Research Projects

The demand for higher productivity, good quality products, and the designing of energy-efficient control charts in detecting process shifts are increasing day by day. Since present variables X-bar control charts are designed under conventional sampling procedures such as single sampling, double sampling, multiple sampling, etc., whose requirement of sample size (or average sample number (ASN)) to monitor the process mean and to detect any shift in the process is high they cannot be used in the costly inspection. In addition, the performance of the present control charts is measured by average run length (ARL) which partially reflects the control chart’s performance. Hence, variables X-bar control charts should be designed under special purpose sampling procedures such as multiple dependent (deferred) state (MDS) sampling, repetitive group sampling (RGS), quick switching sampling (QSS), tightened-normal-tightened (TNT) sampling, skip-lot sampling, etc. to reduce the inspection effort and performance of the control charts should be measured by median run length (MRL) since the median of the run length won’t be affected by the skewness of the in-control run length distribution. Such control charts will detect the process shift with minimum sample size and at a lower cost. Hence, the complete study on the performance of the control charts can be done by MRL. In this project, variables X-bar control charts under special purpose sampling procedures will be designed and the performance of the proposed charts will be measured by MRL. The investigation of the performance of the proposed control charts will be done by using some other performance measures such as operating characteristic function, ASN, etc. Comparison between the performances of the proposed control charts with the same existing control charts designed under conventional sampling procedures whose performance was measured by MRL will be done. The construction and application of the proposed control charts will be demonstrated by real-time data.