Thomas Thurman

Towards next-generation design-for-manufacturability (DFM) frameworks for electronics product realization

This paper elucidates the process architecture of a pilot implementation of a DFM Framework (specifically the SFM DFM Framework or SDF), which consists of four key ingredients. The first ingredient is a Design Integrator that acquires product design information from an ECAD tool and in-house sources (each populating a subset of the design) and consolidates them into a STEP AP210 model. The second ingredient is a Rule-based Expert System (initiated at Boeing) that captures the manufacturability constraints as DFM rules and evaluates printed circuit assembly (PCA) designs against them. The third ingredient is a Design View Generator that extracts design information from the AP210 model (first ingredient) and library database and derives a Kappa design model for the expert system (second ingredient) to evaluate. The fourth ingredient is the Results Viewer that helps the user browse DFM analysis results and identify design improvement opportunities. This implementation of the SDF demonstrates the ability to extract PCA design information and build a higher fidelity standards-based design model. Additionally, it also shows the capability of Rule-based Expert Systems to emulate manufacturability checks on product (PCAs in this case) designs as well as increase analysis coverage and reduce human checking time via automation.

Requirements on information technology for product lifecycle management

Good decision-making is founded on good information. Information technology supporting product lifecycle management ought to provide a high degree of information cohesion and traceability – knowledge of the interrelations among data, and basis for belief. Providing cohesion and traceability is made difficult by differences in viewpoint and ontology employed by the various disciplines and organisations involved in the product lifecycle. This paper describes an analysis of cohesion and traceability into its constituents properties. The paper suggests that process-aware integration schema can improve the cohesion and traceability among product data.