interface design. As a result of careful integration and implementation of innovative, zoom-based interaction and visualization techniques that closely resemble transitions in work practice, the introduced tool design extends the perception of interface specification tools to a broader, more comprehensive coverage of the design process. After describing the technical implementation, the thesis concludes with a review of contributions and remaining issues as well as implications for future investigations.



Gröschel, Carla

Caring for Carers - Design and "in-the-wild" Evaluation of a Mobile System to Support Nurses in Applying Kinaesthetics (2019)

Caregivers are at a higher risk to suffer from workrelated musculoskeletal disorders (WRMD) such as low back pain. One of the factors to be influencing is the daily transfer of patients. Kinaesthetics is one of multiple care concepts that try to communicate safe patient handling techniques to address this issue. Yet, the integration in the curriculum in Germany solely consists of a three-day basic course. A lack of follow-up support leads to a lack of knowledge and therefore appliance in practice.

This thesis examines how a mobile system, in the following called NurseCare, can be designed to face this lack of support in practice. NurseCare is a combination of two components: A wearable to provide real-time feedback on risky movements regarding the back and a smart- phone application to supply additional support. Distributed Cognition, the Behavior Change Wheel Framework, and Participatory Design were taken into consideration to understand the theoretical background. The analysis of related work revealed a gap in research of systems that combine sensory and mobile technology to support safe patient transfer techniques during work. Thereupon, two conducted workshops could help to understand the design space in this complex work domain. The workshops’ results guided the requirement analysis, as well as the design process of NurseCare. The first workshop involved nurses as end-users and the second workshop additionally HCI students as technical experts.

NurseCare was implemented as a prototype and evaluated "in-the-wild" with nine nursing students. Data of questionnaires, interviews, an e-diary, and log files were analysed to investigate how caregivers experience the usage of NurseCare in its natural setting. The findings reinforce the need for the support of Kinaesthetics in practice. Moreover, the design of NurseCare as a wearable and smartphone application was perceived as expedient and easy to use in everyday life. The findings indicate that an integration of a system like NurseCare is wished and revealed ideas for its realisation. The examined benefits and shortcomings were furthermore taken as a basis to generate implications for future work.


Grün, Christian

Development of a visual meta-data Browser for the University library of Constance (2004)

This Master Thesis illustrates the conception, development and implementation of a browser for Meta Data in libraries. The browser, which is called MEDIOVIS, provides on the one hand search facilities in the library data base and offers on the other hand different views on the displayed data, aiming to give the user a quick, simple and still versatile access to relevant information. MEDIOVIS was successfully established in the Mediothek Konstanz, a section of the library which also offers numerous audiovisual media such as CDs, videos and DVDs beside books and papers. The existing library data was enriched by external data in order to give the system additional value on content base. MEDIOVIS was designed generically by the beginning so that it can easily be adapted to other library data bases and application domains.



Grupp, Veronika

Blended Museum – Die didaktische Bedeutung des sozialen und interaktiven Kontextes und dessen Umsetzung durch Informations- und Kommunikationstechnologien (2009)

Ziel dieser Arbeit ist es, die Wirkung dieses sozialen und interaktiven Kontextes auf den individuellen Lernprozess in der Lehr- und Lernforschung zu analysieren und die sich hieraus ergebenden Potentiale für ein Blended Museum zu entwickeln und aufzuzeigen. Maßgeblich hierbei sind Ausgestaltungsszenarien, die die Ergebnisse durch einen sinnvollen Einsatz von Informations- und Kommunikationstechnologien realisieren.


Guliyev, Javid

Investigating the Influence of Display Size on Cognitive Load in Visual Search Tasks using Eye-Tracking Technology (2018)

Over the past few years, researchers are showing interest to study the level of cognition with the modern eye tracking technologies which can accurately measure gaze movements. Rise of the eye tracking technology encouraged researchers to use cognitive load as another measurement in the field of Human-Computer Interaction as well.

In this thesis, we use an eye tracking technology as a tool to assess the influence of the display size on the cognitive load in visual search tasks. We designed and conducted an empirical study with 35 participants to investigate if the display size has any effect on the level of cognition. First, we analysed how gaze measurements were affected while changing the level of cognition in two different sized displays. Later, we investigated if the display size had any influence on the gaze measurements. Using this knowledge, we discussed how the display size influenced the level of cognition in visual search tasks.

During the experiment, we collected not only the gaze measurement, but also the level of performance and users’ subjective thoughts about their mental effort.

Our objective and subjective analysis shows that display size has an influence on the cognitive state and there is a tendency for the large display leading to a higher cognitive load in visual search tasks.



Gundelsweiler, Fredrik

Search and Exploration in Complex Information Spaces by Example of the Digital Sales Channel of DaimlerChrysler (2005)

This thesis is about search and exploration in complex information spaces of the digital sales channel of DaimlerChrysler (DC). Requirements, usability goals and styleguides about search and exploration are collected. More constraints concerning the internet presence of Mercedes-Benz (MB) come from DC. A theoretical overview of search and exploration leads to the examination of different search systems in practice and research. The internet presence of the future is a multimedia experience world where users explore and collect the information they need. These future worlds need innovative interaction techniques and search result presentations. Different studies show that search is no alternative to navigation on websites. But a search function is necessary to let the user locate needed information efficient and effective. Main focus of this thesis is the description of developed search and exploration concepts and the resulting design studies concerning the internet presence of MB. Finally an advanced car search concept is presented and the future development of search and exploration is discussed.



Hankh, Markus 

Entwicklung und Evaluation eines Konzepts für digitale und kollaborative Kreativarbeit im Kontext von Storyboarding (2013)

Storyboarding in HCI and Interaction Design is a common design technique located in all phases of the design process. Thus, storyboards describe user interaction with a system or the graphical user interface. Designers are able to communicate thoughts and concepts as well as emotions to improve understandability and communication between them and all stakeholders. Therefore, storyboarding provides a visual language which talks back to designers. Storyboards have always been created within a design team. Creative teamwork is also called collaboration which consists of communication, cooperation and coordination. While creating a storyboard most collaborative techniques are performed and lead to great advantages and therefore to better results. The downsides of creative teamwork are disturbances which are called social factors. Determining the creation of storyboards, all digital tools and techniques are designed for single usage. At this time, there is no interactive tool which supports collaborative storyboarding. So the purpose of this thesis is to develop a concept for such an application. To maintain the same creativity, this application has to adopt used and well-known design procedures into the digital world. Thus, the system called Collaborative Storyboarding supports individual sketching and public presenting and discussing. Storyboards can be created, organized and shared within a design team. To evaluate the system, a usability test was performed on real users to determine usability problems, collaboration techniques and the appearance of social factors.



Heberle, Siegrun Annette

Förderung eines benutzerzentrierten Software-Entwicklungsprozesses in Unternehmen am Beispiel von UBS Schweiz (2002)

Why are we still confronted with such a poor usability in many software products? What lets all those big companies ignore the obvious opportunity to increase their efficiency, effectiveness, and employee satisfaction? This master thesis demonstrates that we have all required means in both, theory and practice, to develop usable software. Hence, the problem is not the missing solution but the missing awareness of the solution, and, even worse, the missing recognition of the problem itself. Considering user-centredness as innovation and the process to its implementation as organisational learning delivers useful indications to further strategic efforts on the way to user-centred development processes. Following the ideas of several widely distributed software engineering and usability engineering standards this thesis presents a maturity model of an ideal user-centred organisation. The model distinguishes seven dimensions: (1) the development process, (2) methods and tools, (3) the organisational structure, (4) employees qualification, (5) the education system and knowledge management, (6) management activities and culture, and (7) resources; three maturity levels lead the way to real user-centredness. The applicability of the maturity model is demonstrated in a case study about UBS Switzerland. The case study shows the benefit of using a model to support analysis and to direct strategic considerations with regard to user-centredness. Finally, the latest technological approaches to usability engineering and organisational learning are discussed.


Heilig, Mathias

Integration of model-driven software-development methods into a Usability-Engineering-Process by example of Dr. Ing. h.c. F. Porsche AG (2007)

More companies than ever before recognize the high value of useable user interfaces (BURMESTER et al. 2005). Contrary to the development of human-machine interfaces for devices such as mobile phones, televisions etc. the development of driver information systems for automobiles is considered with special basic conditions. The tendency to increase the amount of functions in the vehicles is unbroken (Infotainment, Internet, Information and Telematik services) though many dependencies result. Thus, the complexity of the development of an entire system rises continuously. Parallel to this fact it is imperative to adapt the interfaces between humans and technology to the increased complexity. These require an adjustment of common approaches of the Requirements-Engineering and system specification.

