SPC and Acceptance Sampling - Assignment Example

Task SPC and Acceptance Sampling In order for a producer to ensure that products adhere to and exceed the set guidelines in terms of product usefulness, timeliness, reliability, durability and other quality indicators, then the processes of achieving the final product must be modified accordingly. For instance, the element of chance must be eliminated such that a product conforms by design rather than by chance. The main approaches to reducing the element of chance are by educating employees on the components or quality assurance, and informing them of their role in the quality assurance process. Typically, the quality assurance process involves six main processes including testing of previous products, planning to make improvements, designing processes for implementing improvement, production of new product, reviewing the new product, and testing the new product. However, due to the high number of volumes produced by a firm, the manufacturer must design economic ways of reviewing and testing products, and minimization of variability in product quality. Statistical process control is used as the main tool in ensuring that most of the products fall within the usability range and error minimization. According to Rajput (2008: 870), the producer must design a process by which to test items for acceptability without having to study the whole batch of produced goods or services. This essay reviews the use of statistical process control and acceptance sampling as quality assurance tools for employees in a small manufacturing firm. Statistical process control (SPC) SPC Tools Leitner (2007: 15) states that there are three main tools used in SPC for data collection and interpretation including: Sampling: all items in a batch cannot be tested for adherence, as it would be laborious and expensive; a producer must decide the criteria to be used to decide how much a sample should contain, and the frequency by which such samples should be taken. The volume of production determines sample size while sampling frequency depends on the time that production goes on as well as the level of desired conformity. Operational definition: after the sample has been collected, there has to be a measure of attributes, measuring equipment, levels of conformity to standards, and criteria used to make decisions on the suitability of the product. Anova gauge R & R: this is the tool used to measure, improve and ensure the repeatability and reproducibility of levels of adherence to stipulated quality standards. In this sense, quality is assured by the capacity of people, processes, equipment, and products to retain, sustain and improve desired functionality consistently. Measures of Deviation According to Leitner (2007: 28), a producer must maximise conformity and reduce the rate of rejection of products on grounds of poor quality; the producer must work such that the products adhere to certain minimum conditions. First, the levels of adherence to quality standards should fit in a normal distribution curve, such that most products are within range and as few as possible are on both extremes; poorer quality and better quality than required. These instances of extremes are as a result of production processes affected by chance, which should be reduced as much as possible. The level of standard deviation in the normal distribution curve should be as minimal as possible to ensure that product quality is similar with little variation among them. Distribution Curves The most commonly used distribution curves are noncumulative discrete and continuous distribution curves (Wheeler, 2011: 298). The most suitable distribution curve is the normal distribution; if product quality fits in a normal curve it means that the firm is on the correct path to achieving total quality management. However, many factors must be considered when interpreting distribution curves including range, average, and standard deviation among others. The range of the data indicating product conformity should be minimal, which means that products must have as little differences between them as possible. For instance, the firm should define the maximum difference in length or weight of products, which should be as small as possible depending on the level of quality assurance the firm intends to achieve. The average of product qualities should be an indicator of the level of adherence to standards, such that the value should be equal to the standard. In this case, measurement values vary such that the products that do not meet standards equal those that surpass the standards, and the majority of products are in between. Maintenance of Control According to Wheeler (2011: 374), SPC uses charts classified as control charts for variables and control charts for attributes. Control charts for variables are P charts and C charts while those for attributes are X bar (average control chart) and R charts (range control chart). Before charts are used on a process, upper control and lower control limits are set, and a sample collected. After analysis, if the sample does not fall within the upper and lower control limits, then it is said to be out of control. The P chart indicates the frequency of defects in a system; the C chart indicates the number of defects per unit; and the X bar and R in complementary with each other. Acceptance Sampling According to Schilling and Neubauer (2009), the main aim of acceptance sampling is to accept or reject items in a batch based on results of analysing a sample before they are entered into the production process. The main tool used in acceptance sampling is the OC curve (Operational Characteristic Curve), which serves as an indicator of conformity of products to given standards. The OC curve is a plot of probability against lot of acceptance, with the two values being inversely proportional to each other in an inverted sigmoid curve (in most cases). Before the chart can be used, some variables must be defined including Rejectable Quality Level (RQL), Acceptable Quality Level (AQL), lot size and sample size. RQL is the statement of the quality that one would like to reject regularly, and AQL is the statement of the poorest quality that one is likely to accept. In normal cases, if 10 -95 per cent of probability of acceptance (y - axis) does not lie within the difference between AQL and RQL, then the lot is rejected in its totality; otherwise, it is accepted. Conclusion Quality assurance must be evidence-based practice, such that the results of the production process are the main indicators of the effectiveness and efficiency of the process. Therefore, control charts must be used to ensure that products and processes conform to quality standards, and the process is repeatable and reproducible with a high level of consistency. Since testing cannot be done on all the produced items, samples must be picked, and OC charts are valuable tools in deciding the best sampling method in acceptance sampling. Task 2: QA Charts and Interpretation Samples were collected for processes in a firm, with a specification of 60.0 2.0; values were plotted into a graph and analyzed for process feasibility; values were plotted into graphs and analyzed for process feasibility. Figure 1: Mean graph of process capability This graph indicates the averages of values obtained from sample processes in a production firm. Since the values should be between 58.0 and 62.0, all values that lie below or above this range should be rejected; and the remaining processes analysed further for feasibility. In this case, there were no values rejected for being below the required range of values; however, the ones that fell beyond the values should be considered since the firm may targets that are lower than necessary. Therefore, this graph is not conclusive in its analysis, and the graph of range values for each of the samples is required to ascertain if the company goals are suitable, or need to be adjusted to keep up with company capacity and market requirements. Figure 2: Range graph for process capacity This graph gives process variability for the samples that were analysed. A high range of values is indicative of an unstable process whose quality of production cannot be assured. For instance, the company is required to make a positive or a negative error of 2; this means that the maximum range should be 4. However, since the company has the intention of improving dependability and repeatability, the value of 4 is too large, and any process that comes close to or goes beyond this value of range should either be rejected or modified to reduce this variability. Analysis Process analysis using mean and range involves the consideration of these two values in accordance to the requirements of the quality assurance plan. The mean should not be used in isolation since a process may have the required mean and not meet the standards, as the mean may be a result of a balance between the values that went beyond the required mark and those that did not meet the specifications. Range is the tool used to ensure that every process will result in products that as similar as possible in terms of utility, timeliness, suitably, and other characteristics as specified in the total quality management plan. While the processes that meet the stipulated standards are accepted, those that fall short are either improved or rejected in their totality. References Leitner, A. (2007) Statistical process control, GRIN Verlag. Rajput, R.K. (2008) A textbook of manufacturing technology: manufacturing processes, Firewall Media. Schilling, E.G., & Neubauer, D.V. (2009) Acceptance sampling in quality control, CRC Press. Wheeler, D.J. (2011) Understanding statistical process control, Spc PR. Read More