PRACTICAL APPLICATION OF SIX SIGMA METHODOLOGY TO REDUCE DEFECTS IN A PAKISTANI MANUFACTURING COMPANY

All over the world six sigma is being adopted as a quality improvement approach towards zero defects. Unfortunate-ly, the adoption of six sigma methodology in manufacturing companies is very rare in developing countries due to various challenges. This study demonstrates the practical use of the six sigma Define, Measure, Analyze, Improve and Control (DMAIC) cycle by conducting a case study at a manufacturing company in Pakistan. The potential problem was the external leakage defect in the refrigerator during its production stage. The objective of this study was to improve the process by adopting six sigma DMAIC approach to identify and eliminate the root causes that produce defects in the final product. Project charter, Pie chart, Bar chart of faults, Suppliers Input Process Output (SIPOC), Voice of Customer (VOC) and flow process map were used to define the problem, its scope and process routing. Pareto chart was used to identify sub defects and sigma level was calculated for the existing process. Feed rate, capillary action of filler material, cleanliness, visual inspection and unsuitable heat input were the major causes of external leakage. Cause and effect matrix was used to rank the identified causes. Further, design of experiment (DoE) was performed to improve the process by conducting different alterations in the parameters. In order to control the process, failure mode and effect analysis (FMEA) sheet was prepared to sustain the process improvements. The FMEA control plan needed to be revised at specific time intervals to attain continuous process improvement. This six sigma DMAIC cycle produced a 30% overall reduction of external leakage defect and service call rate (SCR) was improved with lower complaints from customers.


INTRODUCTION
The concept of six sigma was presented by Motorola and General Electric (GE) in the 1980's as an innovative set of management techniques to assist both companies. Both companies benefited with the use of newly innovative technique by improving their production efficiencies and better control over the processes. Since the successful implementation and better performance of six sigma in both companies in 1999, numerous other companies adopted this technique as a part of their management strategies [1]. Six sigma is now a widely used and well accepted method which can eliminate root causes of the unwanted variations and provides good control of the process. The main focus of six-sigma is on process control, continuous improvement, customer needs, statistical analysis and business growth [2][3][4][5][6]. Most widely used approach of the six sigma is DMAIC cycle i-e Define, Measure, Analyse, Improve and Control is a five-step six-sigma improvement model adopted by manufacturing and service organisations to improve their current capacities of existing processes [7][8][9][10][11][12]. There are variety of the supporting tools which can be used at various phases of the six sigma methodology. Table.1 shows the most frequently used tools under six sigma DMAIC cycle. Practical Implementation and use of six sigma methodology is still challenging for manufacturing companies in developing countries. Considering the fact, Pakistani manufacturing industries are also lacking in the adoption and use of six sigma tools and techniques. Thus, this study had made an attempt to show the practical use of six sigma methodology by conducting a case study at a manufacturing company. Various tools and techniques were used through different phases of the six sigma DMAIC cycle.

CASE STUDY
Dawlance is a private limited company which is one of the top electronic and premier home appliance manufacturers in Pakistan having many products. Products of the company are refrigerator, washing machines, freezer, split air conditioner, microwave ovens. Due to the increasing rate of defects from the calculation of service call rate (SCR), company wants to reduce defects to sustain its competitive position in the market.

METHODS
To solve the problem of external leakage in refrigerator Model No: 9166, the project team had used the six sigma DMAIC approach for the improvement. The various tools and techniques that were used under the six sigma DMAIC cycle are shown in Figure. [18] tools and techniques include cause and effect analysis, process mapping, histogram, control chart, process capability analysis, FMEA and Poka-yoke. It seems that tools that offer graphical representation, recognize root causes of bottlenecks are easily adoptable as compared to more complex and sophisticated tools and techniques [17]. Thus, study has selected the most recommended and widely adopted tools and techniques.