The master thesis describes the possibilities of a reasonable integration of model-driven software development methods into a Usability-Engineering process by the example of the HMI-Development-Process of the Porsche AG. This process is a special Usability-Engineering development process for the design of driver information systems at the Porsche AG.

After the analysis of the established Usability-Engineering processes and their qualitative evaluation possibilities, the concept of the model-driven software development is introduced. That includes the presentation of the current possibilities of this approach within the research field of the model-driven generation of user interfaces. Based thereon the theoretical possibilities are described, how model-driven development methods are able to support a Usability-Engineering development.

The next part of the thesis describes a practical scenario of the approach for the development process of a driver information system at the Porsche AG. This happens in the context of an evaluation of alternatives and an explanatory statement of the choice of tools. Furthermore, the used visual modelling language is described. For this language the meta-modelling paradigm is used to create an own domain-specific modelling language. The feasibility to generate an output from the visual specification is proven by the example of an interactive prototype.

Thereafter the effects on the HMI-Development-Process are brought up for discussion. An examination of the quality of the process is accomplished based on the DATech test manual for Usability-Engineering development processes. Thus can be determined, in which stages a qualitative progress can be achieved by the application of model-driven methods.

Finally, the concept is regarded and analyzed from different perspectives. The insights of the thesis are abstractly represented with the purpose of the possibility to transfer these also to other Usability developments.


Höchtl, Anita

Supporting methodic design practices with interactive organization and visualization of design artifacts (2011)

This work describes the support of designers in their design environment. The focus is on the organization and visualization of design artifacts like images, drawings or sketches on digital surfaces. Besides a short introduction in the design process, the physical design environment is one main issue. Therefore, the need for some kind of formal organization in a collaborative design environment is highlighted and the dealing with visual design resources is analyzed. These shared surfaces can be supported by the use of digital technologies such as touch-sensitive surfaces. The visualization as well as the interaction concept are of high relevance in order to allow an appropriate use of these surfaces. For this purpose, different spatial layouts are considered and current existing systems are verified concerning their usability in the design space. Two interfaces for structuring visual artifacts are introduced on the base of the analyses and various design principles. The results of a comparative user study, which was conducted on a digital tabletop, give a first insight into the usability of both interfaces. The following discussion is concentrated on the adequatness and effectiveness of these, the user performances and the user preferences. Moreover, results explain the applied interaction strategies and show the use of bi-manuel and multi-finger interaction. Finally, conclusions concerning the design space are drawn and suggestions for future improvements are made. With regard to future work, a holistic system, which provides a remote control for visualizations on large displays besides supporting the organization of design artifacts on a digital surface, is presented.



Hofmann, Ingo

Entwicklung von Interaktionskonzepten zur Erstellung von wissenschaftlichen Arbeiten an einem interaktiven Tisch (2014)

This thesis deals with the question of how to support the development of a thesis manuscript in the library of the future by utilizing the inherent advantages of vertical and horizontal displays. To start with, the usual working methods of knowledge workers were examined in order to define key requirements for an interactive system to be developed. Following this, findings from related work about how to apply vertical and horizontal displays in the context of writing scientific papers and how various input methods, such as gestures or digital pens, can support the work process efficiently are presented. Based on this analysis, new concepts for a system which combines an interactive desk with a vertical display were developed. The focus was on an additional visualization of the document to be edited on a horizontal display that supports and simplifies the outline and structuring of scientific documents. These concepts had been implemented prototypically and got improved during an iterative expert user study. A final empirical evaluation showed that the developed system has a good usability and was considered helpful by target group users. Furthermore, it was found that most users of the system prefer the use of a second vertical display rather than an interactive table.


Hofmann, Jacqueline

bibox: Konzeption und Evaluation eines Tangible User Interface zur Förderung der Partizipation in öffentlichen Bibliotheken (2015)

The library 2.0 concept aims to transfer existing technologies and principles of the web2.0 to the context of public libraries. It focuses on the idea of active participation and interaction from library visitors to shape the future of their local library. Based on this concept, a system was designed to motivate library visitors to contribute to their local library in form of book ratings. This work opens with an analysis of public libraries and its user groups. Based on an examination of actual motivation concepts and design guidelines,requirements for the development of a participatory system are deduced and applied to the context of public libraries. During an iterative design process, the system was developed and theoretically evaluated against these requirements. To practically evaluate the system,a one-week In-the-wild study was conducted in the public library of Cologne. The system continued to stay on site for an additional two months, for a long-term analysis of the real interaction. The studies revealed important insight regarding the use of tangible objects in interactive systems and how they are being accepted by the public. Moreover, some possible difficulties regarding a flexible interaction process have been discovered during both studies. The thesis concludes with an outlook on how the system can be improved to enhance acceptance by visitors of public libraries.


Höhn, Stephanie

Ideaflow - Supporting the Collaborative Idea-Finding Process for Design Teams On-The-Go (2013)

In the early stages of a design project, designers produce a variety of different ideas. Combining various individual perspectives on a design problem leads to various solutions and ultimately to more creative products. Therefore, these ideas are generated and evaluated preferentially in a collaborative manner. The three basic requirements for effective design work are thus collaboration, communication, and awareness of the work of colleagues.

The desired requirements for effective design work, however, are not met adequately if designers are on the go. The main difficulty lies in the unsatisfactory methods available to designers for recording ideas and communicating them, leading to lost opportunities to record spontaneous ideas generated as a result of inspiration by stimuli from the outside environment, such as smells, colors and shapes. This challenging task of developing software that aims to support collaborative idea generation by design teams by providing designers with the ability to record and share ideas on the go is the focus of this study.

Having determined the requirements of such a software the prototype \"Idea Flow\" was developed with the aim of digitally supporting the collaborative design workflow. This system provides methods for structured recording, archiving, and exchange of ideas created on mobile devices for distributed design teams, and for the collective review of the ideas on a central desktop application. In order to evaluate the efficacy of the system for, and its benefit to, the design work, a case study was conducted with the participation of designers from the creative sector. The study sheds light on the following questions:

  1. Can collaboration and idea generation of creative design teams on the go be effectively supported by system notifications on new shared ideas?
  2. Does the Ideaflows idea archive encourage more awareness on the work of colleagues?
  3. Are more related ideas produced through the awareness of the work of team members precipitated by the archive and the system notifications?

In conclusion, with the provided functionality of Ideaflow we could support the mobile and collaborative idea generation process of creative teams digitally. Further improvements are identified which support designers on the go in a way that is parallel to the effective workflow in design offices. We found that systems should enable spontaneous idea capturing, communication of ideas to colleagues, and awareness of the work of collaborators.


Hubenschmid, Sebastian

MIDAiR - Multimodal Interaction for Visual Data Analysis in Augmented Reality (2019)

The research area of immersive analytics investigates how emergent technologies, such as mixed reality devices, can be helpful in analysing complex data. Prior work in this area shows how a multimodal interaction approach using a large touch-sensitive tabletop can be beneficial for interacting with an abstract 3D visualisation. This work improves upon prior work by extending the design space, utilising more input modalities to address several shortcomings.

Firstly, theoretical background for interacting with visualisations as well as several input modalities are investigated. This leads to functional requirements that guide the selection of used input modalities. Furthermore, related work is analysed in regard to the usage of input modalities in the realm of multimodal data analysis.

Secondly, the Multimodal Interaction for Visual Data Analysis in Augmented Reality (MIDAiR) system is presented. MIDAiR is a collaborative immersive analytics tool for the exploratory analysis of multidimensional abstract data. MIDAiR offers a novel combination of a spatially-aware tablet with an immersive augmented reality device that allows for multimodal interaction with a 3D visualisation. The used 3D visualisation consists of several linked 2D scatter plots, forming a 3D parallel coordinates visualisation. This enables users to easily filter and analyse data, allowing for the detection of clusters, trends, and outliers within the data set. MIDAiR thus expands the design space of prior work, offering a more flexible analysis workflow.

Thirdly, a usability study with eight participants was conducted to test the interaction concepts of MIDAiR. Although users needed some time to get used to the interaction concept, they fully employed most input modalities to accomplish the given tasks. Based on the results of this study, design recommendations and further research directions for immersive analytics tools are presented.


