Shared Virtual Environments

So far, the work in collaborative systems has looked mostly at sharing of 2-D textual/graphics objects. How should virtual environments be shared? One approach is to simulate ``being there,'' that is, create a virtual environment that faithfully mimics the real environment at a remote site. It would be useful, for instance, in supporting telesurgery. The dual approach is to go ``beyond being there'' by supporting (for logical or implementation reasons) modes of collaboration not possible in face-to-face meetings. It would allow, for instance, two users designing a 3-D object to see different views of the object.

The former approach and the difficult implementation issues raised by it are being investigated by the telepresence project of Prof. Henry Fuchs. This project is concerned with the latter approach. It must address several design and implementation issues such as: What are the various dimensions along which coupled VEs can differ? What kind of concurrency control, merging, and security model should be used for VEs? What should be the architecture of the VE application, that is, which parts of the VE application should be centralized, which should be replicated, and which should dynamically migrate? What kind of VR-specific media adaptations must be added to the collaboration-bus kernel to support interoperating VEs? The goal of this project is to generalize/adapt these previous approaches for the world of virtual environments. In particular, it will build on the media bus project at UNC and Maryland.

It will give you experience with research in virtual environments, distributed architectures, interoperability, and concurrency control. For more information, contact Prasun Dewan, Dave Stotts, or Lars Nyland.