Six Sigma and Lean

Everything is a process that can be defined, measured, analyzed, improved and controlled. The underlying proposal of six sigma is continual process improvement. Lean is to remove waste (processes and procedures that add no value).

DMAIC- Define, Measure, Analyze, Improve, Control.

Statistical Process Control (SPC)

Control Charts

Failure Mode and Effects Analysis (FMEA)

Process Mapping

 

Sigma Level

Sigma level is a metric that is utilized to allow one to visualize the process capability to meet the process’s requirements.
     The sigma sign is the representative of standard deviation. At sigma level 1, the percentage of defects is 69%. At sigma 6, it is 0.00034%.
      Organizations aim for the Six Sigma level.
              DPMO, or defects per million opportunities, is used to calculate the sigma level for a product or process.
                  The higher the DPMO, the lower the Sigma Level (Lower DPMO=Higher Sigma, Higher Yield)

Process Capability and Process Performance

 Cp= Process Capability
Cpk= Process Capability Index. Adjustment of CP for non-centered distribution
Pp= Process Performance
Ppk= Process Performance Index. Adjustment of Pp for non-centered distribution

Cpk is a measurement to determine how close to the target value you are and how consistent. This is a measurement of how close the product/process is to running at is determined specification limits. The larger the index is, the less likely any process will be/produce product outside of the limit.

The Cpk value determines how far you are before you reach the limit. Anything less than one is outside the limit, 1 is at the limit, 2 is halfway, 3 is third.. and so on.

Process performance is used to determine the same thing as process capability, however Pp is generated from a specific batch. Pp is only to be utilized when Cp is unavailable.

The difference between these two- process performance will let you know the variation in your process, however process capability will allow you to theorize how the variation affects your overall yield.

You are revalidating a process. This process is run on multiple lines over multiple shifts. You run concurrent validation on each line. The validation data for each line would be Pp, the overall data for all lines run would be Cp.

Pp = C_p = (USL – LSL)/6*Std.dev  
                              This estimates the process capability if the mean is centered between the specification limits ~ normal distribution

Cpk uses R-bar/d2 
         Cpk= min  [(USL-mean)/3*Std.dev, (mean – LSL)/3*Std.dev]
                               This estimates the process capability with non centered mean ~ normal distribution

To estimate process capability around a target, divide the C_p (Can be same value as Pp) by [square root of 1+((mean – target divided by Std. dev)squared)]
         Substitute Cpk for C_p  to determine capability around a target for off centered mean.

Recommended values for a process minimum is usually 1.33 (Sigma 4, 99.99%), with a Six Sigma quality process at 2.00 (Sigma 6, 99.9999998%)