Huber, Stephan

Entwicklung und Untersuchung eines räumlichen Interaktionskonzepts für die Navigation in zoombaren Benutzerschnittstellen (2013)

Digital information mostly consists of different visual contents such as music, videos, pictures or texts and dominates the life of most people. The amount of digital contents generated by humans seems to increase exponentially from year to year and this begs the question of how we can handle this overwhelming mass of information in the future. Especially in the area of knowledge work, it is essential to assist people in efficiently managing information. Current operating systems support this through a hierarchical file system. Zoomable user interfaces (ZUIs), however, offer the possibility to organize digital information in a spatial way and are based on the knowledge gained from the development of Post-WIMP User Interfaces. Post-WIMP User Interfaces have been focusing on the development of user interfaces that involve the user’s understanding of the physical world and motor skills. These relationships were examined in recent studies. However, those studies did not particularly disclose any distinctive spatialbased interaction concepts for the navigation in ZUIs, which are more focused on the motor skills of human beings. Therefore, this thesis introduces a spatial navigation concept for ZUIs, which can be used as an alternative navigation concept compared to previous touch based input gestures. The result of a user study reveals the impact of a spatial navigation concept on the navigation performance and memory capability, in comparison to conventional touch-based input gestures.



Jain, Vidhu

Investigating the Influence of Display Size on Aspects of Spatial Memory (2017)

Spatial memory is an up-to-date research topic within the research domain of human cognition and HCI. Researchers have started to investigate how it can be harnessed in interaction design in order to improve users’ performance (navigation, search, recall etc.). In the first part of my thesis, I identify different aspects of spatial memory: Assembling/Disassembling, Rotation, Visual Recall and Navigation Recall based on an extensive literature review and an Affinity Diagram session with three HCI researchers. Furthermore, I have planned and conducted an experiment in order to investigate the influence of display size on the identified aspects of spatial memory. Comparing a 10.6 inches (SMALL) and 55 inches display (LARGE), participants were asked to perform four tasks (each reflecting one of the four identified aspects) in a controlled within design lab setting. Results of the study show that display size indeed has an influence on the Rotation and Assembling/Disassembling aspect of spatial memory. For the Rotation aspect, Participants performed significantly faster (time) on small display with difficult histogram (6-bar histogram and 270-degree rotation) and significantly more accurate on large display with difficult tasks (6-bar histogram). Participants performed significantly faster (time) when working with the small display in task of Assembling/Disassembling aspect of spatial memory.



Jetter, Hans-Christian

Information Architecture and Information Visualization for the Post-WIMP Era (2007)

This thesis introduces a concept for a fundamental redesign of todays user interfaces for the Personal Computer and the World Wide Web. The presented design ZOIL (Zoomable Object-Oriented Information Landscape) unifies principles of object-oriented user interfaces, zoomable user interfaces, and information visualization in an application-independent concept for navigation and work in complex multimedia information spaces of all kinds. ZOIL can be regarded as an extension of the dominant WIMP and hypertext user interface styles with direct-manipulative and intuitive navigation and novel information architecture and information visualization. It furthermore has the potential of replacing these styles with ZOIL as the new technological paradigm for PC and Web user interface design. 



Kleiner, Eike

Ein realitätsbasierter Ansatz zur Präsentation und Exploration von Bibliotheksbeständen (2013)

Subject of this thesis is the user interface Blended Shelf, which provides a shelf browsing experience beyond the physical location of the library. Shelf browsing offers numerous advantages and users apply it as a research strategy in libraries. Little usable and proven applications exist to provide shelf browsing in the digital domain, which would allow time and location independent shelf access for the users. Therefore, the aim of this work is to develop a user interface, which offers the experience of digital shelf browsing, without losing the essential advantages that are deeply rooted in the physical space.

To accomplish this, the first part of the thesis constructs a collection of basic requirements that need to be fulfilled to emulate the shelf browsing experience. The basics of these requirements are the theoretical background of shelf browsing, as well as an analysis of library specific aspects and user needs. The central parts of the work illustrate how the usage of the requirement collection serves as a foundation for the concrete implementation, the set of functions and the reality-based interaction design of Blended Shelf. Finally, an evaluation in form of a comprehensive field study checks whether the implementation meets the requirements and if the users perceive the User Interface as useful and usable. A description and discussion of the study design and results forms the last third of the thesis. An outlook to and discussion of open questions and future work concludes the thesis. 



Klinkhammer, Daniel

Blended Museum - Steigerung von Besuchererfahrungen durch Interaktions- und Informationsdesign (2009)

In this assignment we introduce the approach to the Blended Musuem. It attempts to increase the visitor experience by intersecting the virtual with the real museum as well as blending information and communication technology and the real museum. Based on the Contextual Model of Learning, various implemented museum-like installations are being introduced. They provide a basis for a taxonomy that classifies the application of information and communication technology to augment visitor experience. Based on this taxonomy tangible user interfaces are being introduced, facilitating the embedding of information and communication technology into the real museum in an innovative manner. By integrating these novel systems new ways of raising visitor experience are formed since these computer interfaces offer new means of human-computer interaction as well as the acquirement of information.


König, Werner A.

Reference model and feasibility study for a novel zoomable user interface paradigm (2006)

Today`s users are confronted with increased information complexity in terms of quantity, dimensionality and heterogeneity. Nevertheless it seems that most of the concepts used in user interface design still derive from the late 60s and beginning 70s. In this thesis a new zoomable user interface paradigm named ZOIL is presented. It aims at merging the concepts of zoomable user interfaces and the more traditional direct-manipulative, object-orientated user interfaces while making use of the benefits of spatial data management. It therefore offers a widely scalable information landscape integrating geometric and semantic zooming functionality. The ZOIL paradigm does not specify a certain user interface or application but offers a consistent set of visualization and interaction technologies and therefore can serve as a flexible basis for many different kinds of applications and application domains. In order to provide proof for the conceptual and technological feasibility of this paradigm it is applied and implemented as a prototype to the application domain of personal information management. Furthermore this thesis presents a new reference model for zoomable user interfaces, which tries to combine the reference model for information visualization as defined by Card and the rendering pipeline used in 3D computer graphics. It therefore aims at offering an integrated and efficient but nevertheless flexible model for the visualization process which meets the requirements of complex zoomable user interfaces in general and the ZOIL paradigm in particular. As an introduction to the subject matter a related work part presents and discusses several applications and scientific work using similar concepts as well as user studies done in the field of zoomable user interfaces. In addition psychological and physiological aspects of the human perception, basic navigation concepts and potential input devices for zoomable user interfaces are analysed and elaborated in regard to the ZOIL paradigm. The thesis concludes with an outlook for possible benefits of future projects derived from this work.



Konstanzer, Robert

Interaction and Filter Concepts for Graph Visualization (2007)

This master thesis deals with navigation and interaction concepts to visualize graphs. On the basis of product data management systems (pdms) a new concept for information access is designed and implemented in a prototype. This concept named Multi-Focus Filtering integrates the dynamic query approach into the graph using interactive clusters and combines it with a navigation concept. The thesis is structured in three parts. In the first part the problem space of pdms in information access is described and requirements are derived. The second part forms the theoretical basis. Scientific concepts dealing with interaction and visualization of graphs are presented. The third part describes the Multi-Focus Filtering concept with its elements, evaluates it and explains under which conditions the concept can be assigned to other domains.


Leifert, Svenja

Theorie und experimentelle Untersuchung des Einflusses von Interaktionsdesign auf das räumliche und inhaltliche Gedächtnis (2013)

Presentation and exploration of huge data spaces do often take place in panning or zooming and panning user interfaces in which the user goes astray easily if he loses track of the surrounding context. By integrating certain interaction components to support spatial memory into interface design, one could try to counteract problems arising while navigating through huge data spaces.

The centre of this work is the planning and execution of a quantitative experiment which analyses the influence of grids on spatial and content memory. The study examined the factors "visibility of grid lines" and "alignment", the structural or chaotic arrangement of objects on the canvas. Results suggest that both methods of structuring the canvas have a positive influence on spatial memory. But only if grid lines are visible, content memory is positively influenced by the structural alignment while, if they are hidden, there are no differences between the two alignments. However, in that case results for content memory are generally better than if grid lines are shown.

Together with two additional quantitative experiments exploring the influence of input devices (mouse versus multi-touch) in panning and panning and zooming interfaces, a more thorough picture of supporting spatial memory by prudent interaction design can be drawn. Suggestions for possible usage are made, and future work for further analysis of spatial memory in human-computer interaction is discussed.


Lorenz, Laura

Animation als Sketchingtechnik für Storyboards - Enwicklung und Evaluation eines Interaktionskonzepts (2013)

This thesis is about the development and evaluation of an interaction concept for animated storyboards in early design phases. Digital applications and their requirements have drastically changed in recent years: they became more mobile, interactive and ubiquitous. Interaction designers need to respond to this change. However, the use of traditional design representations limits the ways designers may express new concepts and communicate them to fellow designers. Animation, in conjunction with storyboards, could offer a solution for this issue. Today, storyboards are already a standard tool in the field of human-computer interaction and the potential of animation has already been discovered by various researchers.

