Part design is the principal source of communicating design intent to manufacturing and inspection. Design data are often communicated through computer-aided design (CAD) systems. Modern analytics tools and artificial intelligence integration into manufacturing have significantly advanced machine recognition of design specification and manufacturing constraints. These algorithms require data to be uniformly structured and easily consumable; however, the design data are represented in a graphical structure and contain a nonuniform structure, which limits the use of machine learning algorithms for a variety of tasks. This paper proposes an algorithm for extracting dimensional data from three-dimensional (3D) part designs in a structured manner. The algorithm extracts face dimensions and their relationships with other faces, enabling the recognition of underlying patterns and expanding the applicability of machine learning for various tasks. The extracted part dimensions can be stored in a dimension-based numeric extensible markup language (XML) file, allowing for easy storage and use in machine-readable formats. The resulting XML file provides a dimensional representation of the part data based on their features. The proposed algorithm reads and extracts dimensions with respect to each face of the part design, preserving the dimensional and face relevance. The uniform structure of the design data facilitates the processing of data by machine learning algorithms, enabling the detection of hidden patterns and the development of pattern-based predictive algorithms.

Part design is the principal source of communicating design intent to manufacturing and inspection. The design data is often communicated through CAD systems. Modern analytics tools and artificial intelligence integration into manufacturing has significantly advanced machine recognition of design specification and manufacturing constraints. This paper is aimed at the collaboration among multiple vendors across supply chains to enable efficient order procurement. To this end, the paper discusses the development of a simple framework for extracting the dimensional data from part design and storing them for enhancing machine readability of the part design at multiple levels of manufacturing.