Problems and Issues in Design and Manufacture of Complex Electro - Mechanical Systems

Project focus: The MADE program focuses on problems of design and manufacture of complex electro-mechanical (CEM) items like seeker heads and disk drives, an area that does not have a single technical focus or well developed engineering models. Practitioners deal in many technologies, including nontechnical ones like organizations and corporate cultures, and make heavy use of their experience to cover gaps in basic engineering knowledge, CAD tools, and multi-disciplinary design techniques. This situation contrasts sharply with that in printed wiring assemblies (PWAs) or digital micro-electronic manufacture, where there is one dominant technical and scientific focus, highly integrated CAD/CAM tools, and standardized manufacturing processes. Both military and commercial CEM items have similar characteristics and challenges, but commercial ones are especially important to understand in an era of dual use. The goal of this project is to identify the key technological and process blockages to improved design and manufacture of CEM items and help ARPA identify how it can make dramatic improvements in the cost, quality, and design time of these items. Findings: CEM products are growing in complexity as they shrink in size and perform more functions. Some can no longer be assembled manually, requiring new production methods. The infrastructure to produce such methods is not strong domestically. Companies with CEM products seek international component and production partners while reserving system responsibility to themselves in an effort to maintain competitive advantage without having to be capable in every precision technology. However, the strategy of relying on systems to remain competitive without having control over key component design and manufacture may not be stable. The system-driven methods of design and manufacture that are so successful in digital VLSI and PWAs are not directly applicable in CEM products for a variety of fundamental reasons. Digital VLSI design can be separated into distinct steps: component design first, then system design. This separation, crucial to VLSI design efficiency, is not possible in CEM design, where components and systems are designed together. In fact, digital VLSI may be a special case, since design of other electronic items like hybrid and RF circuits shares many characteristics with CEM design. Furthermore, VLSI design techniques compromise the designers' freedom and create larger chips (~10%) in exchange for the ability to design extremely complex systems efficiently. CEM designers cannot leverage a small library of verified components into huge systems but must design components from scratch. Isolated design and fabrication tools exist for these components, but multi-phenomenon component analyses (geometry+magnetics+fluids+vibration...) do not exist. Design tools for complete systems are almost totally lacking. Some component and system design problems can be addressed with more powerful computation, but others are faced with fundamental engineering knowledge gaps and lack of common data representations for analyzing multiple phenomena and linking function to design, manufacturing, and assembly. Application of the VLSI approach (library