In this work, first the foundations of sketching, storyboarding and animation, as well as various related works are analyzed. A set of requirements for an interaction concept, which supports the creation of animations in storyboards in early design phases, is worked out. The principle of transformation-by-example, in which a user performs the motion which should be animated, is recognized as promising and is integrated into the interaction concept. A prototype, implementing the aforementioned concept, was developed and finally evaluated in two user studies regarding its applicability and interaction difficulties. Subjects were able to express their ideas in animated storyboards and explain and discuss them with other participants. Furthermore, it was shown that given animations can successfully be replicated with the aid of the prototype. The studies revealed potential for improvement, which was partially taken up in redesign proposals.



Ludwig, Kerstin

VizBox - eine webbasierte Bibliothek für das Design von Visualisierungen (2005)

The subject of this thesis is the presentation and explanation of the VizBox, a web-based library for the design of visualizations. The library consists of two components: the visualization patterns and the structured descriptions of visualizations which both assist the designer in the creation of alternative visualization designs.

The goal of the VizBox approach is to combine the main advantages of two prominent design strategies from traditional design disciplines, that are also discussed in detail. First, there is the pattern approach, originally developed by the architect Chistopher Alexander which generates designs of high quality. The second approach is the so-called “Box-of-Ideas” which originates from the context of product design. This approach functions as an assistance for the designer in developing new and innovative proposals of designs while she is designing a visualization.

The thesis discusses how these approaches can be modified and utilized for the visualization of data. The VizBox combines the advantages of both approaches by adapting them for visualization design and embedding them in the generation of a visualization system.

The thesis presents a new pattern format which enables VizBox for the combination of the existing formats. Corresponding to this, a new description format for visualization is specified. Eventually a set of design suggestions for the VizBox as a visualization system is developed which draws profit from the combination of the two approaches and makes the derived approach applicable for the context of data visualization. An added value of the VizBox is the possibility to design a visualization system by the consequent use of userfriendly visualization techniques. The compilation of these techniques in a single library makes the VizBox also usable as a kind of web-based encyclopedia for the visualization of data, because it contains a huge amount of knowledge about this subject.



Mahendra, Tanveer Singh

CaregiVR: A Virtual Reality System that Provides Concurrent and Terminal Feedback for the Self-Directed Training of Ergonomic Patient Transfers (2021)

Caregivers often care for patients with functional disabilities. While assisting with a patient’s daily activities, caregivers are often required to transfer the patient. However, conducting manual patient transfers poses a risk to the physical health of the caregiver. Although the Kinaesthetics care conception can help address this issue, existing mechanisms for self-directed training are limited. This thesis presents a virtual reality system, in the following called CaregiVR, that can support self-directed training of ergonomic patient transfers by providing concurrent and terminal feedback. CaregiVR is a combination of two components: A feedback system for detecting risky behavior related to ergonomics and a patient transfer task in virtual reality. This thesis focuses on the effect of various types of feedback on self-directed training of ergonomic patient transfers. Theories related to feedback, Kinaesthetics, and information visualization inform the design requirements of CaregiVR. Relevant works from training in nursing-care and feedback systems in motor learning provide a foundation for the design process. A novel sketching template was a part of the design thinking process that facilitates easy conceptualization in the early stages of the design lifecycle. CaregiVR was implemented as a prototype of the design concept. The thesis proposes a study design for conducting a qualitative user study. This study design was iterated after a pilot test. Moreover, this work also discusses various techniques to analyze data collected by the proposed study design. Finally, we look into the benefits and shortcomings of CaregiVR and discuss how future researches could extend this work in various directions.


Martinez, Pablo

Merge Policy Comparison for AR HMD Collaborative Manipulation

Augmented Reality incorporates both virtual and virtual worlds. Head mounted displays provide a hands-free interaction with the virtual environment, without forcing the user’s attention to a screen like hand-held devices. For true immersion to occur, virtual objects need to possess the affordances of real objects, one of them being the concurrent manipulation from multiple users. Literature reveals two approaches for this collaboration: 1.Separation of DOFs splits the object’s degrees of freedom among the users, and 2. Combination of user actions aggregates the user manipulations into a final transform applied ot the object. Since any mathematical function can be calculated for Combination, related work has implemented a variety of them: 1. Sum adds all the user inputs together, 2. Mean averages user actions, and 3. Weighted mean assigns weights to the averaging process. Nonetheless, little comparison among these merge policies is provided in literature and none is available on HMD devices.

Therefore, this thesis provides a baseline comparison for the three approaches in a AR HMD docking task. To that end, a study prototype which implements the aforementioned merge policies is developed. A controlled lab experiment with 36 participant was conducted, where the performance, subjective workload and user experience for each of the conditions were measured using the prototype. While there were no significant differences among conditions in regards to performance, user preferences were polarized between the Sum and Weighted conditions. The Average was preferred by users to make any accurate placement, and the Sum needed the most coordination due to overshooting issues. Based on the obtained results, merge policy selection for future tasks in addition to a comparison frame for future collaborative AR HMD research are presented.


Marx, Stephanie

Considering aspects of book data for user-influenced recommendation in the public library context (2015)

Book recommendations are given by many institutes based on different information criteria. This thesis discusses different book recommendation algorithms and their information sources regarding their usefulness for a user-influenced recommendation process in public libraries. Based on the discussed algorithms a new hybrid approach is introduced, which generates book recommendations based on keywords and reader opinions. Additionally the interactive prototype Booquid is introduced to give the hybrid approach an user-friendly graphical interface. To assure the understanding of the hybrid approach by the users, Booquid is evaluated and advanced in this thesis. The result of the evaluations and this thesis is a hypothesis whether one of the basis data sources generates book recommendations of a better quality than the other.


Memmel, Thomas

An innovative navigation concept for complex information spaces exemplified by DaimlerChryslers digital sales channel (2005)

This master thesis describes the scientific derivation and prototypical realization of an innovative navigation concept for complex information spaces exemplified by DaimlerChrysler’s digital sales channel. The thesis starts with a preface, which explains problems of common navigation concepts in general and shortcomings of the website navigation of DaimlerChrysler in particular. Subsequently, the thesis is structured into three main parts. The first main chapter describes design principles for the prototypical implementation, based on DIN ISO standards and fundamentals of the disciplines usability engineering, human-computer interaction and emotional design. The findings prepare the following examination of models and techniques of information visualization with respect to feasibility of a navigation concept. Both chapters are the groundwork for the subsequent presentation of the new navigation concept showing different usage scenarios and various prototypical specifications. Throughout the explanation of the prototypes, the thesis documents how and where previously defined design decisions and principles were implemented and which trade-offs occurred as a result. The thesis closes with an outlook and a summary.



Müller, Jens

Reality-based Interaction Styles for the Manipulation of Process Variables in Control Room Environments (2012)

Control rooms are facilites that serve as operations centers to monitor and control complex processes, e. g. in power plants or industrial production plants. One essential task in operating control rooms consists in the manipulation of process variables which represent the physical state of the supervised process. In today’s control rooms operators manipulate process variables by means of desktop computers. However, the mediation of process states in terms of keyboard, mouse and computer displays is inappropriate as perception is heavily reduced to the visual sense. Hence, the operator is not sufficiently supported in his/her natural skills. The thesis at hand presents interaction concepts for the reality-based manipulation of process variables which build on the users’ skills and pre-existing knowledge. In the course of a usability test these concepts where compared with current interaction styles and analyzed by their effectiveness. As a result the manipulation of process variables turned out to be faster in traditional interaction manners while participants performed better under the reality-based condition in recalling their performed actions within the task scenario.



Nitsche, Markus

Dynamic Personal Spaces : Supporting Group Interactions Around Interactive Tabletops (2012)

This master thesis presents a technical solution that allows tracking the presence and location of multiple users around an interactive tabletop. Multiple use cases and scenarios will illustrate how this technology can be employed in order to improve existing interaction designs and also allow for new ones. Based on the concept of territoriality, the author introduces Dynamic Personal Spaces, which are virtual representations of a user’s workspace on a table. These spaces can be used to accomplish an automatic display partitioning, coordinate the multiuser process and support group work. Two studies have been conducted based on this approach. A long-term in-the-wild study reveals common distribution patterns of multiple users around a tabletop. An artificial experiment compares dynamic personal spaces with fixed ones in order to study what conditions underly user movements.



