This research develops a novel representation scheme for collaborative 3D product development with multiple levels of detail (LODs) based on integration of feature model and 2D meshes. Different features are selectively hidden in the product model from certain collaborators, depending on their individual authorities and actual requirements in collaboration. The feature switch face is used to control display/hiding of each feature. A tree-like data structure is introduced for characterizing the feature hierarchy, 2D mesh, and the relations of each feature to the switch faces and LODs in a product model. Algorithms are also provided for quick generation of the mesh data from the data structure for displaying the 3D model corresponding to any LOD. An agent-based system, which allows multi-user synchronous collaborative 3D assembly over the Internet, is implemented to demonstrate the effectiveness of this work. A real scenario of 3D mold design using the system shows that the proposed LOD concept enhances the security of information sharing and the efficiency of data transmission while ensures the practicality in network-based 3D design collaboration.
INTRODUCTIONThe challenge most enterprises facing today is to assure their core competences and outsource less-skilled activities in order to remain competitive in global competition. This is especially important for the medium and small sized enterprises (SME's), which rely on close collaborations with customers, suppliers, and other partners of its supply chain, to compete with large companies. Collaborative product development (CPD) has thus emerged as an effective business approach in many industry sectors [6]. CPD involves a wide spectrum of product development activities, in which design is one of the most complex tasks. The design process is referred to as collaborative design when a product is designed through the collective and joint efforts of many designers [5]. There have been a number of studies concerning distributed collaborative design. Li et. al [13] performed a survey and discussed current research status in this field. Integration of horizontal and hierarchical collaboration, new feature-based streaming technology, and security/interoperability are identified as future research challenges. Multiple-resolution techniques [11,14] have been identified for years as an effective method to achieve good performance in terms of shape preservation and data transfer. The most common approach is mesh refinement, or simplification, adopted for polygon meshes with various LODs, i.e. near objects are rendered with a higher resolution while distant ones with a lower resolution. Most mesh generation based on this approach have been focused on graphics rendering with less emphasis on LOD of design semantics. On the other hand, there have been studies that deal with compression and de-compression of 3D CAD in B-Rep in distributed environment. Koo and Lee [12] developed a "wrap-around" algorithm for producing multi-resolution representations from a B-Rep model, which is...