3D Discrete Event Systems: An Efficient Way to Model and Supervise Dynamic Behavior in Virtual Environments

Author(s):  
Ju¨rgen Roßmann ◽  
Michael Schluse ◽  
Ralf Waspe

The combination of Discrete Event Systems with 3D simulation and control technology to set up versatile virtual environments seems to be at the first glance a contradiction in terms because of mixing up two totally different application classes. But here, the newly developed State Oriented Modeling methodology comes into play providing a comprehensive and flexible object oriented framework for the development of complex Discrete Event Systems. State Oriented Modeling establishes a link between these two worlds using object oriented Petri Nets [10] for the discrete part as well as application specific mappings between continuous and discrete state variables realizing so called 3D Discrete Event Systems. Integrated in a versatile 3D simulation system, State Oriented Modeling allows for the efficient realization of a large class of applications within the large field of virtual environments. It provides a comprehensive, easy-to-use but nevertheless efficient framework for behavior modeling in virtual worlds and real-time simulation applications. When modeling virtual production lines for example, State Oriented Modeling provides an easy way for a close-to-reality simulation of automation systems starting with the control of simple actuators or sensors and ending with the supervisory control of the entire virtual factory. In the field of Projective Virtual Reality [4] State Oriented Modeling supervises the user’s work in the virtual environment to derive his “intention” and then generates appropriate command sequences for intelligent automation hardware. This way, even complex automation systems which may be far away or even in space can be controlled — and the user must not be an automation expert. In the field of interactive visualization and training environments State Oriented Modeling simulates a variety of dynamic objects in the virtual world, integrates interaction components like 2D control elements or application specific driver seats into the virtual environment and supervises training sessions to derive performance indicators. And, in addition to this, State Oriented Modeling is the basis for a new multi-agent and discrete event based approach to realize the virtual human.

Sign in / Sign up

Export Citation Format

Share Document