ComputerAided_Design_Engineering_amp_Manufactur.pdf

universities have demonstrated CBR in a wide variety of domains including law, cooking, and American
football. Schank’s student, Hammond (1989), described the architecture and the implementation of a
case-based planner for the generation of Szechwan cooking recipes. Some key components of a CBR
system are
• A means to store and index past cases.
• A means to retrieve the nearest case.
• A means to modify the old case to meet the goals of the new case.
Today, the implementation of these components has been restricted to relatively ‘‘simple’’ domains
such as generation of cooking recipes, criminal sentencing, legal reasoning in patent law, etc. In such
domains, it is possible to describe the criteria used to index the cases and describe the goals as textual
statements for storage and manipulation in a program.
One of the greatest challenges facing the development of a case-based system for die design is to find
a way to describe the product (which is basically geometrical in nature) in simple descriptive statements
such that it can be indexed and stored in a library of cases. Now, it is believed that the feature-tree of a
product can be used to index a metal-stamping workpiece for storage as a library of cases. The featuretree
can also be used to search and retrieve the nearest case for adaptation. However, the feature-tree of
a product can be topologically rather complicated and the pattern matching process required to locate
a similar tree or the nearest tree from a library of products can be tedious and computationally rather
complex. A complete treatment of a case-based approach for die design is outside the scope of this chapter.
Furthermore, there are still many problems that need to be addressed. However, we believe that the
techniques described in this chapter provide some of the initial building blocks to achieve the ultimate
objectives.
7.15 Conclusion
The tool and die industry is facing three severe challenges: First, the increasingly competitive market has
forced companies to push out new products more frequently to attract consumers. This has put a
tremendous amount of pressure on product and tooling designers who are required to work under the
constraints of increasingly shorter lead times. Second, as consumer expectation becomes more sophisticated,
the high product quality demanded invariably requires the use of new materials with high standards
of accuracy and finishing. The tools and dies required to make these products need to be more precise
and sophisticated, stretching the expertise of experienced tool and die designers to their limits. Third, in
the newly industrialized and industrializing countries, young people are reluctant to take up craftsmanshiprelated
courses that require long apprentice training programs. Hence, the tool and die industry will find
it very difficult to maintain the pool of well-trained, experienced personnel it requires to sustain the
growth needed to keep pace with the continuous demands for new and innovative products by the
consumer.
The use of conventional CAD/CAM systems has helped to automate the drafting, NC programming
work associated with design and tool making. This has partly helped to alleviate the problems posed by
the challenges identified above. However, it has not replaced the experience-based, trial-and-error
approach. A longer term solution is to introduce intelligent CAD/CAM practices to the die making
industry to help it meet with these challenges.
Further Information
The authors were awarded funding from the National Science and Technology Board (NSTB) of Singapore
on November 1, 1996 to develop an Intelligent Progressive Die (IPD) system, a knowledge-based software
for the planning, design, and manufacture of progressive dies.