Siggraph 2005 logo
Conference Exhibition Call For Participation Presenters Media Registration

emerging technologies

The Virtual Raft Project: A Network of Mobile and Stationary Computer Systems Inhabited by Communities of Interactive Animated Agents

A novel interaction paradigm in which computer screens serve as islands of virtual space ("iLands"). Mobile "virtual rafts" let people move animated agents between iLands. The project features innovations in heterogeneous computer graphics, tangible human-computer interaction, interactive animated characters, and mobile computing technology, and has applications in education, entertainment, and new-media art.

Art and Science

The Virtual Raft Project offers a unique physical and graphical interaction that can be used for interactive storytelling and learning, and as a "sand box" in which participants can experiment with interactions of animated agents.


1. To develop an interactive platform that enables physical human interaction with virtual characters.

2. To explore the implications of this kind of heterogeneous platform for animation of autonomous characters, as they break away from the constraints of a fixed desktop screen.

3. To enable interactive education and active learning by creating an engaging setting for exploring a range of content domains, starting here with color theory, but extending to more complex domains such as ecology and other system sciences.


1. This project offers an example of animated characters that are able to move seamlessly among heterogeneous graphical systems. If a character exists on only one screen, it is a subset of that screen, a piece of that machine. If, on the other hand, a character can move between screens, then it appears to exist in some broader sense, to be independent of any given machine. Networked multiplayer games do this to a certain extent as well, but there the characters remain "in the box." The characters in the The Virtual Raft Project appear to exist in real space, interacting physically with people, and are able to move between both stationary and mobile graphical screens. Through this dynamic cross-screen animation, the characters appear to be more believable than characters that only exist on fixed computer screens.

2. Through the virtual rafts, the work presents a novel tangible interaction between humans and virtual characters. Accelerometers in the tablet PCs allow characters on the rafts to respond in real time to the physical motion of the device; this visual feedback appears to enhance the realism of the characters and create an enjoyable experience for human interactors.

3. The work demonstrates an "island metaphor" for computational interactions, which helps to frame the relationship between real space and virtual space, and between the real creatures and virtual creatures who inhabit those spaces. As computer programs become more capable and autonomous, the desktop metaphor is growing less relevant to computational systems. Files, staplers and tape dispensers rest on desktops; autonomous entities generally do not. A metaphor that is more appropriate to the growing trend toward autonomous computational entities is needed. While it is unlikely that the island metaphor will supplant the traditional desktop metaphor, it is a useful way of thinking about future computational systems and potentially extending the capabilities of computational technologies.


This example of a novel interaction paradigm enables autonomous animated characters to break free from the traditional desktop screen. Because it allows characters to jump between stationary and mobile computing platforms, the system increases the realism of the characters and helps to frame the way people should interact with autonomous agents. The "island metaphor" for computational devices points toward a new kind of interaction designed to accommodate and enhance the capabilities of autonomous systems. This interaction could be used as a platform for new kinds of educational exhibits, new genres of location-based entertainment, and new forms of interactive media art. In particular, the development team is currently collaborating with several science centers to develop a version of the project based on multi-species ecological interactions.


Bill Tomlinson
University of California, Irvine
wmt (at)


Eric Baumer
Sara Goetz
Jessica O'Connell
Ksatria Williams
So Yamaoka
Man Lok Yau
University of California, Irvine

Jesse Gray
University of California, Irvine
Massachusetts Institute of Technology