Abstract
LED is one of the most competitive products for Optoelectronics industry in Taiwan. Owing to the impact of global competition, how to provide high-quality, low price and varied products is the most important issue. The enterprises must meet the customer’s needs and create the highest profits and economic value under the limited resources. Since 1995 the GE implemented Six-Sigma approach that the company got very high growth rate, so the approach was paid attention to the quality scope. The spirit of the Six Sigma is to utilize statistical method to improve the quality and competition for enterprises. The study focused on improvement of LED process, which considered the customers needs to get the key factors of process by applied Six Sigma program (DMAIC), relative quality tools and Statistic analysis approaches. Finally we used the real world example to certificate the performance. The define phase is clearly to confirm the objective of the project which improved the backed process yield of LED, then to set up procedure of the project. The measure phase is to collect the data of main characteristics. In the analyze phase, we applied the Pareto diagram and causes & effect chart to get the key factors which influenced the process. In the improve phase, we utilized the DOE to find the best parameters of probing process. Finally the control phase is to involve the best parameters to the quality system. In this study based on the internal and external customers’ needs, according to evaluation method of project and executives meeting to select the project theme is “reduce the defect rate of abnormal probing“. In defining phase, by systematic observation, we found out the main cause of abnormal probing that is the defective appearance of chip. Then we used the high-level and detailed flow chart to get the cause of abnormal process, and set improvement targets, to reduce the appearance of defect rate to 4.14% for the AlGaInP product probing . In the measure phase, according to the status to collect data, identify key quality characteristics for the appearance defect rate of probing chip, as a basis for improvement. In the analysis phase is the use of Pareto analysis and Cause & Effect Diagram to summarize the key factors that affect the process, then according to the feasibility assessment to identify the causes for high and easy to solve, the materials and angle of probing needle. In the improvement phase is to use experimental design to identify the best factors of enhancing probing yield. The best factor is that the probing needle is made of alloy, the angle of probing needle is 75 degrees, and validated the average of defect rate from 5.32% reduced to 3.84%, the improvement rate is 125%. In the control phase is standardized the best factors and included in process control item, and then complete this project. Keyword：LED( Light Emitting Diode ). Six Sigma. DOE( Design of experiments ). AlGaInP