Process Capability Overview-Tamil
Process Capability Overview-Tamil
by Sharon K. Rabeneck
The principle associated with all process capability indices is to relate the
performance of a process to the requirements for that process. There will be
adaptations based on long or short term and centering, but the principle remains the
same throughout all of the applications.
The reason we need to have 6 sigma between the process target (nominal) and the
closest specification limit is to preclude the production of defective material (3.4
PPM) if the process shifts 1.5 sigma.
Process shift from short to long term has the effect of widening the
distribution. Until we have actually measured process shift and drift, and have
specific information to the contrary, we assume it to be 1.5 sigma. In fact, historical
empirical information supports the assumption of drift = 1.5.
Process Capability cannot be improved by adding inspection. If the probability
of detecting a defect is 80% and there are 6 consecutive inspectors, about 64
escapes would be expected for each 1,000,000 defects produced.
If the probability of detecting a defect is 80%, it would require more than 7
consecutive inspectors to have a 6s level of confidence that a defect will be detected.
The Standard Normal Transform
The standard normal transform is the method we use to determine the location of
any point in a normal distribution with respect to the mean of that distribution. We
encounter this transform in all phases of projects and the Breakthrough Strategy.
The standard normal transform converts any normal data to the standard
normal so the Z table can be used. If we know Z then we can calculate s or vice
versa.