Distributed interaction

The personal computer as used by most people still to a large degree follows an interaction and technological design dating back to Allan Kay's Dynabook and the Xerox Star. This implies that interaction is confined to a single device with a single keyboard/mouse/display hardware configuration sitting on a desk, and personal rather than collaborative work is in focus. The challenges of "moving the computer beyond the desktop" are being addressed within different research fields. For example, Ubiquitous Computing (Ubicomp) investigates how computing can be embedded in everyday life; Computer Supported Cooperative Work (CSCW) researches collaborative interaction; and many researchers in the CHI and EICS community explores basic infrastructure and technologies for handling multiple devices and displays in e.g. smart room setups. In this talk, I will present our approach to these challenges. Specifically, I will introduce the term of "distributed interaction," which is a research agenda focusing on researching theory, conceptual frameworks, interaction design, user interfaces, and infrastructure that allow interaction with computers to be distributed along three dimension: Devices -- computers should not be viewed as single device but as (inter)networked devices. Hence, interaction is not confined to one device, but should encompass multiple devices. Space -- computers are distributed in space and time, and are not confined to one setting. This includes mobility, but more importantly that devices are to be found in all sorts of odd settings where they need to adapt to, and collaborate with, their surroundings, including other devices, people, interaction devices, etc. People -- computers are to a large degree the primary way of collaboration in distributed organizations. Hence, a lot has changed since the personal computer was designed for small office collaboration and there is a need for incorporating support for global interaction as a fundamental mechanism in the computing platforms. I will present our current approach for supporting distributed interaction called "activity-based computing" (ABC). Based on a strong theoretical foothold in Activity Theory, ABC provides a conceptual framework, interaction design, user interface, and a distributed programming and runtime infrastructure for distributed interaction. I will present ABC and show how it has been applied in building support for clinical work in hospitals and for smart space technology.