Department of Mechanical Engineering

ME Seminar - A Computational Theory for the Generation of Solutions During the Conceptual Stage of Design

 

Mechanical Engineering Seminar
Faculty Candidate
Wednesday, April 4, 2007
11:00 A.M., 3004/3006 Black Engineering Building

A Computational Theory for the Generation of Solutions During the Conceptual Stage of Design

By
Cari R. Bryant
Design Engineering Lab
University of Missouri-Rolla

 

Abstract

The creative nature of design generation demands skills from a designer that must be developed and refined through practice. Advancement in technology is usually made by building on previous experiences and learning from past successes and failures. Although many successful designs are easily identifiable, often it is unclear why or how that success materialized without prior experience dissecting or designing a similar product.  Research has shown that successful component configurations can be dissected and stored for reuse. Even if experience in the form of design knowledge is accessible, both the experienced and inexperienced designer may feel compelled to become fixated on a particular solution or domain restricted set of solutions based on instinct or, perhaps, a subconscious desire to pursue an initial ‘gut feeling.’ So, the challenge in creating useful design tools becomes finding innovative ways to guide an engineer toward the best solution(s) by building on existing design experience while simultaneously discouraging tendencies to make choices or evaluations based on hunches or biased methods.Few computational tools exist to assist designers during the conceptual phase of design. Many well-known manual methods (e.g. brainstorming, intrinsic and extrinsic searches, and morphological analysis) are designed to stimulate a designer’s creativity, but ultimately still rely heavily on individual bias and experience, and are often time intensive, laborious tasks that may not catch solutions that seem unrelated but are, in fact, functionally analogous. Under the premise that quality designs come from experienced designers, experience in the form of design knowledge may be extracted from existing products and stored for reuse in a web-based repository.The research presented covers an automated concept generation tool that augment traditional activities during the conceptual phase of design. The automated concept generator draws on the existing knowledge contained within the repository of existing design solutions to quickly produce numerous feasible concepts early in the design process that each satisfy the functional requirements for a design problem. The computational algorithm enables the development of a computerized design tool that complements other concept generation activities, such as brainstorming and morphological analyses. By quickly presenting numerous concepts from products that have already been developed, this design tool provides a broader set of initial concepts for evaluation than a designer may generate alone when limited by his/her personal experiences. Furthermore, the computational tool further utilizes component information stored within the repository of knowledge to facilitate decisions regarding the merit of concepts by a designer for further development.

Biography

Cari R. Bryant received her BS in mechanical engineering from the University of Missouri–Rolla in 2000. In 2003 she received an MS in mechanical engineering and an MS in biomedical engineering from the University of Michigan while doing research in the University of Michigan Orthopaedic Research Laboratories. She is currently approaching the completion of her PhD degree in mechanical engineering with the Design Engineering Laboratory at the University of Missouri–Rolla under advisors Robert B. Stone and Daniel A. McAdams.                                                                                     

---