Development of a Segemented Fractional Plan (SFP) to aid the implementation of Six Sigma in SMEs

Abstract

Six sigma is a statistically based, project-oriented process improvement strategy with documented successes of its usage in both large and small to medium sized enterprises. Although six sigma has been successful used by some small and medium sized enterprises (SMEs), a number of SMEs have cited the lack of resources as a factor impeding the use of six sigma. This research aims to develop an experimental plan to aid manufacturing SMEs implement six sigma at low cost. A statistical technique used with six sigma which can be resource intensive is the factorial technique used in the design of experiments. As the number of factors in a full factorial experiment increases, so does the number of experimental runs needed to conduct the experiment. This can lead to a considerably high amount of experimental runs when the factors to be studied are each represented in three levels (3-level experiments). To aid the implementation of six sigma in SMEs, this research developed an experimental plan referred to as a Segmented Fractional Plan (SFP) for fractionating 3-level full factorial experiments when 3 and 4 factors are to be studied (33 and 34 full factorial experiments). The SFP was tested using published data on designed experiments and its performance was compared to Orthogonal Arrays (OAs) using the aforementioned data on designed experiments and a laboratory experiment. The findings from the comparisons show that to identify the process setting that produces the desired product quality, with a reduced number of experimental runs, the SFP can perform as well as or better than some OAs thus, providing an option for economic experimentation when fractionating 33 and 34 full factorial experiments.

Date of Award	1 Aug 2016
Original language	English
Awarding Institution	Northumbria University
Supervisor	Philip Hackney (Supervisor)