Designing UbiComp Experiences for Spatial Navigation and Cross-Device Interactions

We are witnessing a considerable growth in number and density of powerful mobile devices around us. Such devices like smartphones and tablets are our everyday companions. If not already at hand, they often wait in our bags and pockets to provide us with a ubiquitous computing (UbiComp) experience. However, most of these devices are still blind to the presence of other devices and performing tasks among them is usually tedious due to the lack of guiding principles.This thesis closes with the gap as mentioned above by investigating in the design and evaluation of spatial and cross-device interactions. As a central theme, presented research fundamentally grounds on embodied practices by exploiting users' pre-existing practical knowledge of everyday life for human-computer interaction. These embodied practices are often applied subconsciously in our daily activities, which unfolds new - yet unexplored potentials for fun and joyful UbiComp experiences.Within this thesis, research is approached through both deductive and inductive reasoning. It begins with a brief history of UbiComp and its overarching vision. Contradicting opinions on this vision are discussed before leading over to recent theories and believes on embodied cognition and models on human spatial memory. This theoretical background eventually thrives arguments for yet unexplored and hidden potentials for spatial and cross-device interactions. Then, the application domain is narrowed down to academic libraries and knowledge work activities. In field studies at the Library of the University of Konstanz, the following two main knowledge work activities, and resonating issues are identified: literature & bibliographic search and reading & writing across documents. Together with the theoretical background the found issues are transformed into potentials for future knowledge work. Thereby, two fully functional research prototypes, Blended Shelf and Integrative Workplace, were implemented to explore the problem space further and to derive research questions covered in this thesis.The research questions are tackled through controlled experiments and implementation of low-cost enabling technology. In the first experiment, the optimal size of a spatially-aware peephole display is studied. As a finding, a relatively small tablet-sized peephole display serves as "sweet-spot" between navigation performance, subjective workload, and user preference. Within the second experiment, peephole navigation is contrasted with traditional multi-touch navigation. The findings indicate that users prefer physical peephole navigation over multi-touch navigation. It also leads to better navigation trajectories, shortens task-completion-time, and hints for longer retention of object identities as well as their locations in human spatial memory. Due to the lack of appropriate technology HuddleLamp was developed in an intermediary step. HuddleLamp is a low-cost sensing technology that tracks multiple mobile devices on a table. It allows implementing spatial and cross-device applications without the need to instrument rooms, equip devices with markers or install additional software on them. This technology is used in the third experiment to understand subtleties of cross-device interactions. Findings show that, for cross-device object-movement tasks, users prefer spatially-aware interactions over spatially-agnostic interactions.Apart from individual findings, this thesis contributes a summary and integration of all findings to general design guidelines for future spatial and cross-device applications. Eventually, these guidelines are applied by researchers and practitioners to develop UbiComp experiences that increase users' task performance, lower their individual workloads such as mental demand, effort, and frustration. At the same time, these guidelines lead to an increase of the cumulative value when working with multiple mobile devices.publishe