Bisociative design: concept exploration by mining design associations across multiple products
Open Access
- Author:
- Kang, Sung Woo
- Graduate Program:
- Industrial Engineering
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- February 04, 2016
- Committee Members:
- Conrad S Tucker, Dissertation Advisor/Co-Advisor
Conrad S Tucker, Committee Chair/Co-Chair
Soundar Kumara, Committee Member
Gul Kremer, Committee Member
C Lee Giles, Committee Member - Keywords:
- engineering design
product design
new product development process
text mining - Abstract:
- More than 250,000 engineering products are released into the market each year. The increasing quantity and diversity of products makes it difficult to search through the entire design space for knowledge that may lead to the successful creation of new product concepts. Researchers have presented engineering design methodologies for new product development processes, such as design by analogy and bio-inspired design, which can discover design inspirations across multiple domains. In order to develop new design solutions, existing design methodologies focus on improving products’ functional aspects and engineering performance based on experts’ manual investigations. Although these methodologies have shown great promise in solving design issues in the engineering design field, they have not fully addressed approaches that can increase next generation products’ market sales. In the context of engineering design, a next generation product incorporates significant improvements in performance and design over the preceding generation’s product by changing or adding with new product attributes that addresses the needs of customers. Current design methodologies rely on manual investigations for analyzing and generating ideas, a process that is time consuming and costly. As products increase in quantity and complexity in the market space, such manual approaches to concept exploration become challenging. In order to mitigate these challenges, this dissertation proposes Bisociative Design; a methodology that efficiently searches for product attributes that increase the novelty, market sales, and sustainability of next generation products. Koestler defined the concept of Bisociation as “a synthesis of elements drawn from two previously unrelated matrices of thought into a new matrix of meaning by way of a process involving comparison.” The Bisociative design methodology proposed in this dissertation provides designers with new product attributes (i.e., elements) that can increase next generation products’ (i.e., new matrix) novelty, market sales, and sustainability (i.e., synergies arise from a synthesis). This dissertation hypothesizes that product attributes discovered using the Bisociative design methodology will the increase sales, novelty and sustainability of next generation products. In this dissertation, a product represents an artifact that is manufactured for sale. Using this definition, a module, which is a subassembly of the artifact, can also be considered a product since it is also manufactured for sale. In the context of engineering design, a product is assumed to be characterized by its form attribute and function attribute. The form attribute represents the product’s shape, while the function attribute represents the product’s objective. The methodology presented in each chapter of this dissertation automatically analyzes product attributes across multiple domains by employing natural language processing algorithms and 3D image mining techniques. Case studies related to each step of the new product development process demonstrate the Bisociative design methodology in chapters 4, 5, 6, and 7. Chapter 4 demonstrates how the methodology provides new product attributes that increase market sales. In Chapter 5, the methodology generates initial design concepts, which improve product distinctiveness compared to existing products, and establishes the feasibility of developing next generation products. Once designers select modules containing product attributes in order to create detailed designs for next generation products, the methodology presented in Chapter 6 automatically quantifies each module’s functional interaction, which reduces the manual efforts needed to analyze the interaction. In order to increase product sustainability, Chapter 7 extends the concept of Bisociative design to End-Of-Life products. Each chapter aims to reduce the manual efforts required for discovering product attributes across multiple products during the product design process. The "new matrix" is analogous to the next generation products that are launched into the market space.