Virtual Reality in Science

Content

Name of the Game

What is virtual reality? As one discover from the following list, it is still not real clear what we are talking about!

Abstract

At the time I'm working on a distributed, non-immersive VR environment based on PVM. By making extensive use of the object-oriented paradigm it is very easy to describe a virtual world. The outline I follow is to have a worldserver which manages registering, unregistering, entering and leaving the world. I distinguish two kind of objects - passive and active objects, where active objects are inherited from the passive ones by adding auto-functionality. Once an object has entered the world or was inserted to it, it informs the worldserver if its properties have changed. The worldserver broadcasts this message to all other objects inside the world, which have to handle that information them self. At the first moment this looks not very effective due to a lot of communication. The advantage is the independence of the objects. E.g. in this way it is possible for a lot of observers to enter or leave the world any time they want to.
The project will be formulated in C++ and later I will do some mathematics on it to get a (hopefully) abstract description of the whole thing.

Some Theoretical Aspects

OO Paradigm

Here we give a short introduction to the system described in Gunter/Mitchell: "Theoretical Aspects of Object-Oriented Programming". We continue with building a formalism for the concepts of polymorphism, class and inheritance. Finaly we give a graph describing the relations between classes and/or instances.
Our final goal will be a distribution strategy for instances on a distributed system. Therefore we appologize solving a min-cut problem on the introduced graph.

A Virtual Laboratory

Brief Description

The outcome of these first lines of code was an exapmle-environment which allows scientist to setup experiments without using any real instrument nor device. I call it the "Virtual Laboratory". Currently a demonstration is produced which simulates an electron bunch in a linear accelerator. For doing this experiment the scientist has to describe The environment is used for As in real life it is possible to interact to the process any time the scientist want to. This means he can stop the simulation any time, may alter the setup and rerun it. This may save time if one discovers an error after half the simulation is done. It also may help to prevent real instruments from beeing damaged.

At all the concept will be totally different to a normal simulation which is done with a sequential/parallelized algorithm. The path I follow really starts with an independent description of the devices, instruments and particles. The next steps explains how they interact e.g. an accelerator tube interacts on electrons, which means it takes control over the electrons as long as they remain inside the tube. This way no user nor any simualation needs to have any a prioi knowledge about the experiment which will be carried out. As long as the descriptions are not too far from reality the whole thing will be a real experiment.

State of Art

So far the worldserver, viewers and some intrinsic classes are implemented. It currently runs on a PVM cluster of a Convex SPP/1000 Exemplar, a Convex C-3820, a C-3220, an IBM RS6000/591 and an IBM/PC with Linux as I/O knot of the parallel virtual machine and - if I can find a PowerGlove (A.G.E. seams to be broken).
If I will succeed in building my own HMD - I will offer a kind of immersion to the virtual world.

Virtual Instruments/Devices
Geometries
Processes
Experiments

Research Institutes and Organizations

Currently there is a project called MuSE in progress at IGD FhG.
CORBA
MUSE, Werner John

Links to other VR places

Ongoing events

Up Department, Up Institute