EVO-HD: A Globally Scalable Standards-based Full-HD Environment for Immersive Collaboration

This project will develop an extensible, globally scalable High Definition (HD) standards-based platform for remote collaboration (EVO-HD) at low cost, packaged for widespread corporate, research and at-home use over existing and future-generation broadband networks. Based on results with a recent prototype, we are confident that we can develop and deploy standards-based HD videoconferencing and remote collaboration end points, embedding many high frame rate HD and lower resolution streams and a wide variety of shared content, in a real-time 3D immersive collaborative environment, and that we can integrate them into a highly capable autonomous system infrastructure that scales from a few to hundreds of end points. The EVO-HD system would be alone in its class in terms of scalability, minimal operational burden and cost due to: 1) its built-in field tested and proven end-to-end intelligence (based on autonomous agents), 2) its automatic real-time topology and data flow optimization, 3) its ability to integrate across a wide range of home and work environments (including low cost mass market desktops, laptops and room-based systems), and 4) its unique user interface in which many interactive video streams, documents, remote desktop areas, and other information are presented live on 3D surfaces using OpenGL technology. The proposed full HD implementation of the EVO collaboration system will leverage recent mass market developments that have sharply lowered the cost of a) 1080i and 1080p (¿full HD¿) video cameras, b) large LCD and plasma screens, c) capture and graphics cards with powerful GPUs, d) libraries supporting HD resolution, e) multi-core CPUs in mass market PCs that, together with the GPUs and libraries, can easily support one or several HD streams. The creation of low-cost HD Television (HDTV) systems running EVO for remote collaboration, and the adaptation of EVO software and network technology to allow the EVO-HD end points to interwork with the full range of existing EVO end-clients (with varying degrees of resolution and interactivity) will leverage the Caltech EVO team¿s 14 years of development and round-the-clock operational experience, as well as work at Caltech on autonomous systems that monitor, support and help manage networks, large scale computational grids and real-time distributed applications