Date of Thesis

5-4-2017

Thesis Type

Masters Thesis

Degree Type

Master of Science in Mechanical Engineering

Department

Mechanical Engineering

First Advisor

Charles J. Kim

Abstract

The aim of this study was to develop the process necessary to identify design knowledge shared across product classes and contexts in Design for the Developing World. A process for extracting design knowledge in the field of Design for the Developing World was developed based on the Knowledge Discovery in Databases framework. This process was applied to extract knowledge from a sample dataset of 48 products and small-scale technologies. Unsupervised cluster analysis revealed two distinct product groups, cluster X-AA and cluster Z-AC-AD. Unique attributes of cluster XX-AA include local manufacture, local maintenance and service, human-power, distribution by a non-governmental organization, income-generation, and application in water/sanitation or agriculture sectors. The label Locally Oriented Design for the Developing World was assigned to this group based on the dominant features represented. Unique attributes of cluster Z-AC-AD include electric-power, distribution by a private organization, and application in the health or energy/communication sectors. The label Globally Oriented Design for the Developing World was assigned to this group. These findings were corroborated by additional analyses that suggest certain design knowledge is shared across classes and contexts within groups of products. The results suggest that at least two of these groups exist, which can serve as an initial framework for organizing the literature related to inter-context and inter-class design knowledge. Design knowledge was extracted from each group by collecting known approaches, principles, and methods from available literature. This knowledge may be applied as design guidance in future work by identifying a product group corresponding to the design scenario and sourcing the related set of knowledge.

Share

COinS