A CAD Methodology for Knowledge Assisted Design

Abstract

Modern computer-aided design (CAD) systems have developed into integral support tools for the product design and development process. Designers must, however, draw upon experiential engineering knowledge such as past experiences, specific design rules and procedures, and heuristic reasoning just as before the advent of CAD. This work develops a methodology for integrating experiential engineering knowledge in an interactive CAD environment that serves as a knowledgeable design assistant and supports a design process controlled by the designer. The knowledge assisted design environment is an object-oriented, domain independent framework based on a blackboard architecture that incorporates a feature-based design environment with multiple, autonomous knowledge sources. The system can be utilized for any domain for which a set of features and knowledge sources have been defined. The knowledge sources provide design assistance by reacting opportunistically to a developing design solution and by presenting advice interactively to the designer. The object-oriented knowledge and hierarchical, feature-based model representations are presented along with the design environment and its functionality. The methodology is applied to the industrial application of engine flywheel design.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Aug 07, 1998
Accession Number
ADA350929

Entities

People

  • Paul A. Hey

Organizations

  • Purdue University

Tags

Communities of Interest

  • Autonomy
  • Biomedical
  • Human Systems

DTIC Thesaurus Topics

  • Advanced Manufacturing
  • Air Force
  • Artificial Intelligence
  • Cognition
  • Computer Programming
  • Computer Programs
  • Computer-Aided Design
  • Computers
  • Control Systems
  • Engineering
  • Engineers
  • Expert Systems
  • Graphical User Interface
  • Inference Engines
  • Manufacturing
  • Mechanical Engineering
  • Ontologies

Fields of Study

  • Computer science

Readers

  • Artificial Intelligence
  • Software Engineering.