Automatic extraction of manufacturable features from CADD models using syntactic pattern recognition techniques

This paper reports a system for automatic extraction of manufacturable features from 'engineering drawings' created in CADD format. It employs the techniques of string based pattern recognition and natural language processing. Drawing entities are processed to derive feature string patterns, which are syntactically analyzed to detect the feature topology and geometry. The information of the features so detected is augmented by non-geometric data like dimensions, position, tolerances and the relevant drawing callouts. The system is capable of detecting generic classes of features like holes, pockets, steps, slots, bosses, etc. The derived information can be easily linked to the downline CAD/CAM activities in the product cycle. The system, thus, demonstrates the feasibility of design automation from CADD models.