Use of computers in advanced manufacturing industries started in the 1960s focused on challenges within design, analysis or production. The shape modelling approaches were tailored to the challenges addressed and the available computational resources. Consequently different shape modelling paradigms were employed within the different application domains and different systems. The need for communicating shape models  within and between organizations has lead to standardization, e.g., STEP (ISO 10303) within Computer Aided Design (CAD). The work on STEP started in 1984 with an initial release in 1994/1995. STEP consequently reflects the ideas of the late 1980ies and the qualities of shape models then envisaged.

Seen from the view of communicating organizations, STEP was a major move  in the right direction. STEP as an international standard for exchange of product model data has allowed organizations to exchange product model data, and contributed to global cooperative distributed design and production. However, the extensive exchange of CAD-descriptions between and within communication organizations and different computational systems is severely hampered by the inherent weaknesses in STEP based CAD-representations.

STEP-based CAD-models represent the inner and outer surfaces of objects by surface patches, and are not required to be  water tight, e.g., small gaps are allowed between the surface patches constituting the shells of volume objects. Analysis systems work on the complete volume, and require wate- tight models (no gaps). Due to this, up to 80% of the time used for analysis of CAD-models is used for gridding (making a simulation model from the CAD model). As little as 20% is spent on the actual analysis. The same shortcomings frequently create problems when the models are used by other systems outside of the CAD-system family in which it was created, e.g., for advanced visualization, production or maintenance purposes, or long term storage in product management systems. The consequence is a need for several representations of the same models  through the value chain, introducing problems in feedback, communication, and collaboration between the involved parties.