Fig. 3 presents the architecture of Sandia's distributed virtual environment platform. Sandia's distributed virtual environment system supports multiple, independent participants. It is device independent and supports VE configurations ranging from flat screen and mouse to a full immersion interface consisting of head-mounted displays, magnetic position trackers, voice recognition, and sound/vocal output. The platform uses standard Ethernet protocols for multicasting information from different processes across the net. Currently, it requires a Local Area Network configuration to ensure information integrity.
Figure 3: Architecture of Distributed Virtual Environment Platform
The platform is open and permits integration of heterogeneous simulation modules, each driving some aspect of the shared virtual environment. For example, for the MediSim project, the system permits multiple users to share a common battlefield VE. Each participant is represented within this environment by his or her avatar. Each avatar is, in turn, created using the Jack software. Sensor information is broadcast to the rendering modules (VR_Station) for each participant allowing the system to update his or her view of the world. In addition, this broadcast information is used by the Jack software to determine the new posture and position of the participant's avatar. Simulated medics, casualties, and computer-based medical decision and planning components are also easily integrated using these techniques.
A trainee may command the semi-autonomous medic (or his/her avatar) to carry out a medical procedure via voice commands. A set of commands is being developed for the tension pneumothorax intervention, as well as for the general preliminary assessment carried out prior to any diagnosis and treatment. When a vocal command is received, the system sends a message to the Jack software which will assess the validity of the command and determine the subsequent actions. This information will be passed back to the Sandia platform, and the required simulation and/or vocal feedback will be presented to the trainee. Fig. 4 shows a trainee with VR devices, his avatar, a simulated casualty, and a SAM, which the trainee may command via vocal commands.
Figure 4: Trainee in MediSim Environment