Towards interacting with smarter systems

Abstract

The aim of this article is the requirement analysis, design, proof-of-concept implementation and user evaluation of a novel user interface that is specifically developed to support interaction with ambient intelligence systems. The focus of the user interface is rather on interacting with ambient intelligence, a meta system that comes into existence as the result of the coherent behavior and collective intelligence of interconnected smart devices and systems of a domain such as meeting rooms or smart homes. To design the user interface, we conducted an empirical user and task analysis and elaborated requirements, defining the scope of human interaction with ambient intelligence. The captured requirements shed light on the characteristics of such a user interface and lay the foundations for describing a meta system architecture. As our major contribution, we provide an interaction model formally specifying human–meta system interaction. Based on this specification, we have designed and developed basic affordances of the specified meta user interface such as overriding default system behavior, monitoring and predicting past, current, and future automations, or exploring the possible effects of a certain type of automation. The prototype and implementation of the meta-UI is then assessed through a Cognitive Walkthrough and a formative user evaluation. The result of the evaluation indicates that the designed meta-UI can increase usability of the human-meta system interaction.