Öttl, Sonja

Visualisierungs- und Interaktionsdesign für multivariate, zeitabhängige Daten in sozialen Netzwerken (2008)

Die Analyse sozialer Netzwerke hat sich in den letzten Jahren zu einem Forschungsfeld von zunehmendem Interesse entwickelt. Untersucht wird, inwiefern und wie die einzelnen Akteure einer definierten Gruppe (soziales Netzwerk) miteinander interagieren und sich gegenseitig beeinflussen, um daraus generelle Einsichten in individuelle und gruppendynamische Handlungsweisen zu gewinnen. Temporale Entwicklungen spielen hierbei eine signifikante Rolle, da inhärente Prozesse und deren Auswirkungen in der Regel zeitabhängig sind.Um den Nutzer bei der Analyse sozialer Netzwerke zu unterstützen, werden ihm unterschiedliche Visualisierungen zur Verfügung gestellt. Ein Blick auf den Markt einschlägiger Systeme zeigt, dass das Potential, das der Einsatz von Visualisierungen für die Analyse sozialer Netzwerke mit sich bringt, momentan noch nicht ausgeschöpft wird. Dem entsprechend gilt es, Problemstellungen bei der Visualisierung von und Interaktion mit zeitbasierten Daten in sozialen Netzen zu konkretisieren und geeignete Methoden zu finden, um diesen entgegen zu wirken. Nach einer kurzen Einführung in die Thematik (Kapitel 1) werden im Rahmen folgender Arbeit zunächst theoretische Grundlagen mit Fokus auf datenspezifische Merkmale und daraus resultierende Nutzeranforderungen (Kapitel 2) dargelegt und der Markt nach adäquaten, vorhandenen Lösungen untersucht (Kapitel 3). Unter Einbezug der Rechercheergebnisse wird auf Basis einer Anforderungsanalyse (Kapitel 4) ein modulares Konzept zur Visualisierung von und der Interaktion mit multivariaten, zeitbezogenen Daten sozialer Netze entworfen (Kapitel 5). In einem weiteren Schritt wird dieses genauer evaluiert und überarbeitet (Kapitel 6), und in Kapitel 7 um einen Ausblick auf weitere potenzielle Ansätze zur Optimierung des entworfenen Prototypen ergänzt. Abschließend werden die entwickelten Bausteine kritisch reflektiert und die im Rahmen dieser Arbeit gewonnenen Erkenntnisse zusammengefasst (Kapitel 8).



Prestipino, Marco

Reifegradmodelle zur wissensorientierten Prozessverbesserung bei der Entwicklung gebrauchstauglicher Software (2002)

The ubiquity of technology in western societies requires high-quality software. The rising frequency of non-experts interacting with computers requires especially a high-quality user interface. Whereas quality has only been considered in terms of the product, focus is increasingly shifting towards the development process. How can usable products be created systematically and independent of individual developers? The thesis deals with process assessment and improvement. Maturity models describe known characteristics of good development processes and allow the assessment of existing processes. The thesis provides an overview of maturity scales for user-centred design processes. It establishes criteria for an analysis of two recent maturity models. As standards contain abstract descriptions of processes, their implementation requires selection of methods. The thesis shows that this task can not be left to textbook methodologies. Software development is still mainly based on experience. Thus, experience must be collected, analyzed and distributed. An investigation on existing approaches for experience exploitation reveals a simplistic view of organizational and individual learning processes. The thesis explains cultural and organizational aspects of experience exploitation. Several publications in the domain of software engineering support the validity of the presented ideas. These findings lead to the identification of weaknesses in capability scales and approaches for experience exploitation. Additions to existing maturity models are proposed in order to create a interdisciplinary approach to experience-based process improvement.

Rädle, Roman

Squidy: A Zoomable Design Environment for Natural User Interfaces (2010)

Today’s interaction between humans and computers does not significantly differ from the interaction that was provided through the early “personal computers” in the 1980s. Hence, keyboard and mouse are still predominant in standard computer products regarding the consumer market. Nevertheless, the popular gaming console Wii manufactured by Nintendo has proven the feasibility of new input device technologies employing full body interaction, which furthermore were exceedingly accepted by known gamers as well as casual users all over the world. This example shows the favor for new interaction techniques that feature a high usability as they are easy to use, easy to remember, and last but not less important enjoyable. In the next section, we briefly introduce the nature of such interfaces and interaction techniques by giving a history of user interfaces. We present an overview of, user interfaces like the first “command-line interface”, “graphical user interface”, and novel interfaces underlying the concept of a “natural user interface”. Such natural user interfaces offer an entirely free space to equip the human-computer interaction with further devices beyond keyboard and mouse. Nevertheless, interaction designers are not well supported by current toolkits and frameworks although the awareness of the necessity of a more natural interaction is widely accepted. The main contribution of this thesis is the provision of concepts for a design environment, Squidy, that supports interaction designers in the creation of more natural interaction techniques.



Rahn, Ariane

Utilizing a Virtual Environment to Measure Cognitive Load using Eye Tracking Technology (2019)

A valid measurement of Cognitive Load (CL) is of great interest in several areas, such as Human-Computer Interaction (HCI), education, medical training, aviation simulations, and the military industry. Task-Evoked Pupillary Responses (TEPRs) are a widely used method to measure CL. Especially the pupil dilation seems to be an indicator of cognitive processing. But TEPRs underlie a luminance bias, which is an empirical obstacle in CL research. This pioneer work used VR Technology to explore a new approach to better address this issue. The n-back task was used with three levels to induce CL: Low CL (0-back), medium CL (1-back), and high CL (2-back). An integrated eye tracker provided eye-related parameters. Furthermore, the impact of CL on the emotional state (SAM), perceived stress (PASA) and subjective CL (NASA TLX) was assessed.

Findings indicate a rather small effect of CL on TEPRs since CL only partly increased significantly with increasing task difficulty. The novel IPA calculation did not render any significance. But self-reported CL and performance metrics were highly sensitive to task difficulty. The impact of CL on perceived stress and the emotional state (Valence, Arousal, and Dominance) was rather small and only partly significant: Results indicate that CL has an impact on stress and emotional response, particularly when a high level of CL is induced.

Furthermore, a general pattern was found that confirms a successful manipulation of low CL(0-back) and high CL (2-back), but the 1-back condition seems to induce rather low than medium CL. The rather small effect of CL on pupil size change could indicate a common overestimation of the pupil dilation’s sensitivity to cognitive processing in the research field. Hence, future work is highly recommended to gain further insights about VR Technology as a promising novel approach in CL research.


Reeps, Inga

Joy-of-Use A new quality for interactive products (2004)

Since production processes have reached a level of sophistication where functions, safety, and costs are no longer competitive advantages, aesthetics and user experiences, so called Joy-of-Use, has become more than a stylistic after-thought to outpace competitors. This masters thesis introduces several aspects of the new topic of Joy-of-Use for interactive products. We analyse its integration with the disciplines of Design, Branding, Usability and Software Engineering, and it is discussed how the cooperation of these disciplines can contribute to successful Joy-of-Use. In the following theoretical, in particular psychological, background for Joy-of-Use is described. Quality criteria are identified as well as methods for their realization and the measurement of Joy-of-Use as felt by the user. After that critical opinions from the community’s actual discussion about Joy-of-Use are pictured and we present different approaches of design concepts for Joy-of-Use, e.g. conclusions drawn from game software. The work concludes with an outlook toward future prospects and required research.



Rieger, Martin

Novel Input Devices: Technophobia, Practice, and Acceptance (2009)

Despite the nowadays ubiquitous availability of technological devices they elicit discomfort and apprehension in some people. This may lead to the rejection of technologies which actually offer advantageous perspectives in the organization of everyday life. Especially novel input devices are at risk of being prematurely devaluated, if their mastery requires an initial learning process. In this study the relevance of anxiety, learning, and acceptance for the evaluation of input devices is being explored using a combined experimental, time-series, and correlational research design. Moreover, the existence of a technology-related anxiety construct, technophobia, is being scrutinized in context of a state-trait anxiety model. Performance data were gathered from 32 participants as they completed 7 practice sessions of a discrete multidirectional tapping test using a novel laserpointer input device on a large high resolution display. In addition, self-report measures of anxiety and acceptance were collected. Two new instruments are introduced: The Technology Anxiety Trait Subscale and the Fake Decision Test. The results indicate performance improvements according to the power law of practice - even though only short pauses were employed between practice sessions - and a potentially negative association between state-anxiety and performance. Initial support was found for a differentiation between the constructs of trait-technophobia and general trait-anxiety. Based on the implemented experimental manipulation, emphasizing the technological nature of the input device did not result in a significant reduction of test performance or an increase in state-anxiety. A supposed association between anxiety and acceptance was not found.


