Modular Technology in the Modelling of Large Virtual Environments in Driving Simulators

Abstract

This paper presents the latest research and developments in Modular Technology. That is, the optimized repetition of the same geometry or module, for the generation of large virtual environments for the simulators that are designed by CITEF. The current trend is on redirecting the maximum possible share of graphical calculation to the GPU to lighten the load of the CPU as far as possible, as this is slower and less suitable for graphic calculation. Modular Technology, which has been successfully used for some time in driving training simulators, is particularly suitable for the implementation of operations in the GPU through programming shaders. In this way a substantial reduction in the number of geographical entities in the environment can be attained and an increase in the diversity and flexibility of the environment. This is achieved by discretizing the environment into a series of instantiated modules and using shaders in an appropriate manner based only on a few parameters. There are many advantages to be had from this form of generation: savings in scene size, loading times and resource requirements, greater flexibility and clarity of the scene graph and a more substantial upgrade capacity of the environment.