Supplying concurrent engineering information to the designer: The Conceptual Design Information Server

Conceptual design is distinguished by the type and amount of information furnished to and required by the designer; informal design specifications must be formalized through the incorporation of requirements from design 'stakeholders' and the interaction of these requirements with the space of design possibilities. This transformation requires knowledge not only of the issues important to each stakeholder but also of how their interactions have been manifested in prior successful designs. Knowledge of design possibilities gained from experience, from catalogs, or even from reverse engineering is also vital. The importance of good conceptual design cannot be understated; by defining the 'neighborhood' of the final design, conceptual design strongly influences its ultimate success.;The Conceptual Design Information Server (CDIS) organizes and presents concurrent engineering and life cycle design experience in two distinct representations: networked hypermedia documents and mathematical models. Implemented in an open architecture based on network standards, the three CDIS applications--Concurrent Engineering Case Studies, the Design Discussion Server, and the Concept Database--encode design experience ranging from industry 'best practices' to issues resolution in various designs to the evaluation and selection of engineering components. Hypermedia captures information about design process and context that is absent in typical case representations. The tradeoff for this expressiveness comes in the difficulty of structuring this information for reuse: while hypermedia provides strong local structuring, finding the appropriate localities requires global indexing by design context. Context is extracted from the design case base along three distinct axes: concurrent engineering and life cycle design issue, functional decomposition, and component classification. These axes are represented as abstraction hierarchies of textual terms and descriptive phrases which, coupled to a network-standard information retrieval engine, greatly improve access to the CDIS case base.;The component abstraction hierarchy is also used to help structure the design space as represented by component catalogs. Decision and information value theory operate over a probabilistic representation of the design space learned from these catalogs and structured by the component abstraction hierarchy. By evaluating the impact of decreasing abstraction level, the CDIS can help the designer understand the tradeoffs of committing to a component class.