Rinn, Johannes

Methods and tool‐support for interdisciplinary requirements modeling and user interface specification (2008)

The development of graphical user interfaces is a key aspect in the creation of interactive systems, since their distribution has risen dramatically and natural interaction is increasingly becoming a crucial requirement. Therefore, not just the demands on the system to be developed augmented, but also the relevance of the underlying development process. A particularly important phase within this process is the requirements analysis. This phase is characterized as a cooperative process accompanied by communication and learning between people with different knowledge, experience and methods. Consequently, it is a complex task to specify a user interface unambiguously in an interdisciplinary context. Since formal, text-based and disciplinespecific specification documents harm mutual understanding among stakeholders, they tend to increase overall costs and time consumption. This thesis introduces a novel, lightweight approach to the specification of user interfaces. It respects the interdisciplinary nature of the requirement analysis by employing a simple structured process model and a reduced set of models in order to enhance mutual understanding. The theoretical foundations of this approach are based on an analytical consideration of formal, semi-formal and informal expression of participating disciplines. The core concepts of the developed specification method are based on a study of modeling approaches for user interface development, existing design and specification options and well-established interdisciplinary approaches. The result is a structured model, which relates an interdisciplinary consensus of models within multiple abstraction layers. The relationships between the identified design artifacts are highlighted to make the overall process traceable and understandable. Eventually, the theoretical approach is put into practice in conjunction with an innovative tool, which finally resembles an interactive visual specification as an alternative method for text-based documents. The thesis concludes with a summary on elaborated concepts, critical reviews and recommendations for further investigations.



Schmid, Patric

Improving data-gathering in field studies by using electronic devices (2011)

In the current years consumers spend more and more time using their electronic devices within the mobile context of their life where they are on the go. For this matter the researchers who try to observe the usage behaviours of those consumers also need to move from their laboratories into the field. To save resources and manpower during an observation researchers are now able to use modern Smartphone technology for contextlogging and querying experience samples. Those phones can also be used as multi modal diaries to gather more information on the subject’s experiences when they happen. This thesis describes the origin and first development of PocketBee - a multi modal ESM and diary tool for field research. While the first prototype of the system was already used in two pilot studies a re-design is proposed, which introduces a new generation of triggering framework as well as a user friendly graphical interface for the researcher. Both parts enable the researcher to configure studies using conditions and actions by putting them together into condition-chains. Prior to this thesis a state-of-the-art analysis was performed which led to multiple findings about existing ESM and diary tools which were already developed by researchers. In addition to this analysis and further theoretical research two field-studies were conducted using the pilot system of PocketBee which was already developed as a prototype for this purpose. As already mentioned this thesis provides two main contributions to the field:

  • The first contribution is the new architecture of the trigger-framework which represents the backbone of the system. It was designed to match requirements like combination of ESM, diary and logging studies and easy to extend with future sensors. The challenge here was to give researchers a tool for building complex concatenation chains of conditions and actions which then represent the aspired study design within the system. These condition-chains are as modular as possible, while simultaneously being as flexible as possible to combine several study designs.
  • The second contribution is a user interface including an interaction design which enables the researcher to create this complex condition-chains within an easy graphical context. Here the user can arrange and concatenate conditions and actions using the simple drag and drop technique. To furthermore extend the used Pipe/Filter concept the researcher is able to view details of a single conditionelement on demand by using semantic zooming which only shows the needed details for the current zoom level. The challenge here was to provide a concept and interaction technique which can also be used by non-computer experts from different backgrounds like psychology, sociology and sports.



Schnauber, Jochen

Empirische Untersuchung von interaktiven Werkzeugen zur Textverarbeitung (2013)

In the context of the project Blended Library the usage of new technologies in scientific libraries is being investigated. Work processes looked into are scientific investigation, acquisition of knowledge and text generation. This work is focusing on text generation. The process of writing scientific paper in today’s libraries is performed at desktop computers with the help of standard text processors. These text processors often act isolated from their environment and direct interaction between books and software is limited. Based on this situation an interactive tabletop is presented, which bridges the gap between real and digital world by providing reality-based interaction techniques. Therefor the work process in scientific libraries is being investigated. Based on the results a set of interaction techniques are developed. Finally a usability-test is conducted to validate these interaction techniques. This test shows that users consider a system designed by following these techniques enjoyable and natural. In addition the usability-test provides insight into possible directions of future development.


Schönhals, Alexander

FluidHaptics: Taktiles Feedback für Greif-Interaktionen in Mixed-Reality Anwendungen (2017)

This work introduces the development and implementation of a system that provides tactile feedback within virtual or augmented reality applications. The specific focus is on the realization of a tactile stimulus, which supports users in grasp interaction in virtual environments. Besides the necessary foundation with regard to the anatomy of the human hand and the grip taxonomies, different technologies for simulating haptic and tactile stimuli were taken into consideration. In addition, an overview of existing work in the literature, that realize or introduce approaches for identified needs is provided. Also, related systems are verified against predefined requirements. Based on this preliminary work an iteratively designed prototype is presented, followed by a description conducted user study. Finally, the fulfillment of the requirements is reviewed and an outlook for future work is described.


Schreiner, Mario

Extending Web Technologies to Enable Ad Hoc Cross-Device Interaction (2015)

When we look around us today, we see digital devices everywhere: The smartphone in our pocket, tablets and desktop computers on our office table, the laptop in our bag, or the smartwatch on our wrist. And still, while all these devices are interconnected and even share data, for example through cloud services, they are still, in a sense, isolated. Devices today are mostly unaware of all the devices that surround them – and although there is great potential in combining devices to use them in concert, this kind of interaction is rarely seen outside of research labs. Subject of this work is the definition of current obstacles in ad hoc cross-device interaction and to evaluate novel ways of solving them. This is attained through the development of a web-based prototype cross-device framework that includes concepts for a) creating multi-device web applications and b) enabling ad hoc combination of off-the-shelf devices in everyday scenarios. These concepts are evaluated, and the results will aid us in proposing new web standards that can help to proliferate cross-device interaction in everyday life.


Schubert, Sören

Auswirkung verschiedener Eingabegeräte auf das Erlernen räumlich-visueller Merkmale einer virtuellen Informationslandschaft (2010)

In recent years research in the field of input devices has attracted increasing attention in the IT-industry as well as in university research. It seems that in this context the question whether the emergence of new input devices holds additional benefits for the interaction with differentcomputerprograms is of particular importance. The touchscreen is a popular representative of such innovative input devices. In the scope of this thesis an experiment has been designed and conducted in which differences in the performances of touchscreen interaction and traditionalmouse interaction were evaluated in order to gain new insights.To find certain objects which provide access to the functionalities of different pieces of software and to find those objects again afterwards plays a significant role in the operation of computer programs. What is more, an explicit memory for the locations in which certain objects can be found is of importance. It was because of this fact that the navigation which test subjects carried out was observed and evaluated during a search task consisting of learning a spatial arrangement of objects. In an additional task the accuracy of explicit memory concerning the locations of the objects which had been investigated in the search task was captured. During the experiment it could be observed that under certain experimental conditions test subjects performed better when they used a touchscreen than when a mouse was used for the interaction. The performance was evaluated in terms of the initial finding and recovery of objects, as well as concerning the explicit memory for the objects’ locations. The result that touchscreen interaction outperformed mouseinteraction was significant if subjects were only allowed to issue panning commands in the interface during the search task. On the other hand, in an interface in which subjects could also issue zoom commands the results were not significant.In order to provide a broad understanding for the described experiment and the reported results, one part of this thesis is concerned with a detailed description of the experiment as well as with the considerations which have been taken into account during the development of the experiment and eventually led to the final design. Another part of this master thesis provides a theoretical background which played an important role during the planning and on the basis of which a close examination and interpretation of the results can be conducted.


Schwarz, Tobias

Konzeption eines Unterstützungswerkzeugs für die Gestaltung ergonomischer Benutzerschnittstellen am Beispiel der Dr. Ing. h.c. F. Porsche AG (2007)

As the development periods of service panels or service systems are getting shorter and as the stress of competition increases, the need of supporting tools for ergonomic Human-Interface-Devices becomes much more important. Pressure is also created by raised norms and standards for automobiles, which have to be considered by today’s developers. The aim of this thesis is to provide a module that supports the entire development process for service panels within the Porsche corporation. At present, existing tools cover the configuration of service panels only partly and not through the entire process of development, as a detailed analysis has shown.
Hence a concept for the supportive module PoCKET (Porsche Consulting and Knowledge Engineering Tool) was developed within this thesis. It has been integrated in a tool-support chain, which is elaborately described in Bock (2007) and Heilig (2007). On the one hand PoCKET is a system that supports developers during the configuration of service panels in the dimensions interface, information and interaction design. On the other hand it provides a knowledge management system in which collaborative evolved information can be stored and accessed.
In order to develop a system that perfectly fits the HMI development process of Porsche, a in depth analysis of the development process has been accomplished. An interview in type of a questionnaire has been conducted quantitative as well as qualitative to determine the developer’s requirements as detailed as possible. Therefore these results have been the initial point for an adequate configuration of the new supportive module. The consideration of existing tools for the configuration of ergonomic user interfaces has shown that these do not fit the particular industrial requirements of Porsche. For that reason they can only contribute as a connecting factor to this thesis. The aim of this thesis is to provide a component, which supports the HMI configuration process of the Porsche corporation entirely. Hereby the focus of this work lies on the support of specific process periods as well as the employment incrossproject contexts. The surplus of the supportive module PoCKET is shown by considering the DATech testing manual for usability development processes. Thus an increasing quality of the HMI development process can be justified by the application of this module. 



