Statistical process control, Applied Statistics

Statistical Process Control

The variability present in manufacturing process can either be eliminated completely or minimized to the extent possible. Eliminating the variability completely may not always be possible and therefore we should aim to reduce it and consistently strive to improvize the process or at the least maintain that state. The first instance of applying statistical methods to quality control can be traced back to the 1920s when Walter A Shewhart, a researcher at Bell Laboratories, USA, has developed a system for tracking variation in the manufacturing process. This technique not only provided for reducing the variation but also helped to identify the causes responsible for such variations. The methodology adopted by W A SheAwart is called 'Statistical Process Control (SPC)'. It was further developed and popularized by W Edwards Deming, who was a colleague of Shewhart. Ironically this method was first put into practice by the Japanese and not by the Americans. For the managers in USA, it was more of a compulsion to adopt this technique in the face of increasing competition from the Japanese automobile and the consumer electronic goods industries.

The variations in the manufacturing process referred above are generally studied under two heads called as random and non-random variations. The random variation is also referred to as non-systematic or common or inherent variation, whereas  the non-random variation is referred to as assignable or special cause variation. To get a better view of this let us take an example. Piston India Ltd. manufactures pistons which is an important component in an automobile. Though there are many parameters which are important and hence require a lot of attention, we consider the diameter of the piston to be most crucial as compared to others. In this case, the diameter of the pistons will not be uniform throughout. There will be at least some amount of variation in the diameter of the pistons. This variation can be due to the factors like hardness of the metal used for manufacturing pistons or errors made while taking the measurement of the diameter or else it can be due to the fact that the cutting edge of the machine getting blunt due to continuous use. If we observe, the first two reasons are not instrument specific but rather general in nature, while the third reason is instrument specific. That is, the first two reasons are said to cause random variation and the last one causes non-random variation. At this juncture  it is important to note that it is mandatory that the entire process has to be redesigned for the reduction of the random variation, whereas the systematic non-random variation can be reduced or eliminated by dealing with a specific issue, the issue being strongly related to the machine rather than the personnel who are operating it. That is, if the process is out-of-control, which indicates the presence of non-random patterns, the management should first identify the cause of that variation and eliminate it. This elimination or the reduction of the systematic variation results in the process being brought "in-control". Once this is done, the whole process can be redesigned to improve or reduce the incidence of random or inherent variability.


Posted Date: 9/15/2012 6:42:40 AM | Location : United States

Related Discussions:- Statistical process control, Assignment Help, Ask Question on Statistical process control, Get Answer, Expert's Help, Statistical process control Discussions

Write discussion on Statistical process control
Your posts are moderated
Related Questions
give me question on mean is the aimplest average to understand and easy to compute

Muti linear regression model problem An investigator is studying the relationship between weight (in pounds) and height (in inches) using data from a sample of 126 high school

If the test is two-tailed, H1:  μ ≠  μ 0  then the test is called two-tailed test and in such a case the critical region lies in both the right and left tails of the sampling distr

#quesgraphical representation of data

For each of the following situations choose the statistical model that you find to be the most appropriate. Justify your choice. a) We are interested in assessing the effects of

The 4 assumptions of regression: 1.       Variables are normally distributed 2.       Linear relationship between the independent and dependent variables 3.       Homosced

Explain the characteristics of business forecasting.

For the circuit shown below; Write a KCL equation for Node A, Node B, Node C and Node D. Write a KVL equation for Loop 1, Loop 2 and Loop 3.   A simple circ

Read the following data on the economy of Angoia and answer/respond to the questions/instructions that follow. Unless otherwise stated, the monetary figures are in real billions o

There are situations where none of the three averages is fully satisfactory. For example, if the number of items in a series is very small, none of these av