Polyhedrika

  Statistical Process Control (SPC) is used to control a critical process output acting on the factors affecting it only when required. The purpose is to differentiate  between a statistically significant change in the process and common cause variation because the actions required will be different. Real Time Statistical Process Control enables the operators running the process to know when action is required, as described in    Real Time SPC There are two extremes to avoid:            Over-reaction (adjusting after every output) makes the process worse: dispersion is increased. Under-reaction is just as bad: a slow process degradation passes undetected until it is too late. SPC Charts for Variables and Attributes We will now analyse some of the charts used for variables and attributes. SPC Chart Interpretation SPC uses some rules, developed by the Western Electric company, to detect symptoms of  statistically significant  variation. The center line is the average of all values and th

  Process variation analysis is essential in order to improve it. There is always variation. Sometimes we only have random variation such as in the case of dice throwing: https://polyhedrika.blogspot.com/2020/12/dice-throwing-exercise.html   Normally we have additional variation caused by special causes and these are the ones we should analyse and reduce first. Hourly Production  We are going to analyse the hourly production data collected during one month: Download this Excel file Trend Analysis.xlsx from OneDrive folder Process stability SPC charts allow us to check for stability: see if all variation is random. https://polyhedrika.blogspot.com/2020/12/spc-analysis.html We now apply an individuals SPC chart to our production data to check this: The green and red dots are a clear indication of lack of stability: there are special causes affecting production. Timestamp Data Collection When we collect process data either manually or automatically it is essential to collect the timestam