Schweitzer, Daniel

VITT — Virtual Transfer Trainer: Design and Evaluation of a Virtual Reality System to Support the Self-directed Training of Ergonomic Patient Transfers

There is a high prevalence of musculoskeletal disorders among medical care professionals, in parts resulting from frequent patient transfers. The so called kinaesthetics care conception is a movement system to promote ergonomic patient-transfers, that both support the caregiver as well as the patient. The educational support structure for kinaesthetics is very limited in Germany, as only limited resources are allocated. This issue extends into, and is afflicted by, a general shortage of medical care professionals in Germany: there are not enough nurses, which affects the quality of work and their own health. This also impacts usage of kinaesthetics during every-day work. There are not enough instructors, which affects the education of students and potential re-education. The last part of this vicious cycle is a high number of students, who abandon this education as a result. A virtual reality learning application may help nursing students to partially overcome these difficulties by providing a readily available and immersive training environment, without the need for instructors or training partners. Research into existing work regarding virtual reality, immersion and learning effects has been done, which led into the creation of a concept for such an application. It has been iteratively created, making use of information gathered via concept walkthroughs from a kinaesthetics instructor as well as three experienced nurses of varying ages. Finally the concept has been iteratively implemented and partially evaluated. A pilot study has been executed, revealing well-functioning as well as critical areas of the application. Within the limits of a pilot, results were mostly encouraging for a theoretical full-fledged usability study, though one important area has shown lacking robustness. Based on this, sensible changes and additions have been made to the implementation.


Söter, Andrea

Theorie und Praxis von Benutzerbefragungen, Benutzungsstudien und Benutzbarkeitstests in der Automobilindustrie (2009)

In order to develop automotive systems, it is necessary to consider certain general constraints which arise from a special context of application different from desktop PCs and therefore pose special challenges to concept design, development, and evaluation. As an introduction to the subject of the thesis, these general constraints as well as the appropriate research questions will be explained. Subsequently, conventional methods of consumer surveys, usage studies and usability tests pertaining to human-computer interaction will be introduced, and the applicability, both theoretical and actual, of said methods to the automotive industry will be discussed. An exemplary lifecycle will show the practical application of theoretical methods. The focus will be on the data collection and evaluation techniques used to plan, conduct, and analyse two evaluation studies in the context of the development process of an automotive internet browser.


Stasche, Anton

Combining pointing and gestures: Novel interaction concepts for large high-resolution displays (2008)

Interaction with large high-resolution displays offers the embracing of alternative input modalities due to the users’ increased freedom of movement. As opposed to the standard mouse-keyboard combination, pointing devices such as laserpointers that possess motion-sensing capabilities can provide the advantage of integrating direct pointing with the eliciting of discrete commands to the system via gesturing. In the Human-Computer Interaction (HCI) domain, gestural interfaces have been largely developed and investigated using computer-vision and glove-based approaches. However, most of these findings do not regard the type of gestures applicable for the handheld motionsensing device employed in the current study. This Master thesis explores the theoretical background of gestures for the sake of establishing a coherent classification of device gestures in HCI. These are then placed firmly into a framework for gestural interaction by augmenting the work of Karam (2006). Results and suggestions from a series of semistructured interviews inform the creation of two collections of gestures for the context of electronic mind mapping performed on the Powerwall, a large high-resolution display at the University of Konstanz. Conclusions from a formal evaluation of the overall usability and learnability of two final gesture sets suggest the promise of using gestures as an additional input channel for large display interaction.


Stößer, Gloria

Blended Museum – eine didaktische Begründung des Einsatzes vonInformations- und Kommunikationstechnologienim Hinblick auf die Informationsverarbeitung und –strukturierung (2009)

Museen haben „auf Änderungen in der Sozialstruktur und Lebenshaltung der Gesellschaft ebenso zu reagieren wie auf die Entwicklung der Informationstechnik“ (WAIDACHER, 1993, S. 27). Traditionell steht im Museum das reale Objekt im Vordergrund. Aufgrund gesellschaftlicher Veränderungen, neuer technischer Möglichkeiten sowie veränderten Erwartungen der Besucher wurden die Museen in den letzen Jahren umgestaltet. Vermehrt werden im Museum „neue Medien“ zur Informationsvermittlung eingesetzt, um auf die sich entwickelnde Informationskultur zu reagieren. Auch „Wissenschaftler und Museologen sehen die Hauptfunktion des Museums nicht mehr nur im Bewahren und Vermitteln seiner Objekte, sondern vielmehr im Bewahren und Vermitteln der Informationen über diese Objekte“ (SCHWEIBENZ, 2008b, S. 12).

Allerdings werden Museen durch die rasante Entwicklung im Bereich der Informations- und Kommunikationstechnologien (IKT) vor große Herausforderungen gestellt. Die zunehmende Digitalisierung von Informationen scheint nur schwer vereinbar mit den realen physischen Objekten im Museum zu sein. Das Museum, als ursprüngliche Institution zum Bewahren von Objekten und zum Vermitteln von Wissen, steht in der heutigen Informationsgesellschaft demnach einem Konflikt zwischen den realen Objekten und der zunehmenden Virtualisierung der Informationen gegenüber (SCHWEIBENZ, 2008b, S. 11). Genau dieses vermeintliche Spannungsfeld soll jedoch durch eine sinnvolle Verbindung beider Perspektiven zu lösen versucht werden.

Mit dem Konzept des Blendend Museum wird ein realer sowie virtueller Zugriff auf Museumsinhalte ermöglicht. Der Besuch der Institution Museum muss mittlerweile nicht mehr abhängig von Standort und Öffnungszeiten sein. Durch die Informations- und Kommunikationstechnologien und dem Medium Internet entwickeln sich neue Möglichkeiten und Perspektiven für die Wissensvermittlung im realen sowie im virtuellen Museum. Die Herausforderung für Museen besteht nun darin, durch einen sinnvollen Einsatz von Medien bzw. Informations- und Kommunikationstechnologien, nicht nur Erlebnischancen zu bieten, sondern auch das Lernen im Museum zu unterstützten.


Tennie, Oke

Midas – Ein Toolkit für das Erkennen und Analysieren von Gesten (2016)

In this thesis, a software technical solution for the identification of hand and arm gestures is being presented. The real-time detection used is based on sensors that record the movement of the arm and the hand. The hardware used to record these gestures is the so called Myo bracelet. After the data is being recorded, it is sent to an implemented Windows application named Midas, where the actual gesture recognition is calculated. Before Midas can recognize any gestures, it is necessary, for the user to train some gestures for the system, in order to make sure, that the used algorithm knows how the gesture looks like. To find the right algorithm, many approaches have been tested and analyzed. As to ensure the flexibility of the data pipeline, Midas was designed as Node Designer. This allows the user to model and test different algorithms, using the machine-learning pipeline. Finally, one came to the conclusion that the dynamic time warping algorithm works best for the task of a gesture recognition.

With this knowledge, two scenarios were implemented in order to test the dynamic time warping algorithm in detail and to verify its suitability. The first scenario contains a touch gesture recognition. The implemented system, called Midas Touch, allows an assignment of a touch interaction on a touch-sensitive display to the executing user. The second scenario enables the user to translate any gesture into spoken language. The particularly therefore developed system Gvoice, implements the training and the testing of gestures, in the most possible user friendly way. Both systems are supported by the Midas Node Designer, that runs in the background. This enables the user to change the used pipeline in real time in order to test new approaches. The two developed scenarios - Midas Touch and Gvoice only touch the surface of the possibilities Midas can offer. For this reason, a model of this system was created to illustrate its potential. This model is based on the Buxton State Model (Buxton 1990).

Subsequently, both scenarios were evaluated with users in two studies. The evaluation of Midas Touch showed, that the average detection rate of touch gesture accuracy is 88%. The system Gvoice identified different gestures with an 85% probability correctly. On average, twelve gestures per subject were tested. This result indicates that a gesture recognition using sensors, in combination with the dynamic time warping algorithm, lies within the scope of feasibilities.


