Six Sigma Project Case Study Improved Productivity of Fans Through Reduction of Defects/Defectives at Spindle Manufacturing
Six Sigma Project Case Study Improved Productivity of Fans Through Reduction of Defects/Defectives at Spindle Manufacturing
Preamble/Project Selection
High Loss of Production of Fans due to short supply of Critical Components Spindle was an important component. It was not possible to augment the production of However, it was observed that, rejection of spindles was very high (6.5%) If rejection is brought down, it will have great impact on business in the form of higher productivity of fans and reduced losses at spindle manufacturing Hence, management initiated a Six Sigma project at spindle manufacturing
The Team
Sponsor: GM, components Manufacturing Leader : Manager (QC), a Black Belt Team Members: Two Green Belts and one supervisor Trainer & Guide : Consultant
Scope:
All types of Spindles manufactured
A high level process map was drawn Schedule for the project milestones was drawn using a Gantt diagram. Target
It was decided to achieve a five-fold improvement in rejections. The current level being 6.5%, the target was set at 1.3%
Defect
Any dimension which is high or low is considered as a defect. There were as many as 19 dimensions. Hence number of opportunities is 19
Measure Phase
Base line Data Past 1 months go-no-go inspection data was considered. The number of spindles inspected, the total number of defective spindles and the number of defects due to each dimension were noted down. Total no. inspected : 17,000 Number of defectives : 1130 Total no. of defects : 1350 P = 0.065 Yield = 93.5% dpmo = 4180 Sigma Rating for product: 3.01 (Z= 1.51 for p=0.065) Sigma for defects: 4.13 (based on dpmo) Target p = 0.013 Target sigma rating for product: 3.72
Analyze Phase
Pareto analysis was carried out with the Data on defects. Four dimensions accounted for over 90% of the defects (A group). These were selected for further study. There were 8 dimensions with negligible defects (C group). Remaining 7 dimensions (B group) were average. For each of the selected dimensions, data on actual measurement were collected for 150 components. Histograms were drawn The histograms suggested that the process capabilities were good. However, there was severe lack of control in 2 cases, improper setting in one case and shift in average dimension due to tool wear out in the fourth case.
Improve Phase
1. Improvement suggestions for A group items: For cases where lack of control was evident, control chart was recommended. ( X R ) For the case where there was improper setting, first off inspection (approval inspection) was recommended For the case where there was tool wear out, statistical analysis (regression analysis) was carried out and optimum tool grinding/ tool replacement were recommended
2. For B group items maintenance of p/np chart shift wise were recommended