Feature ontologies for the explicit representation of shape semantics

Current CAD systems encode shape semantics as so called parametric features on different levels of abstraction. Here we discuss an approach that combines parametric modeling with techniques from the field of knowledge representation and ontological reasoning. Parametric models refer to feature ontologies that model feature semantics on several levels of granularity. On higher levels the interrelation between features and feature interoperability is captured whereas on lower levels a feature is described in terms of geometric, topological, and parametric entities. Different engineering tasks can utilize feature ontologies as a basis for application specific shape reasoning across several modeling layers.