Statistical Process Control

 

5.	Statistical Process Control Statistical Process Control (SPC) is a widely used sampling technique which checks the quality of an item which is engaged in a process. SPC can also be used to inform management of improved process changes (Krajewski et al., 2010). SPC identifies the nature of variations in a process, which are classified as being caused by ‘chance’ causes or ‘assignable’ causes.
	5.1.	Chance Causes of Variation Processes will have some inherent variability due to factors such as ambient temperature, wear of moving parts or slight variations in the composition of the material that is being processed. The technique of SPC involves calculating the limits of these chance-cause variations for a stable system, so any problems with the process can be identified quickly. The limits of the chance-cause variations are called control limits and are shown on a control chart, which also shows sample data of the measured characteristic over time. There are control limits above and below the target value for the measurement, termed the upper control limit (UCL) and lower control limit (LCL) respectively. The behaviour of the process is observed by studying the control chart and if the sample data plotted on the chart shows a random pattern within the upper and lower control limits then the process is ‘in-control’. However if a sample falls outside the control limits or the plot shows a non-random pattern then the process is ‘out-of-control’.
	5.2.	Assignable Causes of Variation An assignable cause of variation is a variation in the process which is not due to random variation but can be attributed to some change in the process, which needs to be investigated and rectified. However in some instances the process could actually be working properly and the results could have been caused by sampling error. There are two types of error which can occur when sampling from a population. A type I error is indicated from the sample output when none . A type II error is when an error is occurring but has not beenaactually occurs. The probability of a type I error is termed. Type I errors may lead to rectification workbindicated by the sample output. The probability of a type II error is termed  which is unnecessary and even the unnecessary recall of ‘faulty’ products. Type II errors will lead to defective products as an out-of-control process goes unnoticed. Customer compensation and loss of sales may result if defective products reach the marketplace. The sampling methodology should ensure that the probability of type I and type II errors should be kept as low as reasonably possible
	5.3.	Types of Control Charts   Two types of control charts are for variable data and for discrete data.. Control charts for variable data display samples of a measurement that will fall in or out of a range around a specified target value. Examples of variable data could be a customer transaction time in a bank or the width of an assembly component. Two control charts are used in measuring variable data. An X,–– chart shows the distance of sample values from the target value (central tendency). An R chart shows the variability of sample values (dispersion). Attribute control charts measure discrete values such as if a component is defective or not. Thus there are no values, as in a variable control chart, from which a mean and range can be calculated. The data will simply provide a count of how many items conform to a specification and how many do not. Two control charts will be described for attribute data. The p-chart which shows the proportion of defectives in a sample and the c-chart which shows the number of defectives in a sample
		-	ISO 9001	applies when the supplier is responsible for the development, design, production, installation, and servicing of the product.
		-	ISO 9002	applies when the supplier is responsible for production and installation
		-	ISO 9003	applies to final inspection and testing of products.
		-	ISO 9004	provides guidelines for managers of organisations to help them to develop their quality systems. It gives suggestions to help organisations meet the requirements of the previous four standards.
		The standard is general enough to apply to almost any good or service, but it is the specific organisation or facility that is registered or certified to the standard. To achieve certification a facility must document its procedures for every element in the standard. These procedures are then audited by a third party periodically. The system thus ensures that the organisation is following a documented, and thus consistent, procedure which makes errors easier to find and correct. However the system does not improve quality in itself and has been criticised for incurring cost in maintaining documentation while not providing guidance in quality improvement techniques such as statistical process control.