Collaborative Work in Virtual Environments
Kai-Mikael Jää-Aro
Interaction and Presentation Laboratory
Department of Numerical Analysis and Computing Science
Royal Institute of Technology
Stockholm
Sweden
What are Virtual Environments?
The popular view: Immersive virtual environments
Mårten Stenius at the Interaction and Presentation
Laboratory, wearing a Virtual Research Flight Helmet stereoscopic
head-mounted display and wielding an Ascension Technology Flying Mouse. In the background
is the console of an SGI Onyx VTX driving the display.
Projection-based virtual environments
The CAVE environment has been designed at the Electronic Visualization
Laboratory of the University of Illinois at Urbana-Champaign. The CAVE
projects stereoscopic images on the walls of a cube, large enough for several
people to stand inside.
Screen based virtual environments
3D stereoscopic pictures can also be generated on an ordinary workstation
screen. The stereo effect can be produced in many ways, but a fairly standard
method is using CrystalEyes LCD shutter glasses, which allow allow one eye at
at time to see the alternating left- and right-eye views that give a 3D
impression. This can be combined with head-tracking to give the ability to
look around objects.
Multi-spectator virtual environments
Several people can experience the environment, but only one person can actually interact with it
- CAVE
- CrystalEyes projection system
Multi-user virtual environments
Several users can concurrently be present in and interact with the virtual environment
Networked virtual environments
Several users distributed over a network can participate in a common virtual environment.
What is Computer-Supported Cooperative Work?
Very little work is actually done by single persons, but rather by smaller (e g design teams) or larger (entire corporations) groups.
It should be possible to support the cooperation within these groups. This is the goal of CSCW.
Why use Virtual Environments for CSCW?
Virtual environments are presumably a Good Thing, since they give us:
- The ability to present a large amount of information
- Natural information lensing
- Support for many sensory modalities
- Natural multi-user interaction
- Natural awareness of co-workers' activities
Teleconferencing
A necessary function for most CSCW activities is communication with co-workers.

A phone booth in the DIVE environment, created by Emmanuel Frécon.
A text-based 'talk'-window is even in the presence of better communication channels a useful backup and good for indicating file names and the like.
Audio communication is natural, immediate and hands off.
Video communication gives facial expressions of participants.
Teleconferencing in Virtual Environments
Ordinary teleconferences suffer from a lack of spatial cues. Virtual environments can add
- Spatialised sound support for directing attention and pinpointing sound sources outside the field of view
- Gaze direction as indicator of attention
- Spatial presence and activity of participants

A conferencing environment created in DIVE. The environment is embellished
with texture maps and local lighting models.
Shared artifacts
Applications can be represented in the virtual environment and used concurrently or alternatingly by the participants.

Two virtual bodies in a DIVE model of the atrium at Electrum. Between them
stands a whiteboard application on which all participants can draw.
Embodiments
User representations in the virtual environment indicate
- Presence
- Identity
- Position
- Attention
- Activity
- Hardware capabilities
- Personality

Conferencing participants gathered around a table in the MASSIVE system.
Embodiments with ears indicate users with audio capability, those with Ts
embossed in their faces are limited to a text-only client with graphics
capability. The "owner" of the supine embodiment has probably left his
workstation in the real world and indicated this by lying down.
The Spatial Model
The spatial model gives
support for direction of attention, public addressing and private conversations
using as close to normal body language as possible. The spatial model was
first tested in DIVE. MASSIVE was built specifically to test the spatial model
for collaboration.
We define a set of subspaces around the user's embodiment, a pair for each sensory modality.
The focus defines the extent of the user's attention.
The nimbus defines the extent of the user's presence in the world.
The focus and nimbus are enveloped by the aura.
When one user's focus intersects another user's nimbus the first becomes aware of the second.

Two users are surrounded by their auræ which have just collided. Awareness
calculations are performed and the users' views and hearing of each other is adjusted.
The environment may contain aura modifiers or adaptor objects that enhance or shape the user's aura.

The conference table is an aura modifier that folds in the auræ of all those
around the table so that they can hear and speak to each other regardless of
the actual shape and size of their individual auræ.
Populated Information Terrains
PITs are information retrieval applications in which the users are aware of each other's presence and activity.
Information is laid out in 3-space and use shape and colouring to aid in organisation and search.
Current applications include VR-VIBE and QPITs.

A visualisation of 1500 bibliographical references, laid out in space
according to their affinity to the keywords placed at the nodes of the pyramid.
Shared Editing and Design
Preliminary results show that concurrent collaborative editing can be very efficient, at least for some tasks.

A collaborative
3D graph editor, created by Rolf
Staflin in DIVE. The graph being edited is a state machine representation
of a communications protocol. In front is a set of buttons for actions on the
objects in the display.

A collaborative 3D
modeller created by Mårten Stenius in DIVE. At the bottom is a tool menu
for creating objects in the environment, at the top right is an indicator of
the position and orientation of the selcted object (the cone with yellow
markers). At the right is a skeleton showing the hierarchical organisation of
the compong object being created. (No, I do not know what it is.)
Monitoring and Control
Still unproven value, but the hope is to utilise some kind of VE technology for e g process control.

A design for a factory monitoring application, created by Worldesign, Inc.

A LAN
monitoring application created by Kai-Mikael Jää-Aro in DIVE. Visible is a
number of workstations and a printer, their colour indicating their current
system load.
Subjective Representations
The different participants do not necessarily have to see their environment in the same way. How they are going to be able to communicate sensibly if they see different things is still unsolved.
The Greenspace
project, teleconferencing in a virtual environment, jointly undertaken by
the Human Interface Technology
Laboratory, in Seattle, Washington and Fujitsu Research Institute in
Tokyo. On the left is a view from the US end of the conversation and on the
right is a view from the Japanese end.
European Research
Projects
Organisations