Paper Reading #12 - TwinSpace: An Infrastructure for Cross-Reality Team Spaces

Comments:


Reference:
TwinSpace: An Infrastructure for Cross-Reality Team Spaces

Derek F. Reilly, Hafez Rouzati, Andy Wu, Jee Yeon Hwang, Jeremy Brudvik, W. Keith Edwards

UIST '10

Summary:

TwinSpace is an application that connects interactive workspaces with virtual worlds, the main purpose of which is to provide a "...flexible infrastructure for exploring and implementing collaborative cross-reality (CoXR) envrionments."  This expands from the current assumption that people interacting in virtual worlds are physically individually isolated from each other.  TwinSpace assumes individual collaborative groups sharing a virtual space.

Some of the problems encountered were how to make events in the physical world correspont to events in the virtual world, and vice-versa.  What kinds of actuators could be mapped between the two worlds?  They tested a layout based on activity mappings, where a team room is composed or assemble/array, aside table, and aside wall regions.  These are used for extended group discussions, comosing and modifying content, and generating whiteboard content, respectively.  Brainstorm mode promotes group collaboration and decision making, and present mode provides facilities for users to give a presentation.

Through recognizing common abstractions in virtual and physical spaces, TwinSpace provides a flexible fromework for promoting group collaboration and interaction in cross-reality environments.

Discussion:

I would like to see some data from test users to see just how intuitive and useful this environment is.  Most of the paper is simply describing how things work with the system, and the only case studies provided consisted of the design team's experiments.  To be fair, it appears that they were still in the prototype stage at the time the paper was written.  This is reflected in some of their own criticisms, where they note the need for better awareness of who's who in the virtual environment and the lack of representation of remote collaborators.  At the time, I must admit that this feels like a gimmick, but given enough time and modifications I think it could lead to practical and efficient cross-reality environments.