No earlier issue with same topic
Issue
Previous article
Article
SIGCHI Bulletin
Vol.30 No.4, October 1998
Next article
Article
No later issue with same topic
Issue

CVE'98

Collaborative Virtual Environments: A Conference Report

Dave Snowdon, Elizabeth F. Churchill

Introduction

A Collaborative Virtual Environment (CVE) is an application that uses a Virtual Environment to support human-human and human-system communication. Within such virtual environments, multiple users can convene, communicate and collaborate. Virtual Environments vary in their representational richness; for example environments may be graphical (2D or 3D) or may be purely text-based as with many MUDs/MOOs. Users are often represented by embodiments (or "avatars"), which also can vary from rich moving characters (such as in the Massive environment [1] or in Microcosm [2]) or simple 2D static cartoon images (as in The Palace, [3]).

The design, development and use of CVEs is a growing research area with much potential for inter-disciplinary collaboration between researchers in the fields of computer science, psychology, sociology, cultural & media studies, architecture & urban planning, Artificial Intelligence, human computer interaction and CSCW.

The Second International Conference dedicated to the design, development and evaluation of such Collaborative Virtual Environments (CVE'98) recently took place in Manchester. CVE'98 was part of Manchester's Digital Summer, celebrating the 50th Anniversary of the "Birth of the Baby": the demonstration of the first stored computer program. The conference took place over two days, from the 17th until the 19th June in the MANDEC Centre on the University of Manchester campus, and attracted 95 attendees from all over the world, drawn from computer science, human factors, sociology, psychology, graphic arts and other disciplines.

Workshop and Technical Program

CVE'98 started off with a one-day workshop led by Michael Capps from MIT on systems aspects of CVEs.

The technical program for CVE'98 was single track, over two days with 15 presentations. Discussion sessions also took place over lunch each day. Some of the highlights of CVE'98 are discussed below.

Keynote Speech

The keynote speech was given by Adrian West and Roger Hubbold from the Advanced Interfaces Group at Manchester University. Their presentation focused on the design of the `E' in CVE, highlighting some of the current research issues such as handling complex worlds, providing programmer-friendly techniques for specifying complex behaviours and navigating in complex environments such as CAD models of oil-rig platforms. Consideration was given to design challenges for CVEs such as support for large numbers of users, the embedding of applications within different worlds and the embedding of worlds within worlds. A central issue was use, considering what people will do within such environments after they say "hello". Finally, the disjunction between the richness of the `real' world and that of the virtual world was underscored. New metaphors and ways of exploiting the power of virtual environments were called for, and the challenges for underlying architectures raised.

Technical program

The intention behind CVE'98 was to cover a wide range of issues from different disciplines and this was reflected in the mix of papers presented. 3D applications for virtual environments were well represented. Gunter Bellaire (Robert-Rössle-Klinik, Germany) described a high-end projected VR system for supporting visualisation of CT scans. Such visualisations enable the reconstruction of segmented data from multiple scans in support of collaborative exploration and analyses which lead to the diagnosis of tumors. Laurie McCarthy (Lockheed Martin, USA) described a virtual environment designed to support the training teams to use the engine control rooms of US navy ships. This application used an intelligent, Soar-based, embodied agent, Steve, in the VE to act as a mentor to the trainees. Steve exists within the virtual task environment and is capable of taking the role of a team member, leading exploration of the environment and responding to spoken questions. Help is provided help via speech and gesture. James Pycock (Xerox Research Centre Europe, UK) described how a virtual environment was created to illustrate, in combination with other data representations (e.g. video segments), the results of an ethnographic study of a large printing organisation. The system thus enables multi-leveled envisionment of work practice. Ideas gained from experimentation with the VR model resulted in a 2D prototype system, based on the fieldwork, designed to support distributed document production among the various sites belonging to the organisation.

