PUB - Publikationen an der Universität Bielefeld

Augmented Reality (AR) based assistance has a huge potential, especially in the context of Industry 4.0. AR links digital information to physical objects and processes in a mobile and, in the case of AR glasses, hands-free way. AR can replace and optimize paper-based instructions, guide to relevant locations and at the same time automatically accomplish documentation.

AR-based assistance usually comprises two steps: Guiding towards a location where actions shall be conducted and instructing the user to execute these actions. This dissertation puts the focus on the first part. The initial guiding phase is the basis for engineering AR-based assistance systems. Thus, designing such systems depends on profound systematic knowledge on the topic of AR-based guiding.

An initial systematic literature analysis reveals a number of approaches for guiding users in different contexts. However, these were mostly developed for specific scenarios and compared to analogue forms of guiding. The two main types of target scenarios in the literature are workbench and picking tasks: In workbench scenarios, relevant locations are distributed in front or around the user and orienting towards a target usually suffices. Picking scenarios usually consist of more complex environments where users have to navigate towards their target.

Starting from this basis, this dissertation contributes a structured, systematic view on AR-based guiding. To achieve this goal, a taxonomy of guiding techniques is developed in the context of different requirements in form of scenarios, tasks, user groups and investment. Five evaluation studies cover all relevant scenario types scaling from workbenches to picking in large environments. In these contexts, existing and newly developed guiding techniques for specifically matching requirements of the different scenario types are systematically compared.

Based on the guiding techniques identified in the review, the initial evaluation compares basic and well-known approaches for workbench-size scenarios. Building on the insights gained, new guiding techniques optimized to the requirements of such scenarios are proposed and evaluated in two studies. Then, additional modalities are considered for enhancing guiding techniques. A focus is put on eye-tracking-based guiding in the visual periphery. Finally, more complex and larger environments are addressed. Path-based guiding techniques supporting navigation towards a target are developed and evaluated in two studies.

An additional contribution of the dissertation is a framework for simulating AR in Virtual Reality (VR) which guarantees reproducible results in standardized environments. This way, guiding techniques can be evaluated on different simulated AR devices. This allows to integrate and test technologies like eye-tracking for usage in VR and AR HMDs. In the context of guiding in small workspaces, this enabled the evaluation of peripheral guiding techniques which would not have been possible using available standard hardware.