DEVELOPMENT OF AN INTELLIGENT PRODUCT DESIGN SYSTEM: INTEGRATION STRATEGIES

Abstract

Conventional computer-aided design (CAD) packages have drastically reduced the workload of the human designer and shortened the product design cycle. However, the degree of effort and volume of information required to use these tools limit their use to the later stages of design. Intelligent computer-aided design (ICAD) systems have sought to provide a more complete design tool to assist the designer in all phases of design. ICAD systems incorporate conventional CAD elements as well as knowledge engineering constructs. The level of integration between different components of an ICAD system determines its usefulness. Most commercial intelligent CAD packages are tied to a specific set of CAD tools, restricting their application domains. This dependence on specific software tools can be reduced by using general purpose modules to interface with available CAD packages. This paper discusses a method of introducing knowledge engineering technology to help develop an advanced intelligent product design system by integrating ICAD's Concept Modeller with SDRC's l-DEAS package for engineering product design. This integration is necessary because neither the Concept Modeller nor the I-DEAS package provides any unified design environment where users can access both symbolic and numerical design tools as needed to carry out design synthesis and analysis activities. Interfacing engineering design and knowledge processing together is not an easy task. The task is further complicated since it needs to be done only by those who have enough knowledge of both technologies, and also because it may result in reorganization of the traditional design process altogether. The proposed intelligent product design system uses artificial intelligence techniques to take care of human experts knowledge and it advocates the use of several commercial software packages that come from a variety of sources (and are proven to be robust) to perform design synthesis in a cost-efficient and timely manner. The technique described here is relatively easy to implement and is well suited to industrial needs.

Notes

The authors would like to thank ICAD Inc. for providing the Concept Modeller software for this research project on the development of an intelligent product modeling system. The authors also acknowledge the help provided by Chris Ludden, Syracuse University, in many aspects of this work. The authors wish to thank the anonymous referee for constructive suggestions to improve the quality of the paper.

Note: Concept Modeller and IDEAS are registered trademarks of ICAD Inc. and Structural Dynamics Research Corporation, respectively.

Additional information

Notes on contributors

UTPAL ROY

Address correspondence to Utpal Roy, Knowledge Based Engineering Laboratory, Department of Mechanical, Aerospace and Manufacturing Engineering, Syracuse University, 151 Link Hall, Syracuse, NY 13244, USA.