| This article comes from a real problem which “A” company must dissolve. When theytest Ano product surface color with routine methods, the inspection data is not sameas visual feeling. Because there is no good experience before A company’s study, Acompany have to study the project based on textile experience. In the color controlmethod improving process, A company use six sigma DMAIC method to control theirstudy process.In the define process, use the general method, such as brain storming and fish bone toanalysis what factors will influence the color inspection. For inspection, we can findfactors from Human, machine, material, method and surrounding; for color, we canfind factors from light, target and observer. In this article, we find the most importantfactors are human, gage, parts surface and color space. In the measurement process,we qualify upper important factors, such as test the human threshold, gage R&R, partcolor distribution and how to select color space. In the analyze process, we use thedata from measurement to find the relationship between some factors and find the rootcause of inspection deviation. Based on the visual threshold and GRR, we can findwhy sometimes gage data is not same as visual feeling. We also build a model toanalysis the relationship between our parts physical scale and color distribution. In theimprovement process, we build a new color model to show the relationship betweenLab data and visual feeling and use some correction factors to correct the influencefrom uneven color, mottling and cut mark. In the control process, we lead to newcontrol method to the plant and record the new control method performance. Wefound the control rate rise dramatically.This article use many tools to analysis the method defects root cause and build modelto analysis problems and dissolve problems. These tools can also be applied in other projects. On the other hand, this project is a good sample to show that the six sigmaapplication in a analyze method study. |
PDF Full Text Request