UC Santa Barbara

Augmented Reality (AR) is an interactive technology that delivers rich and immersive computer-generated perceptual information overlaid onto the real world. AR technology continues to improve over time, potentially enabling an "always-on" AR future, where wearable AR devices are as comfortable to wear as glasses and have an all-day battery life. This concept has been termed "Pervasive Augmented Reality", in which AR is the predominant form of personal computing, instantly accessible, constantly providing information and available to assist in everyday tasks.

Pervasive AR represents a shift in how, when, and where we use computers, but there are few guidelines for how to effectively create these types of systems. Current research focuses on single-purpose use cases, designed for specific tasks and environments, such as navigation or maintenance. They rarely consider AR as a mobile experience that can be taken anywhere, where computing happens continuously and persistently. This dissertation addresses that insufficiency by analyzing the design and evaluation of AR artifacts for 3 use cases: AR Recommender Systems, AR Language Learning, and AR Multitasking. We conduct our investigation in 3 parts. 1) What benefits do AR applications provide over non-AR based systems? 2) What additional inputs and signals can AR applications benefit from? and 3) How do we evolve current AR systems towards a pervasive AR future? Our work makes contributions through empirical user studies, prototype systems, and the development of new interaction techniques, revealing insights into the tools and techniques needed for developing pervasive AR, as well as the opportunities and new possibilities it enables over existing technologies.