Emergent Interaction –
A Pre-study

Niklas Andersson -Niklas.Andersson@dh.umu.se
Anders Broberg -bopspe@cs.umu.se
Agneta Bränberg -Agneta.Branberg@tfe.umu.se
Lars-Erik Janlert -lej@cs.umu.se
Erik Jonsson -Erik.J.Jonsson@epl.ericsson.se
Kenneth Holmlund -holmlund@hpc2n.umu.se
Jonny Pettersson -jonny@cs.umu.se

UMINF 01.16
DEPARTMENT OF COMPUTING SCIENCE
SE-901 87
SWEDEN

Abstract

This is the final report from a pre-study on Emergent Interaction. The purpose of the pre-study was to get a clearer idea of what emergent interaction is, what it can be used for, and what the problems are, as a preparation for a larger study. The project is an Umeå Center for Interaction Technology (UCIT) and Ericsson Erisoft collaboration, with participants from four UCIT labs (CCL, IDL, DML and VR-lab) and Ericsson Erisoft Research. It represents the first step of a UCIT research program on emergent interaction.
An Emergent Interaction Systems consists of an environment in which a number of individual actors share some experience/phenomenon. Data originating from the actors and their behaviour is collected, transformed and fed back into the environment. The defining requirement of emergent interaction is that this feedback has some noticeable and interesting effect on the behaviour of the individuals and the collective – that something ‘emerges’ in the interactions between the individuals, the collective, and the shared phenomenon as a result of introducing the feedback mechanism.
The immediate effect may be enhancement of the individual experience – with resulting effects on the individual’s behaviour, choice of action, and so on. The immediate effect can also be some kind of change in the observed, shared phenomenon. In particular the feedback might effect or establish some kind of collective control. The effect could also involve some kind of organizing and controlling of the collective. ‘Organization’, in this case, need not imply uniformity and regularity, it could just as well be to diversify or even randomise behaviours.
Systems in which people interact in a shared feedback loop already exist. What is new with the emergent interaction research program is a unified approach to such systems, and an agenda that seriously addresses the task of designing EIS. This is made possible and necessary by the new information technology. Computer, communication, and interface technologies crucially change the conditions and possibilities. First, the amount and variety of data that is possible to collect, and the speed of collection increase radically. Second, the new information technology offers completely new possibilities to design and control the feedback function and thus ultimately the behaviour of such systems. Third, the feedback loops can be speeded up many orders of magnitude to match the ‘natural’ time scales of individual and collective behaviour, thus also making the existence and importance of such systems more easily recognisable. Fourth, in this new time scale, with these new capabilities, there are great opportunities as well as possible hazards that we so far only can guess about.
A number of pre-study activities are summarised in the report. Some of the concrete results are: a list of focus areas for emergent interaction; a categorisation of emergent interaction applications into nine categories; a classification of four different aspects of emergent interaction applications; a list of prototype requirements.
This report will serve as a basic framework for the continued study of emergent interaction within the UCIT emergent interaction research program. A number of suggestions for the next actions within this program are also given. First, three possible first prototype implementations (spinning, collaborative art, campus communities) are suggested. They cover a number of aspects that have high priority for closer investigation. Second, the EI concept should be established in the scientific world as well as in the commercial world. Third, we have identified a number of research issues to study and ideas to develop (Emergent Architecture, Emergent Architecture Protocol, Emergent Design are some) within the emergent interaction field. Other important activities, closely related to but outside of the proper research program, are market studies and work to g et external financing for the program.