A NEW EWMA SCHEME FOR JOINT MONITORING OF PROCESS MEAN AND VARIANCE.

Quality monitoring has become an inevitable business strategy in all most all the sectors such as manufacture, transport and service (Montgomery, 1996). The dimensions of quality are performance, reliability, durability, serviceability, aesthetics, features, perceived quality and conformance to standard (Garvin, 1987). The quality parameters which are often called quality characteristics are several elements that jointly describe the quality of a product in the view of consumer. All the business want to supply the products identical from one unit to other in terms of quality characteristics and it is the expectation of the customers (Gravin, 1987). This task is very expensive and difficult because of the unavoidable variability present in the quality characteristics. To monitor and control these variations among the quality characteristics, statistical process control tools were introduced (Montgomery, 1996). These tools are histogram, check sheet, Pareto chart, cause and effect diagram, defect concentration diagram, scatter plot and control chart. Among all these tools, control chart is most favored in the industry because, control charts help to identify the variations in a very meaning full way that process variation attributed to assignable causes is distinguished from unassignable causes (Shewhart,1939). The first quality control chart was introduced to the world by Shewhart (1939) for controlling sample mean of a quality characteristic. Since then control charts becomes an essential tool in quality monitoring. Subsequently advanced control charting schemes such as cumulative sum (CUSUM) and exponentially weighted moving average (EWMA) charts were introduced by Page (1954) and Roberts (1959) respectively. These advanced charting schemes have the ability to detect small shifts in quality characteristics but it is not as sensitive as the Shewhart chart when the shift is large (Gan, 1995). Shewhart, EWMA and CUSUM charts were first proposed to monitor the process mean alone and later these charts were altered to monitor the process variance alone.