The SOFA architecture relies on several innovative concepts, in particular the notion of multi-model representation.
In SOFA, most simulation components (deformable models, collision models, instruments, ...) can have several representations, connected together through a mechanism called mapping.
Each representation can then be optimized for a particular task (e.g. collision detection, visualization) while at the same time improving interoperability by creating a clear separation be tween the functional aspects of the simulation components. As a consequence, it is possible to have models of very different nature interact together, for instance rigid bodies, deformable objects, and fluids. At a finer level of granularity, we also propose a decomposition of physical models (i.e. any model that behaves according to the laws of physics) into a set of basic components. This decomposition leads for instance to a representation of mechanical models as a set of degrees of freedom and force fields acting on these degrees of freedom.
Another key aspect of SOFA is the use of a scene-graph to organize and process the elements of a simulation while clearly separating the computation tasks from their possibly parallel scheduling.
