Product clustering for focused factories and cellular manufacturing

Show full item record

Title: Product clustering for focused factories and cellular manufacturing
Author: Batista, Judith
Abstract: In the past, clustering analysis techniques have been broadly utilized to segment markets. Recent studies (Berry et al. 1991) have used market-based clustering to examine manufacturing operations required to be competitive in different market segments. These authors proposed a methodology to categorize a set of products based on their similarity across a set of key variables, such as order-winning criteria, time considerations and volume levels. Their method incorporated the Agglomerative Hierarchy with k-Means Refinement Model to identify groups of similar products to be manufactured in a focused-factory. While many other multivariate clustering methods exist in the literature, the application of these methods to focused factory development is missing. Therefore, this thesis explores the use of other clustering methods for this purpose and incorporates the application of multivariate clustering analysis to support the formation of manufacturing cells. In the case of cellular manufacturing, existing methods rely solely on process similarity to define which products should be grouped together and manufactured in the same cell. Additional attributes, such as volume, setup time, quality and delivery lead times, have been ignored in the literature. Motivated by the need to consider multiple criteria in the formation of focused factories and manufacturing cells, and the lack of analytical methods to support the decision-maker, this research has explored the performance of five clustering methods, each with differing clustering strategies: the Agglomerative Hierarchy with k-Means Refinement, the Plant Location Model, the Covering Model, the Average-Weighted Distance Model and the Fuzzy Set Method. The use of these five methods has been illustrated using two cases drawn from the literature and results illustrate how different clustering strategies may produce different solutions. Findings in this research suggest that some sources of variability in results come from the choices made throughout the analysis process. However, since this thesis has only initiated the extension of multivariate clustering analysis to the definition of focused factories and manufacturing cells, there is a significant opportunity for future research.
Record URI: http://hdl.handle.net/1850/14258
Date: 2000-07

Files in this item

Files Size Format View
JBatistaThesis07-2000.pdf 32.66Mb PDF View/Open

The following license files are associated with this item:

This item appears in the following Collection(s)

Show full item record

Search RIT DML


Advanced Search

Browse