Interaktion in Ambient Intelligence
According to the vision of Ambient Intelligence (AmI), our everyday environment and its objects will be pervaded by sensing, computing and communication capabilities. A major characteristic of such environments is the increasing amount of intelligent devices and their complexity; Computers will be ubiquitous. As household appliances grow in complexity and sophistication, they become harder and harder to use, particularly because of their tiny display screens and limited keyboards. At the same time, devices will disappear into the background and will be invisible to the user. With the emergence of newly available technology, the challenge to maintain control increases, while the additional value decreases. After taking a closer look at Ambient Intelligence environments (AmI-E), there will come up the (general) question of how to build a more intuitive way for people to interact with such an environment. This thesis discusses such challenges of interacting with complex, disappearing, and adaptive environments (e.g., such as AmI-E). It discusses existing research focusing on some important challenges of Human-Environment-Interaction. Based on this discussion, the present thesis shows major advantages and weaknesses of explicit interaction vs. implicit interaction. As a result, it will motivate to follow a mixed-initiative approach to interact with Ambient Intelligence environments. In the main part of this thesis, an interaction model is presented that is designed to overcome identified interaction challenges. Following on that model proof-of-concept implementations are presented. In final chapters of this work, the author presents user evaluations to validate the developed interaction models. The author argues that his approach increases user control over adaptive environments. This is achieved by using a mobile assistant which provides intuitive and explicit access to AmI-E thus allowing always the user to stay in control. Another benefit is that using intuitive metaphors and conflict management mechanisms, the user can define and restrict the behaviour of the implicit actions initiated by the AmI-E. This allows the user to avoid, stop, or undo inappropriate adaptivity, which will increase the reliability of the overall system. Allowing the user to remain control and increasing the reliability of the AmI-E will increase trust in automated system thus motivating the user to accept and use the technology.