An experimental investigation of composite priority rule scheduling for quality improvements in multi-stage production systems
MetadataShow full item record
Considerable research has been conducted in the area of sequencing and scheduling of manufacturing operations. There has been a great deal of research work on investigating the performance of various priority rules. Previous studies have concluded that in general composite priority rules perform better than some of the well known simple rules. A combination of two or more simple scheduling rules is defined as a composite rule. Many different performance measures have been considered by previous researchers to investigate the effects of priority rules. Recently in the manufacturing industry and in the literature, integrating quality control activity at every stage of production process has been gaining attention. The objective of this research is to investigate the effect of priority rules on detecting defects and hence improving the overall quality levels of the production processes. An analysis of the effect of combination of simple rules such as Last-In-First-Out (LIFO) and First-In-First-Out and combination of LIFO and a due-date-based Job Slack (SLACK) rule is performed. The principal advantage of adopting the composite priority rule is that by periodically selecting the next job using the LIFO rule, a defective item produced by the preceding station can be detected shortly after it leaves that station. Computer simulation was used to analyze the problem. Three independent variables were considered for the statistical experiments. These are elapsed time for the process to be "in-control"(3 levels), system determined queue length (3 levels) and the frequency of switching between priority rules (3 levels). LIFO performs best in improving the fraction "good" items produced. LIFO is followed by SLACK and FIFO in that order. FIFO performs best when the variance of waiting time is used as the performance measure. LIFO performs worst. The combination of FIFO-LIFO produces variance that fall between these extremes. Thus a switching policy that maximizes the benefits of improving quality levels can be identified for different problem situations. The best switching policy is identified for one example problem. Based on the analysis, it can be concluded that the best switching policy depended upon the weights assigned to the cost parameters. General models are developed which enable the identification of the best switching policy.