Define Phase
This phase defines the problem, process mapping, scope and effect of the problem for both external and internal customers. The tools used by the project team in this phase are Project charter, SIPOC, VOC, flow process map, trend chart. The project charter shown in Table.   To understand the main root causes of the problem of external leakage associated with Brazing stations, a flow process map was made by six sigma project team with consultation of the internal experts shown in Figure.3.
To have more in depth understanding of the whole process and its different elements SIPOC was prepared as shown in Table.3. Further voice of customer (VOC) was established for both internal and external customers of the company considering their requirements. The six sigma team made the VOC with coordination of the quality department and by communicating with the external customers is shown in Table.4.

Measure phase
The purpose of Measure phase was to identify the current performance level of the process and its sigma level. The data collection was done to measure the process inputs and outputs. The Pareto analysis was used in this phase to prioritize the sub defects and determine the most critical defects that were causing the company to pay for the cost of poor quality. The team used sigma level calculations to find out the sigma level of the process and also to set the target of increasing sigma level. To study the defects and sigma level calculations, around 12 months data was used. 3. Joint the suction line.
Completed joint of capillary with filter-drier, joint of suction pipe and the joint of suction line on the refrigerator body.   Pareto analysis is shown in Figure.4, it was observed that compressor oil joint leak, compressor charging joint leak, bottom condenser and condenser dryer joint leak were most serious sub defects that were causing the external gas leakage in the refrigerator.

Analyze phase
The analyze phase helped in identifying the main root causes and their effects. Thus, in this phase team had used a cause and effect diagram which is one of the basic and remarkable quality tools in assessing the real root causes of the problem is shown in Figure.5. To rank the root causes once identified, cause and effect matrix was used as shown in Table.5. The project team sat along with experts and carefully assigned the weights to each cause based on expert's opinions. Details of the experts are shown in Table.6.

Figure 5: Cause and Effect diagram
Results of cause and effect matrix showed that improper capillary action of filler material, feed rate, brittleness of filter drier joint, cleanliness were the major causes that resulted in external leakage defect in refrigerator.

Improve phase
The Analysis phase helped the team to identify the inputs that might affect the strength of the joint which ultimately caused the failure. Thus, in this phase to optimize the process performance and improve process quality, Design of Experiment (DoE) was performed using Minitab software. The experiment was aimed to investigate the effect of four input factors (Feed rate, pressure of oxygen, pressure of acetylene, filler material) on the output strength of joint (Measured in Mega Pascal). These input factors might affect the output i-e strength of the joint, so the experiment is designed to find out which are the most important input factors and to quantify their effect. A "2-level full factorial design" approach of DoE is used. Process specifications and 2 level experiments for brazing process and are shown on Table.7 and Table.8 respectively. Each level represents one trial and total sixteen trials were run during the experiment as shown in Table.9, there were four input factors and each experiment was set at two different levels. The Pareto and Normal Effect Plots showed the main input parameters which have a statistically significant effect on the process        The main effect plot shows that the average joint strength increased from around 101.5 to 103.5 when the feed rate was increased from 4 to 6 seconds. The average joint strength also increased from 100 MPa to 105MPa when copper filler material was used instead of brass.

Figure 8: Main effect plot
The results of the design of an experiment on Brazing process can be seen above. The experiment concludes

Control Phase
After the identification of the root causes of the problem and their possible solutions it was time to standardize the process and sustain the changes that were proposed. The team made the use of the FMEA approach to create standardized documents for monitoring and improving the process by integrating it in daily production activities as shown in Table.10. Major actions recommended to improve the process were Inspection by mirror, brazing with brass, cleaning of tube with nitrogen at 5bar pressure and standard time per joint in order to control the process continuously.

CONCLUSIONS AND FUTURE RESEARCH DIRECTION
Six sigma is a statistical, systematic and data driven methodology for reducing the defects rate of any process. In this study, six sigma DMAIC approach was adopted in a manufacturing company to reduce the defect rate of external leakage in the refrigerator during its production process. Improper capillary action of filler material, feed rate, brittleness of filter drier joint and cleanliness were the major causes that resulted in external leakage defect in the refrigerator. Major actions recommended to improve the process were Inspection by mirror,