Von Bauer, Philipp

Cognitive State Classification Using Psychophysiological Measures (2018)

Assessing users’ cognitive state is one of the visions in Human-Computer Interaction (HCI). On the one hand it would allow the building of intelligent adaptive systems and on the other hand, could serve as an in-place evaluation tool of workload in contrast to the current practice of using questionnaires post hoc. Building adaptive human-computer systems is a challenging task. This work addresses one aspect of it by investigating the use of psychophysiological measures -pupil diameter, heart rate and skin conductance - as input for supervised machine learning to classify the cognitive state of users. From 24 subjects data was collected during the n-back task - an abstract working memory task. A Random Forest Classifier (RFC) was trained across users using statistical features of the pupil size.Classification accuracies reached up to 89% in discriminating between the 1-backand 2-back task with a window size of 60 seconds. The other physiological measures were not sensitive to the task’s manipulation, thus, left out for classification.

Cross-user classification yielded promising results for the n-back task with letters.In a second explorative small-scale study, six participants performed the n-back task with three new stimuli types instead of letters - spatial, audio and images. The attempt to use the classifier of the first study for cross-task classification showed promising results with accuracies between 79% and 94%.

The cross-user and cross-task classification performed well showing the feasibility of using pupil measures for classification. Accounting for individual differences needs to be addressed further along with a methodology to evaluate classifier performance beyond using task difficulty which cannot be reliably and equally created across tasks.

Some related work investigated the use of EEG for cross-task and cross-user classification. However, most focus was on building individual models. Thus, this work investigated the use of the pupil diameter for cross-task and cross-user classification using the n-back task and variations of it as related work did for EEG.



Weber, Rebecca

Egocentric Motion Guidance in Mixed Reality - A comparison of visualizations guiding bilateral arm movements (2018)

Support through motion guidance is important for many fields in which conscious motor learning takes place. Traditional ways to learn a movement can be broadened by systems which make use of current Mixed Reality technologies. They enable motion guidance from an egocentric perspective.

This thesis presents an egocentric motion guidance system in which different visualizations have been implemented to guide the user’s bilateral arm movements from an egocentric perspective. The visualizations differ in their appearance as well as in the way the movement is visualized. In a study, these different visualizations are compared with respect to their effect on performance and user experience and preference.

The results of the study showed only few statistically significant differences between the movement visualizations which do not produce a profound outcome. This might be caused by several aspects influencing the measurements, e.g., the selection of movement tasks. However, results showed a statistically significant difference between the different ways of appearance indicating that realistic looking visualizations result in a better performance of the movement with guidance as well as in the repetition of the movement without guidance.



Wieland, Jonathan

Comparing Different Approaches for the Collaborative Manipulation of Virtual 3D Objects in Augmented Reality (2019)

Applications on hand-held Augmented Reality devices can create interactive experiences that enable users to see virtual objects seamlessly integrated into their physical environment. One use case for such applications is interior design, which requires users to place virtual furniture in their homes, relying on 3D object manipulation. Although furnishing a place is typically a collaborative task, IKEA only offers a single-user AR application. Besides, different studies show that a collaborative manipulation of virtual 3D objects can be more efficient and increase user experience. Literature analysis revealed three different approaches: 1) Separation of DOFs separates the manipulation of the degrees of freedom of an object between users, 2) Composition of Users’ Action implements a merge policy to combine users’ inputs to one composed manipulation, and 3) Hybrid Approaches combine both variants. However, based on related work, it is difficult to decide which of the three approaches one should use as they have not been compared for hand-held AR devices yet.

Consequently, this thesis addresses the question of which of the three approaches a hand-held AR application should implement that enables two users to furnish their homes collaboratively. Therefore, a study prototype was developed that implements all three approaches. In a controlled lab study with 48 participants, the prototype was then used to compare the three approaches in terms of performance, workload, and user experience. While the measures for performance and workload suggest using the implementation for Separation of DOFs, participants’ preferences were polarized between the implementations for Separation of DOFs and the Hybrid Approach. Based on the results of the experimental comparison and their discussion, this thesis draws implications for future applications and highlights further research questions to be addressed by future work.


Zagermann, Johannes

Is web technology ready for NUI? Evaluation of a cross-browser and multi-display system for collaborative web search (2015)

Collaborative group tasks such as searching, organizing, or problem solving in general are usually facilitated by shared group spaces. For these spaces, tables are often used be- cause of their physical affordances that are essential for group work activities (e.g., allow deictic references or face-to-face collaboration). Therefore, in CSCW, interactive tabletops have been part of various studies around group work activities. However, the influence of tabletop size on such activities has not yet been researched.

The TwisterSearch system was used as a tool to analyze the effects and influences of tabletop size on group work activities. Groups can work collaboratively on search tasks facilitated by personal devices to enable individual search and reading activities as well as a shared group device that allows to structure and share information in an around-the- table situation.

In an exploratory study the three different tabletop sizes 10.6′′, 27′′, and 55′′ were used as shared group spaces in a between-subjects design. The focus of the evaluation lay on groups’ activities as well as their communication and collaboration behaviors. The results reveal differences in communication and collaboration behaviors, as for example groups working on the midsize tabletop tend to discuss longer and more frequently. Both larger tabletops invite their users to work more collaboratively and thereby structure information on the shared group space more actively. The smallest tabletop lets groups rather evolve different roles. These trends are reflected in the outcome as groups working on both larg- er tabletops show higher individual and group scores.



Zimmermann, Tobias

Augmented Laying Board - Design and Evaluation of a System to Support Wire Harness Assembly using Augmented Reality (2021)

In appliances and machines, the cabling is usually not installed one by one, but assembled in advance, bundled together and fitted in as one piece. These bundles are called Wire Harnesses. They provide multiple benefits, for example a simpler installation process or the ability to standardize their production. For their assembly, the workers at Engeser, a company manufacturing wire harnesses, employ so-called Laying Boards, upon which the cables are laid out, and printed instructions that tell them which cables to lay out where. As the Wire Harnesses are still hand-made, despite the advances in automation, the company sought ways to support their workers to increase their efficiency and to improve the training process for new workers. For this reason, this thesis explores how a system can be designed to support workers during the assembly process utilizing Augmented Reality.

To guide the development, the Wheel UX Lifecylce has been employed in this thesis. In accordance with it, the first part of this thesis explores the context of the application, the work environment as well as the requirements that are placed on such a system. Together with the analysis of related work for this topic, a design was then created from the results. As the next step, a prototype was then developed based on this design. The prototype, called Augmented Laying Board, uses the Microsoft HoloLens 2 to project a digital overlay directly onto a laying board. This overlay displays the positions and route for a specific cable as given by assembly instructions. To evaluate the prototype, an ”in-the-wild” study has been conducted at Engeser. The study highlighted the potential benefits a system like Augmented Laying Board can provide to support the workers and improve their training. Based on the findings of the study, a set of improvements and further evaluation directions were proposed.


Zöllner, Michael

A state machine framework for Post-WIMP interaction design (2012)

After years of dominance, classic desktop-based WIMP (Windows, Icons, Menus, Pointer) systems are slowly being replaced by modern post-WIMP systems. Such systems do not stick to a certain user interface or interaction paradigm, but rather contain a heterogeneous set of characteristics that stem from multiple fields of research. These characteristics induce a variety of different challenges that designers and developers of post-WIMP systems have to face and overcome. In this thesis, it is argued that formal methods, and in particular finite-state machines, are an important means to tackle certain of these challenges. In order to adapt finite-state machines to the requirements of post-WIMP systems and to improve their expressivity, specific additions to their default notation are suggested. One such addition allows the specification of animated transitions. The other addition is a notation to differentiate multiple input points. Although there is a consensus between many researchers and developers that finite-state machines are a rather natural formalism for the specification of complex interactive systems, their implementation is not yet supported appropriately by user interface toolkits or programming languages. Thus, a finite-state machine framework for post-WIMP interaction design, the Reactive State Machine framework, is presented in the main part of this thesis. Due to its declarative nature, it greatly facilitates the transformation of a graphical state machine model into code. The Reactive State Machine framework supports all important state machine concepts, such as states and transitions. What sets it apart from other similar frameworks is its full support for input events, its support for animated transitions and its support for the multi-point notation that is introduced in this thesis. To show the utility and value of the Reactive State Machine Framework its application in three assorted use cases is demonstrate. For one of the use cases, the Facet-Streams system, a comparison is conducted between the old naive implementation based on low-level implementation techniques and the revised implementation based on the Reactive State Machine framework. Finally, the threshold and ceiling of the Reactive State Machine framework are assessed in a brief informal evaluation. To conclude the thesis, its main contributions are summarized and an outlook on potential future work is given.