Another topic was the design of user embodiments. Anthony Guye-Vuillème (Swiss Federal Institute of Technology, Switzerland) presented an application to support non-verbal communication in virtual environments using facial, hand, arm and whole body gestures. Nic Earle (NVRCAD Research Centre, UK) described a study of a virtual environment which was then used to inform the design of a user embodiment or avatar. This presentation illustrated use of story-boarding and scenario exploration for the design of gesturing, expressive avatars which are capable of non-verbal cues. In this instance a model was built and animated to illustrate how body movement and facial expression can be used as indicators of degree of interest or activity.

In addition to 3D based VEs a number of talks focussed on text-based environments such as MUDs & MOOs. For example Markus Hoff (Universität-GH Paderborn, Germany) described a MOO based environment to support group work between university students, enabling document sharing and synchronous chat facilities. Rick Giles (Arcadia University, Canada) described a MUD based virtual environment designed to support on-going collaborations between software developers.

A number of papers explicitly addressed the nature of social interaction within virtual worlds. Barbara Becker (GMD, Germany) presented the results of a study of three different virtual environments both 3D and textual (Active Words, OnLive! Traveler and Lambda MOO) and described some of the social conventions used by the inhabitants of these environments. Her talk highlighted how social conventions varied between the different environments; affordances specific to the different environments clearly affected the interaction style. Jenny Preece (University of Maryland, USA) focused on the expression of empathy within virtual communities. The data presented showed clearly that consideration of emotional as well as factual aspects of communication are of import for design of environments which support group interactions.

Informal communications and chance encounters within physical environments are increasingly being recognised as arenas wherein work collaborations are initiated and maintained. Avon Huxor (Middlesex University, UK) described a virtual space constructed within Active Worlds which was designed to promote chance encounters with both his colleagues and students.

Evaluation of VEs is becoming increasingly important. Anthony Steed (UCL, UK) and Jolanda Tromp (Nottingham University, UK) described an evaluation performed on VR applications built by the EC funded COVEN project. In addition, one of the lunchtime discussion sessions was dedicated to the evaluation of CVEs.

Other presentations at CVE'98 included an architecture designed to enable interaction between different VR applications, a 3D interface designed to support collaborative working based on the metaphor of a conference table, and a presentation on `dual augmentation', which presented a description of tools (e.g. the `CyberFridge') which blur distinctions between `real' and `virtual' worlds (Jen Mankoff, Georgia Institute of Technology, USA).

Feedback from the CVE'98 delegates was very positive. Plans for the next CVE conference are already underway; planned for the year 2000, CVE2000 will probably be held in the USA. For further information on CVE'98 look one of the CVE'98 web sites at http://www.crg.cs.nott.ac.uk/events/CVE98/ or at http://www.fxpal.com/cve98/Report. Alternatively, send email to cve98@cs.nott.ac.uk.

References

[1] Benford, S. D., Bowers, J. M., Fahlen, L. E., Greenhalgh, C. M., Snowdon, D. N. (1997) Embodiments, Avatars, Clones and Agents for Multi-user, Multi-sensory Virtual Worlds. In Multimedia Systems, 5, (2), Springer-Verlag 1997, pg 93-104.

[2] Isaacs, E.A. (1998) Microcosm: support for virtual communities via an on-line graphical environment. In CHI'98 Summary, Los Angeles, 18-23 April, 1998. New York: ACM Press, p 5.

[3] The Palace Inc. (1997), Virtual World Chat Software, http://www.thepalace.com

Authors' Addresses

Dave Snowdon,
Xerox Research Centre Europe,
Grenoble, France
Dave.Snowdon@xrce.xerox.com

Elizabeth F. Churchill
FX Palo Alto Laboratory Inc.,
Palo Alto, CA 94304, USA
churchill@pal.xerox.com

No earlier issue with same topic
Issue
Previous article
Article
SIGCHI Bulletin
Vol.30 No.4, October 1998
Next article
Article
No later issue with same topic
Issue