Manufacturing Process Optimization to Increase Product Reliability by Control Charts

The need to produce highly reliable products is essential for the survival of many industries in today’s world. Consequently, one of the main quality characteristics of products is their lifespan or reliability. The production process must be monitored to produce highly reliable products, and in this regard, control charts are one of the widely used process monitoring tools. In this study, the lifetime of the products, which is monitored as a quality characteristic, follows the Weibull distribution with a variable scale parameter and a fixed shape parameter. Monitoring the average lifetime of the products is conducted by examining the Weibull distribution scale parameter. To design a control chart, first, the control limits of the chart for different parameters are obtained and presented. Relationships and values regarding control limits show that the control limit of a one-sided chart is always larger than the control limit of a two-sided chart. For evaluating the performance of the proposed control chart, the average run length in the out-of-control mode for various parameters is presented, and sensitivity analysis is performed. The computational results show that the one-sided control chart has a better performance in detecting reduced lifetime compared to the two-sided control chart. Failure in the type II censoring life test is an important parameter that increases the performance of the control chart. As the number of failures increases, the test continues for a longer time, which causes the chances of detecting the changes to be improved. Finally, to show the performance of the proposed control chart, a practical example is provided.